1. SUMMARIZING VASUBANDHU: SHOULD A BUDDHIST PHILOSOPHER HAVE A PHILOSOPHY?
1. I do not really believe in any strict calculus of the “greatness” of great philosophers, but this is a statement I like to make to challenge the casual certitude with which the top position is so often claimed for Nāgārjuna. Nāgārjuna definitely had a masterly philosophical idea in his Madhyamaka defense of the doctrine of emptiness, and he presented it with elegance and consistency. But to use the famous terminology of Isaiah Berlin, Nāgārjuna was a hedgehog—he knew one thing very well, and united everything under a single vision—whereas Vasubandhu, whose insights transformed every arena of philosophical investigation, was a fox—the one who “knows many things.” Though, if the thesis of this book is correct, then perhaps Vasubandhu was, as Berlin says of Tolstoy, “by nature a fox, but believed in being a hedgehog” (Berlin 1978:24).
2. See Takakusu 1904 and Butön 2013:241–245.
3. There were at least two Vaibhāṣika commentaries on the AKK.
5. On Tibetan doxographies, see Cabezón 1990, Hopkins 1996, and Harter 2011. Cabezón explains how the classification of distinct philosophical schools according to the siddhānta (Sanskrit) or sgrub mtha’ (Tibetan) scheme has had significant effects upon Tibetan readings of Buddhist doctrine, since it came to define what counted as Buddhist for Tibetans, and had an even greater authority than the classification of what texts were included in or excluded from the canon. I am arguing that we reflect this classificatory rigidity when we allow our interpretations of Vasubandhu to be stunted by the almost universal, but anachronistic, reliance on these categories. See Trivedi 2005 for a similar claim.
7. In describing Vasubandhu’s view, I start by explicating the philosophical articulation of the core doctrinal concerns of the AKBh. Since it is not possible to lay out every compelling philosophical argument in the AKBh, I have chosen to begin with a series of linked arguments that are characteristic and representative, but that, most importantly, are aimed at refuting what Vasubandhu calls the distinctive doctrine of the Vaibhāṣika tradition, namely the view that the three times all exist (the source of the name “Sarvāstivāda”). These arguments make use of a number of other points from the AKBh, which I also elucidate when they come up. Since one of these references is to the famous ninth chapter, on the “Refutation of the Person” (see Gold 2011), I am able to establish these many arguments, taken together, as representing the key concerns of the author of the AKBh. The next chapter is a study of three passages that, together, fill out the motivational and theoretical implications of the views studied in the “three times” chapter. With the AKBh thus established as gravitational center, we can then measure the ways in which other works attributed to Vasubandhu are in its orbit.
9. When Frauwallner claims that Paramārtha or his students could have been responsible for the integration of these two narrative strands, he is forgetting that Xuanzang, Butön, and Tāranātha all provide “combined” versions—none of which can be traced back to Paramārtha. If there was an erroneous combination, it had to be quite early, and Indic.
10. Still, Anacker (1998:24–25, n13) says that the Chinese gives indications that the author was aware of telling the story out of chronology, so this is only an error if there were, indeed, two separate Vasubandhus.
11. Jaini 1958:50ff. As one example, with regard to the view that claims that the meditative state called the “attainment of cessation” (nirodhasamāpatti) still has a mind, which the AKBh does cite but attributes to Vasumitra, the commentary says it is “non-Buddhist” (abauddhīyam). See Griffiths 1991:122–128 for a translation of the AKBh passage and a study of this issue, which has been claimed to be one of the central motivations for the famous Yogācāra assertion of the “storehouse consciousness” (ālayavijñāna). See Schmithausen 2007 for that argument. If it is right, then it does seem that to claim Vasumitra’s view “non-Buddhist” blocks this particular Mahāyāna solution, and thus cuts off a path to the Mahāyāna that the AKBh had left open.
12. Frauwallner 1961:131–132. Frauwallner writes that this similarity in their biography makes their confusion all the more likely.
13. Here I am replicating the judicious reasoning of D’Amato (2012:40–41).
15. Hirakawa et al. 1973.
18. For example, see the connection forged between the VyY and the MSABh in Horiuchi 2009. Although his analysis is in a somewhat different vein, I would specifically mention Corless (1989) here as an example of an author whose vision of philosophical unity for Vasubandhu I am extending in my work. As another recent example, Braarvig (1993:cxxix) suggests that although it is unlikely that Vasubandhu wrote the Akṣayamantinirdeśasūtra commentary attributed to him (the Akṣayamantinirdeśaṭīka) himself, because it contains quotations from works that postdate Vasubandhu, including works by Sthiramati, the commentary cites the AKBh in a way that suggests that the author of the Akṣayamantinirdeśaṭīka associated Vasubandhu the author of the AKBh with the early Yogācāra. Braarvig thinks the commentary may have been begun by Vasubandhu and completed by Sthiramati.
19. Kritzer 2003a, 2003b, 2005. Here also Park (2007) must be noted as a gradualist counterpoint to Kritzer, and a strong supporter of continuity in Vasubandhu’s oeuvre. I am very much indebted to Park’s insightful analyses, especially for his pointing to the passage from the VyY that I study in chapter 4.
20. Jaini: “It does not contradict the fact of two (one elder and the other younger) Vasubandhus. The Vṛddhācārya Vasubandhu certainly existed, as is clear from the statements of Yaśomitra. He may well have been the author of a commentary to the Abhidharma-sāra of Dharma-śrī and also author of many Mahāyāna works” (1958:53).
21. Buescher (2013) is an exception here; he calls one Vasubandhu the “Kośakāra” for his having written the AKBh, and the other the “Bhāṣyakāra” for his commentaries on Yogācāra treatises. The latter is a confusing name, since the AKBh is, after all, a Bhāṣya. But it at least provides us a key to the identity of the “second” Vasubandhu as the author of Yogācāra commentaries. Unfortunately, although Buescher does distinguish these sets of texts stylistically and doctrinally, and although he is right to acknowledge the importance of the Yaśomitra passages (see below), his analysis fails to connect the “Bhāṣyakāra” Vasubandhu with the supposed reference of the Yaśomitra passages. The main wedge separating the Bhāṣyakāra from the Kośakāra in Buescher’s argument relies upon Von Rospatt’s point (1995:187, n408, cited in Buescher 2013:392) that the defense of momentariness in the MSABh neglects to take advantage of the AKBh’s argument that destruction is not an effect (see chapter 4). Thus, if Vasubandhu wrote the MSABh after the AKBh, he must have decided not to include his own somewhat radical, if brilliant, innovation in this commentary on a Maitreya text. How do we weigh this evidence? If, as Kritzer believes, Vasubandhu was already a Yogācāra thinker when he composed the AKBh, there is nothing to prevent his having written the MSABh first. To me, it seems quite plausible that he might have chosen not to repeat an argument in a new context for a different audience. It seems, rather, an unduly centrifugal presumption to deduce distinct authorship from this. But if these commentaries are not “his,” there is very little substance to the second “Vasubandhu.”
22. Frauwallner 1951:21. Śāstrī (2008) provides an edition of Yaśomitra’s commentary, the Sphutārthā.
23. Anacker 1998:24–25, n13. I find myself in general agreement with Anacker’s commonsense approach in this footnote. Mejor (1989–1990), in his careful analysis of the Yaśomitra passages, points out that Yaśomitra seems to have used sthavira and ācārya interchangeably, and hypothesizes another intriguing explanation of the meaning of vṛācāryavasubandhu: perhaps, given that the second passage refers to a view that agrees with the position advocated in the commentary (and is seen as non-Vaibhāṣika), the point is just that this is what Vasubandhu used to think, meaning what he thought before (ostensibly) converting to Mahāyāna. Instead of “the elder Vasubandhu,” perhaps we should speak of “the early Vasubandhu.”
24. Skilling (2000:293) “cannot dismiss” these references to a previous Vasubandhu, though he admits that we have no known texts attributed to that Vasubandhu. Buescher (2013) dismisses Anacker’s readings with little discussion, but his analysis, though truly state-of-the-art, provides no smoking gun.
The main question keeping the “two Vasubandhus” hypothesis afloat today is whether the commentator on various Yogācāra texts—the author of MSBh, MAVBh, MSABh, and DhDhVV—might be a different, though roughly contemporary, “Vasubandhu,” known to be such by Yaśomitra. (The authorship of the TSN is also widely doubted, but there is little support among the doubters for its attribution to the “second” Vasubandhu, either.) The resolution of this question awaits a detailed, close reading of each relevant passage from the AKBh together with Yaśomitra’s and Sthiramati’s commentaries, elaborated in the context of the larger arguments made by the author of the AKBh, and placed side by side with a full assessment of the commentarial accomplishments of the above four works. Buescher does some of this work, pointing out that these commentaries represent a different style from Vasubandhu’s other attributed works, and emphasizing the significance of a conspicuously missing argument (namely, Vasubandhu’s characteristic defense of momentariness, on which see my chapter 4). Yet these texts are commentaries on other people’s works (Maitreya’s or Asaṅga’s), and so might be expected to follow different genre norms than Vasubandhu’s autocommentaries.
My feeling, as I detail in the paragraphs that follow, is that the significance of the Yaśomitra passages in question is tremendously overblown. As I’ve said, we have different texts that we need to try to collate (see Buescher for an entrée into the evidence so far), and we do not presume to say that all of them are definitively composed by one author. But so far we have no biographical information and no convincing story to tell about any “second” or “third” Vasubandhu.
25. AKBh III.15 (127.18). Śāstrī 2008:339.29: pūrvācāryāḥ | yogācārā āryāsaṅgaprabhṛtayaḥ.
26. Frauwallner 1951:21, 22.
28. See Park’s (2007) hypothesis of a unified tradition that was the source of both the Yogācāra and Dārṣtāntika/Sautrāntika.
29. AKBh IV.4 (197.4ff) and AKBh VI.10 (338.5ff).
31. This point goes double if we allow for Mejor’s reading (1989–1990) of “the old Vasubandhu” as referring to Vasubandhu (some Vasubandhu) before his ostensible conversion (see note 23 above) to Mahāyāna.
32. Here I am not attempting to leverage the weight of post-Enlightenment philosophical hermeneutics to accomplish much beyond affirming this platitude. The notion of the “hermeneutic circle” harkens back to the medieval interpretive theory, based upon the necessary coherence of scripture, which required one to move back and forth between part and whole. In Gadamer’s Truth and Method, this process of return and renewal of understanding is transformed into a philosophy of meaning that advocates a never-ending movement toward a “fusion of horizons” of one’s own interpretive presuppositions with those of the text (Mueller-Vollmer 1989:37). Both uses affirm the interpreter’s ongoing responsibility to make sense of each new interpretive move in the light of all that has come before. I take from Gadamer that this means that we, as historically conditioned interpreters, should never imagine ourselves to be outside the process of interpretation, never objectively distinct from the “content” of the texts we attempt to disclose. Yet each textual moment is also an expression of a human moment, a historical event of intentional composition. My goal as an interpreter, approached at best asymptotically, is to disclose the intentions embedded within the text. This includes a reading of context and content, but seeks to elucidate, further, the human yearnings enacted through the specific social means that have produced enduring text traditions. It is an imperfect process, tainted by subjectivity and the limits of the present interpreter. But to set one’s sights lower than this would be to abandon the search for understanding. What it entails, at a minimum, is a commitment to a continual renewal of the interpretative impulse, and the continual shifting of perspectives, to explore ever further the possible motivations of each textual choice.
33. Far preferable is Deleanu’s (2006) modest admission that it is impossible to account for all of the evidence. Ironically, this admission makes the “two Vasubandhus” hypothesis far less satisfying.
34. Frauwallner 1951:42, 49.
35. Frauwallner 1995:128.
36. Anacker (1998:25, n13) asks the same question. At AKBh I.26a (16.12) Vasubandhu’s commentary mentions an Abhidharma text with six thousand verses called Dharmaskandha. This, too, may be the text misidentified as authored by Vasubandhu, but neither the AKBh nor Yaśomitra’s commentary identify (any) Vasubandhu as its author.
37. A similar point can be made about the narrative of the author of the AKK/AKBh, since the central narrative is his intellectual development from curiosity about Vaibhāṣika to his Vaibhāṣika teachers’ dismay at his having composed a commentary that counters their doctrine. If Vasubandhu did not compose both texts, the entire narrative is a fabrication. That may be, but if it is, it makes no sense to say that the “earlier Vasubandhu” composed one, and the “later Vasubandhu” composed the other. If the biography is a fabrication, we have no author information whatsoever about the root text, and only the name “Vasubandhu” as author information for the commentary.
39. Frauwallner 1951:27–28.
40. Frauwallner 1951:29–30.
42. I think it is equally possible that Paramārtha named the wrong Vikramāditya, or some other similarly misleading error is present here, but Frauwallner does not entertain these ideas.
43. In fact, it is well known to textual critics that alterations are far more likely when the topic is known and attended to carefully by a copyist, because such readers tend to call upon their own understanding to smooth away difficulties.
46. This rhetorical convergence may account for the tendency, bemoaned by Buescher (2013), for scholars treating the question even after Jaini’s critique to ignore Frauwallner’s complex defense, in which he avers two separate Vasubandhus, each of whom was both Hīnayāna and Mahāyāna.
47. This reading suggests a possible interpretation of the name whereby to be a Vasubandhu is just to be a bridge-building boundary-crosser. Paramārtha’s “Life of Vasubandhu” opens with a famous story, explaining the name of Puruṣapura, where the Vasubandhus’ father lived. The name hails back to a battle between the great god Viṣṇu and a demon, Indradamana, who had earned a boon that allowed him to be instantaneously healed of any wound, so Viṣṇu was unable to kill him with any weapon. The demon’s sister Prabhāvatī gave away the secret, by dividing a lotus and walking between the parts; Viṣṇu is able to kill Indradamana by slicing the demon in two, and walking between the divided parts. This story, with the evident lesson of “divided we fall,” then sets the stage for the naming of the three Vasubandhus with the same name. Is it over-reading to see Vasubandhu’s life story, thus framed, as a Buddhist inversion of this tale, saying that if the goal is to prevent a Tīrthaka victory and the resultant destruction of the dharma, Buddhist scholastics must stand united?
2. AGAINST THE TIMES: VASUBANDHU’S CRITIQUE OF HIS MAIN ABHIDHARMA RIVALS
1. The main arguments for distinguishing the author of the verses from the commentary involve the difference in style and viewpoint. While of course verses and commentary reflect different styles, there are countless examples of authors who have written verse texts and autocommentaries, and indeed the same style of verse-with-commentary is evident in the Viṃś. A notable distinction may be made, however, between the Vaibhāṣika content of the verses and the predominantly Sautrāntika perspective of the commentary. I have already suggested a few reasons for not jumping to the conclusion of distinct authors based on this kind of doctrinal difference. In chapter 3, note 19, I give a further reason for suspecting that the author of the AKK verses shared the Sautrāntika perspective of the commentary in a number of significant instances, which is evident in the verse author’s occasional placement of the term kila. In Butön’s (2013:243) biography of Vasubandhu this term is highlighted as having been offensive to some of the Vaibhāṣikas even before they saw the AKBh.
3. And, of course, this understanding fits with the wider use of the term dharma in the Indian cultural world in reference to the basic rules of the universe.
4. When arguing with opponents, Vasubandhu says that the scriptures (sūtras) and not treatises (śāstras) of Abhidharma philosophers are to be taken as authoritative evidence of the true nature of reality (AKBh 146.3–4: sūtrapramāṇakā vayaṃ na śāstrapramāṇakāḥ). This seems to be what he means when he adopts the name “Sautrāntika.” Yet he also says that scriptures are sometimes meant to have a secondary meaning (abhiprāya), as opposed to Abhidharma, which is always direct and literal. (AKBh 133.15: abhiprāyikaḥ sūtre lākṣaṇiko ’bhidharmaḥ). See chapter 4 for further discussion of this issue.
5. AKBh 77.3: ata evātra sūtre saṃskṛtasyotpādo ’pi prajñāyata ity uktam. This topic is discussed in chapter 3.
7. “Abhidharma is flawless discrimination [prajñā] together with its retinue” (AKBh I.2 [2.3]: prajñā ’malā sānucarā ’bhidharmaḥ). This includes the entirety of the living being associated with that discrimination. The term prajñā is glossed as dharmapravicaya, which is explained in the commentary to II.24 as referring to the classification of dharmas. See Hall 1983:44–45, n3.
8. AKBh 460.1: prāyeṇa hi kāśmīravaibhāṣikāṇāṃ nītyādisiddha eṣo ’smābhir abhidharma ākhyātaḥ. This comment comes at the conclusion to the eighth chapter of a nine-chapter commentary, and it is thought by some that this indicates that the ninth chapter was composed separately. It may also be taken to signal the pivot, in the last chapter, from a focus on the Vaibhāṣika position articulated in the verses, to a focus on the mistaken views of other groups—specifically, the Vātsīputrīyas and Vaiśeṣikas.
9. On the bhāṣya style, see Tubb and Boose 2007:173ff.
10. Dreyfus (2003) provides a fascinating and insightful analysis of the functions of debate in Tibetan Buddhist intellectual traditions. Nance (2011) and Griffiths (1999) thoughtfully examine the purposes of commentaries for Buddhist, and more generally religious, text traditions.
11. The argument over the three times, the topic of this chapter, is a paradigm case.
13. A special issue of the Journal of the International Association of Buddhist Studies (2003:2) was dedicated to “The Sautrāntikas.” See Kritzer 2003a and 2003b and Honjō 2003 for studies of Vasubandhu’s views. My view, as will become clear in later chapters, coheres with Honjō’s interpretation of the term as indicating a position that “does not recognize the authority of Sarvāstivāda Abhidharmaśāstras” (328). I am less certain that it indicates an adherence to “the Sarvāstivāda sect.” In chapter 4, I explain further how Vasubandhu replaces views previously based upon authoritative passages from śāstras and sūtras with new interpretations based on reasoning that takes a more holistic approach to scripture.
14. Vasubandhu as a Sautrāntika is considered by many to have been a follower of the view of one Śrīlāta, none of whose works are extant; Yao 2005:97, 118, n2.
15. It is also possible to read this compound to mean that its advocates hold that “things always exist” or that “all things exist”—including things that are past and future. These alternate interpretations, for Vasubandhu, amount to the same view.
16. When I say the verse text is not unreadable without the commentary, I only mean that it is possible to translate the meaning of the verses into English without including text from the commentary. I do not mean that I have read the verses and discerned their meaning without glancing at the commentary; I do not assume that that would be possible. In fact, I am doubtful that such a hypothetical is sufficiently well articulated to be meaningful. If I am to read the verses without commentary, does that mean that I am to read it with, or without, any knowledge of the doctrinal positions to which it refers? If without such knowledge, then I would surely be unqualified to read even the commentary. The question is therefore just how much knowledge is allowed. Should we assume that a reader of these verses should be familiar enough with the Vibhāṣā, which it summarizes, to know all of the details of the various subschools of fourth-century Indian Vaibhāṣika and their arguments with competitor schools? If so, then in order to understand the verses we may find ourselves insisting that they provide no nontrivial new information. The point is the same as that emphasized in the opening chapter. In order to read Indian philosophy, it is necessary to enter into the hermeneutic circle. It is not possible to gain a scientifically exact, perfectly objective analysis of textual meanings that has not already assumed a foundation of references and norms.
17. Note that verses are in bold type, as they will be throughout. Commentary will be in regular type. I present these verses as they might have appeared prior to Vasubandhu’s having written his commentary, but in fact we do not have a Sanskrit original for the verses without the commentary. The Tibetan translations do contain the verses independent of the commentary, but in the extant Sanskrit manuscripts these verses are broken up, embedded within many pages of commentary. The following is an agglomeration, then, of the Sanskrit from the commentary, which runs from AKBh 295.6 to 297.1. The Sanskrit phonological transformations that take place at word boundaries, called sandhi, are not in proper operation where the words have not been written or spoken together, so the syllable count in these verses is also improper. I insert hyphens (—) where commentary has broken into the verses, so that those with the requisite skills may hypothesize the proper implicit sandhi and versification: sarvakālāstitā—uktatvāt—dvayāt—sadviśayāt—phalāt |—tadastivādāt sarvāstivādā iṣṭāḥ—caturvidhāḥ || (25) te bhāvalakśaṇāvathā’nyathā’nyathikasaṃjñitāḥ |—tṛīyaḥ śobhanaḥ—adhvānaḥ kāritreṇa vyavasthitāḥ || (26).
18. Verse agglomeration from AKBh 297.18–301.12 (see note 17 for an explanation of the inserted hyphens): kiṃ vighnaṃ—tat kathaṃ—nānyat—adhvāyogaḥ—tathā sataḥ | ajātanaṣṭā kena—gambhīrā khaludharmatā || (27).
19. This verse presents, I believe, a significant challenge to those who would claim that the author of the AKK held an opinion of the Vaibhāṣikas significantly different from that of the author of the AKBh. It supports the analysis of the term kila I present in chapter 3, note 19.
20. Clearly it is past actions and their karmic results to which one can be “bound” and from which one can be “freed.” The pronoun “it” is probably intended to pick up the topic of the surrounding chapter, which is the “afflictive tendencies” (anuśaya) that bind one to samsara.
21. Lit. “because they are bound to the quality of what is conditioned” (saṃskāralakṣaṇayogāt).
22. AKBh 295.2–6: kiṃ punar idam atītānāgatam ucyate ’sty atha na | yady asti sarvakālāstitvāt saṃskārāṇāṃ śāśvatatvaṃ prāpnoti | atha nāsti | kathaṃ tatra tena vā saṃyukto bhavati visaṃyukto vā | na saṃskārāṇāṃ śāśvatatvaṃ pratijñāyate vaibhāṣikaiḥ saṃskṛtalakṣaṇayogāt | pratijñāyate tu viśadaṃ sarvakālāstitā.
23. A question that does not loom large, which we might think would, is: Why would anyone want to believe in things existing in the past and the future? What kind of a world is that? In this passage, the reasons adduced for believing in a “block universe” do not attempt to salve the vertiginous character of this view. Perhaps, since Vasubandhu’s view of universal momentariness has no more claim to common sense than the Sarvāstivāda view, there is no advantage to his attempting to highlight the counterintuitive nature of either side.
24. AKBh 296.9–14: bhāvānyathiko bhadantadarmatrātaḥ | sa kilāha | dharmasyādhvasu pravartamānasya bhāvānyathātvaṃ bhavati na dravyānyathātvam | yathā suvarṇabhājanasya bhittvā ’nyathā kriyamāṇasya saṃsthānānyathātvaṃ bhavati na varṇānyathātvam | yathā ca kṣīraṃ dadhitvena pariṇamadrasavīryavipākān parityajati na varṇam | evaṃ dharmo ’py anāgatād adhvanaḥ pratyutpannam adhvānam āgacchann anāgatabhāvaṃ jahāti na dravyabhāvam | evaṃ pratyutpannād atītam adhvānaṃ gacchan pratyutpannabhāvaṃ jahāti na dravyabhāvam iti.
25. “Among them, the first is dispatched by reference to the argument against the Sāṃkhyas, because it advocates [their] view of change.” AKBh 297.4: eṣāṃ tu prathamaḥ pariṇāmavāditvāt sāṃkhyapakṣe nikṣeptavyaḥ.
26. AKBh 159.18–22: na tu khalu yathā sāṃkhyānāṃ pariṇāmaḥ | kathaṃ ca sāṃkhyānāṃ pariṇāmaḥ | avasthitasya dravyasya dharmāntaranivṛttau dharmāntaraprādurbhāva iti | kaścātra doṣaḥ | sa eva hi dharmī na saṃvidyate yasyāvasthitasya dharmāṇāṃ pariṇāmaḥ kalpyeta | kaś caivam āha dharmebhyo ’nyo dharmīti | tasyaiva tu dravyasyānyathībhāvamātraṃ pariṇāmaḥ | evam apy ayuktam | kim atrāyuktam | tad eva cedaṃ na cedaṃ tatheti apūrvaiṣā vāyo yuktiḥ.
27. AKBh 101.19–102.17. See Hayes 1988, Katsura 2003, and Gold 2011 on this argument. Patil (2009) treats later developments in Indian thought, as this becomes a central set of arguments between Hindus and Buddhists.
28. On momentariness in Vasubandhu, see the discussion of AKBh 193.2–8 in chapter 4, and also Von Rospatt 1995, Katsura 2003, Bronkhorst 2006, and Gold 2011.
29. “The Sautrāntika says that each and every unconditioned thing is insubstantial [adravya]. For it is not a separate existent like form, feeling, and so on. Why? Space is only the nonexistence of a touchable. For in this way, where in the dark, physical resistance is not found, one says ‘space.’ ” AKBh 92.3–5: sarvam evāsaṃskṛtam adravyam iti sautrāntikāḥ | nahi tad rūpavedanādivat bhāvāntaram asti | kim tarhi | spraṣṭābhāvamātram ākāśam | tadyathā hy andhakāre pratighātam avindanta ākāśam ity āhuḥ. See chapter 4 for an elaboration of the epistemological implications of this view.
30. It may be mentioned that the non-Buddhist Vaiśeṣika system contains a somewhat more developed view of substances and qualities that may yet elude Vasubandhu’s attack here. See my discussion in chapter 3 of the argument at AKBh III.100, in which the non-Buddhist opponent claims that changes are attributed to qualities, which are attributable to groups of substances. That argument requires a different line of attack, which is not on the table at the moment.
31. AKBh 296.15–18: lakṣaṇānyathiko bhadantaghoṣakaḥ | sa kilāha | dharmo ’dhvasu pravarttamāno ’tīto ’tītalakṣaṇayukto ’nāgatapratyutpannābhyāṃ lakṣaṇābhyām aviyuktaḥ | anāgato ’nāgatalakṣaṇayukto ’tītapratyutpannābhyām aviyuktaḥ | evaṃ pratyutpanno ’py atītānāgatābhyām aviyuktaḥ | tadyathā puruṣa ekasyāṃ striyāṃ raktaḥ śeṣāsv avirakta iti.
32. La Vallée Poussin (1923–1925.4:55) translates here “possession de la concupiscence,” and proposes the Sanskrit rāga-prāpti, with a reference to II.36. Although the Sanskrit is not precisely correct, LVP’s reference to prāpti in II.36 is (characteristically) right on target. There (AKBh 62.16), Vasubandhu uses samanvāgama, the term here, as a defining term for one kind of prāpti. See note 34 for the implication of this terminological choice.
33. AKBh 297.4–6: dvitīyasyādhvasaṃkaraḥ prāpnoti | sarvasya sarvakṣaṇayogāt | puruṣasya tu kasyāṃcit striyāṃ rāgaḥ samudācarati kasyāṃcit kevalaṃ samanvāgama iti kimatra sāmyam.
34. This argument makes oblique reference to Vasubandhu’s critique of the Vaibhāṣika belief in a separate entity called prāpti, or “possession,” where he provides a parallel critique of the reification of a distinction without a difference (see note 32). Here, Vasubandhu’s implicit point is that Ghoṣaka’s argument relies upon an equivocation in the notion of “possession” that must fail if the Vaibhāṣika/Sautrāntika is to admit the distinct, separate entity called prāpti, “possession.” For in that case there could be no question that prāpti would be considered equivalent to active engagement; the latter can produce karmic results, whereas the former cannot. If, however, prāpti is a dharma that exists in all three times, then the very difference between the three times is granted equivalent existence in all of the three times, which is a vicious regress. Vasubandhu’s implication, then, is that his own, nonrealist understanding of “possession” does better even at supporting one of the Vaibhāṣikas’ own attempts to salvage their temporal theory (not that it does save the view): even within nonrealism, we can provide a way to distinguish among the daydreamer’s mental events. This compact implication becomes evident only once the full argument here is understood.
35. AKBh 296.19–21: avasthā’nyathiko bhadantavasumitraḥ | sa kilāha | dharmo ’dhvasu pravartamāno ’vasthām avasthāṃ prāpyānyo’nyo nirdiśyate avasthāntarato na dravyāntarataḥ | yathaikā vartikā ekāṅke nikṣiptā ekam ity ucyate śatāṅke śataṃ sahasrāṅke sahasram iti.
36. Here, kāritra means a specific activity that is the characteristic action of the dharma, its doing what it does.
37. AKBh 297.10–13: tasya kila adhvānaḥ kāritreṇa vyavasthitāḥ || (26) [I. 26c–d] yadā sa dharmaḥ kāritraṃ na karoti tadā ’nāgataḥ | yadā karoti tadā pratyutpannaḥ | yadā kṛtvā niruddhas tadā ’tīta iti | parigatam etat sarvam.
38. AKBh 297.13–14: idaṃ tu vatkavyam | yady atītam api dravyato ’sty anāgatam iti | kasmāt tad atītam ity ucyate ’nāgatam iti vā.
39. “What is the activity of the present tatsabhāga eye?” AKBh 297.15: yady evaṃ pratyutpannasya tatsabhāgasya cakṣuṣaḥ kiṃ kāritram. See notes in appendix A for some discussion of this argument.
40. The term “grasping” (grahaṇa, pratigrahaṇa), which becomes central to the Vaibhāṣika view of the distinction among the times, will be central as well to Vasubandhu’s Yogācāra definition of the illusion of “duality.” For Vasubandhu as a Yogācāra commentator, what the Vaibhāṣikas think of as the defining character of “presence” is the very essence of the reifying errors of the dualistic mind. In chapter 5 we will return to this Vaibhāṣika meaning of “grasping” as a causal result, made into the quintessence of error. To foreground the error of “grasping” is, potentially, to critique the erroneous belief in the illusion of the self. It is a fundamental mistake of deluded sentient beings to imagine that events that have limitless conditions are caused by one event, or oneself.
41. The arguments against kāritra here may be profitably compared to Vasubandhu’s rejection (at II.46a) of the Vaibhāṣika position on the qualities of “arising,” “endurance,” “change,” and “destruction” attributable to all conditioned things (saṃskāra). There, as here, the Vaibhāṣika is criticized for positing a concept (there, a “being,” vṛtti) that seems to account for the issues in question but must itself be accounted for in the same way—opening up an infinite regress that can only be closed by arbitrary stipulation.
42. Akimoto 2004:110–111.
43. AKBh 297.17–298.3: idaṃ ca vaktavyam | tenaivātmanā sato dharmasya nityaṃ kāritrakaraṇe kiṃ vighnam [I.27a1] yena kadācit kāritraṃ karoti kadācin neti | pratyayānām asāmagryam iti cet | na | nityam astitvābhyupagamāt | yac ca tat kāritram atītānāgataṃ pratyutpannaṃ cocyate tat katham [I.27a2] [298] kiṃ kāritrasyāpy anyad asti kāritram | atha tan naivātītaṃ nāpy anāgataṃ na pratyutpannam asti ca | tenāsaṃskṛān nityam astīti prāptam | ato na vaktavyam yadā karitraṃ na karoti dharmas tadā ’nāgata iti.
44. AKBh 298.4–22: syād eṣa doṣo yadi dharmāt kāritram anyat syāt | tat tu khalu nānyat [I.27a3] ato na bhavaty eṣa doṣaḥ | evaṃ tarhi sa eva adhvāyogaḥ [I.27b1] yadi dharma eva kāritraṃ kasmāt sa eva dharmas tenaivātmanā vidyamānaḥ kadācid atīta ity ucyate kadācid anāgata ity adhvanāṃ vyavasthā na sidhyati | kim atra na sidhyati | yo hy ajāto dharmaḥ so ’nāgataḥ | yo jāto bhavati na ca vinaṣṭaḥ sa varttamānaḥ | yo vinaṣṭaḥ so ’tītaḥ iti | etad evātra vaktavyam | yadi yathā varttamānaṃ dravyato ’sti tathā ’tītam anāgataṃ cāsti | tasya tathā sataḥ | ajātanaṣṭā kena [I.27b2–c1] tenaiva svabhāvena sato dharmasya katham idaṃ sidhyaty ajāta iti yo vinaṣṭa iti veti | kimasya pūrvaṃ nāsīd yasyābhāvād ajāta ity ucyate | kiṃ ca paścān nāsti yasyābhāvād vinaṣṭa ity ucyate | tasmān na sidhyati sarvathā ’py atrādhvatrayam | yady abhūtvā bhavatīti neṣyate bhūtvā ca punar na bhavatīti | yad apy uktaṃ saṃskṛtalakṣaṇayogān na śāśvatatvaprasaṅga iti | tad idaṃ kevalaṃ vāṅmātram utpādavināśayor ayogāt | nityaṃ ca nāmāsti sa dharmo na ca nitya ity apūrvaiṣā vāco yuktiḥ | āha khalv api svabhāvaḥ sarvadā cāsti bhāvo nityaś ca neṣyate | na ca svabhāvād bhāvo ’nyo vyaktam īśvaraceṣṭitam.
45. Here La Vallée Poussin (1923–1925.3:52–53, n2) imports the term anyonya from the description of the view into the list of names of views, making Buddhadeva a defender of anyonyathātva—a “mutual-differentialist.” That term works to describe this relativistic view, but the wording we have in both Sanskrit manuscripts, which is also reflected in the commentaries and Tibetan translations (pace LVP), is anyathānyathika, a “difference-differentialist.” What does this mean, exactly? Perhaps that there are different ways to differentiate between the different times, depending on your perspective. But perhaps the manuscript traditions for this passage have been corrupt since before they were translated into Tibetan, and LVP’s version, based on the Chinese implication of apekṣa, “relation,” is closer to the original.
46. AKBh 297.1–3: anyathānyathiko bhadantabuddhadevaḥ | sa kilāha | dharmo ’dhvasu pravartamānaḥ pūrvāparam apekṣyānyo ’nya ucyate avasthāntarato na dravyāntarataḥ | yathaikā strī mātā vocyate duhitā veti.
47. AKBh 297.6–8: caturthasyāpy ekasminn evādhvani trayo ’dhvānaḥ prāpnuvanti | atīte ’dhvani pūrvapaścimau kṣaṇāv atītānāgatau madhyamaḥ kṣaṇaḥ pratyutpanna iti | evam anāgate ’pi.
48. It is not insignificant that this argument is, in fact, placed first, before any of the others (see “Order of Exposition in AKBh V.25–27” earlier). The passage is a discussion of the question of the reality of the three times, but it might just as well be considered a discussion of the question of the meaning of the Buddha’s having claimed that things exist in the three times. The key doctrinal importance of this discussion is thus highlighted from the start.
49. AKBh 295.9–12: uktaṃ hi bhagavatā ’tītaṃ ced bhikṣavo rūpaṃ nābhaviṣyan na śrutavān āryaśrāvako ’tīte rūpe ’napekṣo ’bhaviṣyat | yasmāt tarhy asty atītaṃ rūpaṃ tasmāc chrutavān āryaśrāvako ’tīte rūpe ’napekṣo bhavati | anāgataṃ ced rūpaṃ nābhaviṣyat na śrutavān āryaśrāvako ’nāgataṃ rūpaṃ nābhyanandiṣyat | yasmāt tarhy asty anāgataṃ rūpam iti vistaraḥ.
50. Alternatively, it is possible to read the Buddha here as saying, effectively, “As I am about to say, and I quote. …” This may seem like an absurdity for anyone to say, even a Buddha whose words deserve to be heard and remembered, yet it is not difficult to imagine that such a passage might appear in a text that brings together different versions of a story or event.
51. AKBh 299.1–2: vayam api brūmo ’sty atītānāgatam iti | atītaṃ tu yad bhūtapūrvam | anāgataṃ yat sati hetau bhaviṣyati | evaṃ ca kṛtvā ’stīty ucyate na tu punar dravyataḥ.
52. AKBh 299.2–4: kaś caivam āha | varttamānavat tad astīti | katham anyathā ’sti | atītānāgatātmanā | idaṃ punas tavopasthitam | kathaṃ tad atītam anāgataṃ cocyate yadi nityam astīti. Note the use of ātmanā, meaning “with a nature of …,” in this passage. Since it literally means “as a self,” this term is used ironically by Vasubandhu to indicate the false appearance of an essential nature. Vasubandhu uses the same term in this way in TSN 4.
53. See my discussion of III.100 in chapter 3.
54. AKBh 299.4–6: tasmāt bhūtapūrvasya ca hetor bhāvinaś ca phalasya bhūtapūrvatāṃ bhāvitāṃ ca jñāpayituṃ hetuphalāpavādadṛṣṭipratiṣedhārtham uktaṃ bhagavatā asty atītam asty anāgatam iti.
55. This is called satkāryavāda, the view of the existent effect. See Larson 1979:10. Vasubandhu is clearly not attempting to address the complexities of this view here, so I see no need to delay the argument by delving into Sāṃkhya causal theory.
56. Dan Arnold (pers. comm.) says that Vasubandhu is here adopting the idea from Jain philosophers that the existential verb acts as a nipāta. The point is that the word connects other words, rather than affirming some new entity.
57. Lit. “it has nonexistence.” The point here is that although you say this, you do not intend to say that the light has some reality as a nonexistent.
58. AKBh 299.6–8: astiśabdasya nipātatvāt | yathā ’sti dīpasya prāgabhāvo ’sti paścād abhāva iti vaktāro bhavanti yathā cāsti niruddhaḥ sa dīpo na tu mayā nirodhita iti | evam atītānāgatam apy astīty uktam | anyathā hy atītānāgatabhāva eva na sidhyet.
59. Vasubandhu noted the ambiguous nature of the word “is” fifteen hundred years before Bill Clinton’s famous “It depends on what the meaning of the word ‘is’ is.” Clinton’s point was just that by the word “is,” he meant that he was not presently engaged in inappropriate relations, and it was therefore not technically perjury to have allowed his listeners to assume incorrectly that he was denying past liaisons. Strange as it seems, this set of fine points reflects, in more than form alone, the relevant moral questions about the three times in Vasubandhu’s Buddhist perspective. First, Vasubandhu’s analysis reflects a similar assertion of the technical correctness of the Buddha’s utterance, even if it is known to have been uttered in a context in which its proper meaning would be misinterpreted. Second, Vasubandhu’s claim that past events may seem to exist—but do not—in the present tense use of “is” exactly reflects Clinton’s claim that he was not (technically) denying past events by his denial of what “is.”
60. AKBh 299.8–11: yat tarhi laguḍaśikhīyakān parivrājakān adhikṛtyoktaṃ bhagavatā yat karmābhyatītaṃ kṣīṇaṃ niruddhaṃ vigataṃ vipariṇataṃ tad astīti | kiṃ te tasya tasya karmaṇo bhūtapūrvatvaṃ necchanti sma | tatra punas tadāhitaṃ tasyāṃ saṃtatau phaladānasāmarthyaṃ saṃdhāyoktam | anyathā hi svena bhāvena vidyamānam atītaṃ na sidhyet.
61. For an intricate study of this term, see Broido 1985.
62. AKBh 299.12–16: itthaṃ caitad evaṃ yat paramārthaśūnyatāyām uktaṃ bhagavatā cakṣur utpadyamānaṃ [correcting utyadyamānaṃ] na kutaś cid āgacchati nirudhyamānaṃ na kvacit saṃnicayaṃ gacchati | iti hi bhikṣavaś cakṣur abhūtvā bhavati bhūtvā ca pratigacchatīti | yadi cānāgataṃ cakṣuḥ syān noktaṃ syād bhūtvā na bhavatīti | varttamāne ’dhvany abhūtvā bhavatīti cet na | adhvano bhāvād anarthāntaratvāt | atha svātmany abhūtvā bhavati | siddham idam anāgataṃ cakṣur nāstīti.
63. AKBh 295.13–16: dvayāt [I.25b1] dvayaṃ pratītya vijñānasyotpāda ity uktam | dvayaṃ katamat | cakṣū rūpāṇi yāvat mano dharmā iti | asati vā ’tītānāgate tadālambanaṃ vijñānaṃ dvayaṃ pratītya na syāt | evaṃ tāvad āgamato ’sty atītānāgataṃ yuktito ’pi.
64. This is a description of the connection between the sixth link in the chain, the “contact” (sparśa) of sensory organ and sensory object, and the seventh link, “sensation” (vedanā). I discuss Vasubandhu’s presentation of the twelve links in chapter 6.
65. The “experiential object condition” is one of the four kinds of conditions. See AKBh II.61–62 (97.22–100.17). Vasubandhu provides no technical definition for a “producer condition” (janaka-pratyaya), though this might be thought to correspond imprecisely to the “causal condition” (hetu-pratyaya), since at AKBh II.46 Vasubandhu argues that “producings” come about as a result of the assemblage of causal conditions.
66. AKBh 299.16–21: yad apy uktaṃ dvayaṃ pratītya vijñānasyotpādād iti idaṃ tāvad iha saṃpradhāryam | yan manaḥ pratītya dharmaś cotpadyate manovijñānaṃ kiṃ tasya yathā manojanakaḥ pratyaya evaṃ dharmā āhosvid ālambanamātraṃ dharmā iti | yadi tāvat janakaḥ pratyayo dharmāḥ kathaṃ yad anāgataṃ kalpasahasreṇa bhaviṣyati vā na vā tad idānīṃ vijñānaṃ janayiṣyati | nirvāṇaṃ ca sarvapravṛttinirodhāj janakaṃ nopapadyate | athālambanamātraṃ dharmā bhavanti | atītānāgatam apy ālambanaṃ bhavatīti brūmaḥ.
67. This could be taken as an equivocation, attempting to counter an argument that works for all senses by focusing only on the mental sense. But it could also be taken to be just an opening step in a multistage argument. First, show that for mental objects, the Buddha must mean only experiential objects, not producer objects; second, point out that if the Buddha meant the quote to apply this way to mental objects, we are entitled to read it in this way for all objects.
68. AKBh 145.5–13: kathaṃ sahotpannayor janyajanakabhāvaḥ sidhyati | kathaṃ ca na sidhyati | asāmarthyāt | jāte dharme dharmasya nāsti sāmarthyaṃ pratijñā ’viśiṣṭam | yad eva hīdaṃ sahotpannayor janyajanakabhāvo nāstīti tad evedaṃ jāte dharme dharmasya nāstīti | anyonyajanakaprasaṅgāt tarhi | iṣṭād adoṣaḥ | iṣṭameva hi sahabhūhetor anyonyaphalatvam | iṣṭam idaṃ sūtre tv aniṣṭṃ sparśavedanayor anyonyaphalatvam | cakṣuḥsaṃsparśaṃ pratītya utpadyate cakṣuḥsaṃsparśajā vedanā na tu cakṣuḥsaṃsparśajāṃ vedanāṃ pratītyotpadyate cakṣuḥsaṃsparśa iti vacanāt | janakadharmātikramāc cāyuktam | yo hi dharmo yasya dharmasya janakaḥ prasiddhaḥ sa tasmāt bhinnakālaḥ prasiddhaḥ | tadyathā pūrvaṃ bījaṃ paścād aṅkuraḥ pūrvaṃ kṣīraṃ paścād dadhi pūrvam abhighātaḥ paścāc chabdaḥ pūrvaṃ manaḥ paścāt manovijñānam ity evam ādi.
69. In the light of Kritzer’s thesis (2003a, 2003b, 2005) that Vasubandhu’s identity as a Sautrāntika masked his true identity as an advocate of Yogācāra, it is of interest to note how this argument against mutual causality may serve as an intervention in the issues that shaped the Yogācāra-vijñānavāda as a doctrinal system.
Lambert Schmithausen (2007; first published 1987) argued that the initial, crucial motivation for one of the defining Yogācāra doctrinal innovations—the “store consciousness” or “latent consciousness” (ālayavijñāna)—was the problem of continuity, as it applied specifically to the meditative state called nirodhasamāpatti. One problem, to put it briefly—see Griffiths 1991 for a careful study—was that a consciousness needs to be caused by a consciousness, but this meditative state is defined as being completely free of the six sensory consciousnesses. So the question arose, How does consciousness start again once it has stopped? Schmithausen argued that the “store consciousness” or the “hidden consciousness”—the consciousness that’s tucked away in the body—was first introduced to solve this problem. In reply to Schmithausen, Buescher (2008) has recently argued that, in fact, the passage that Schmithausen cites as the initial use of this new concept does not in fact need to be read as an example of the “latent consciousness” of full-blown Yogācāra-vijñānavāda. It is, instead, a continuation of what Buescher terms the “bi-polar bīja-model,” wherein the two “seeds” of consciousness bring one another about through mutual causality. See D’Amato 2009 for a summary of Buescher’s thesis.
If Buescher is right, then there were two potential solutions to the problem of continuity available to Buddhists of the emerging Yogācāra, one of which was the ālayavijñāna and the other of which relied upon mutual causality. In that case, Vasubandhu’s argument against mutual causality here, and his linear approach to causality in general (see chapter 3), could work to preclude a key, previous solution to the problem of continuity and might be taken to push those who took this problem seriously toward the more mature ālayavijñāna as the only truly viable solution.
70. This strict refusal to grant causal efficacy to merely apparent objects is at the crux of Vasubandhu’s Yogācāra distinction between the causal “dependent nature” (paratantrasvabhāva) and the merely experienced “fabricated nature” (parikalpitasvabhāva). His extensive defense of the Mahāyāna view that all things are “merely apparent” (vijñapti-mātra) also picks up on this understanding of what is “merely an experiential object” (ālambana-mātra). See chapter 5.
71. AKBh 295.17–19: sadviṣayāt [I.25b2] sati viṣaye vijñānaṃ pravartate nāsati | yadi cātītānāgataṃ na syād asadālambanaṃ vijñānaṃ syāt | tato vijñānam eva na syād ālambanābhāvāt.
72. This argument is replicated in the TSN and the Viṃś, which trace the Yogācāra path of mental cultivation whereupon once one has eliminated the reification of the mental object, the consciousness also goes away. Of course, the Vaibhāṣika point in this passage is a reductio—we know there is consciousness, so its objects must exist—but the logic in this passage that deduces the impossibility of consciousness without its object fits Vasubandhu’s later view that even the subjective side of consciousness is an illusion. See chapter 5.
74. I note this here because it distinguishes this argument from the other three arguments, which Vasubandhu addresses directly by saying: “As to what is meant by ‘Because it is said … ’ ”; “As to the expression ‘The arising of consciousness is dependent upon two things … ’ ”; and “As to the expression ‘Because of the result.’ ”
75. AKBh 299.21–25: yadi nāsti katham ālambanam | atredānīṃ brūmaḥ | yadā tad ālambanaṃ tathāsti kathaṃ tad ālambanam abhūt bhaviṣyati ceti | na hi kaścid atītaṃ rūpaṃ vedanāṃ vā smarann astīti paśyati | kiṃ tarhi | abhūd iti | yathā khalv api varttamānaṃ rūpam anubhūtaṃ tathā tad atītaṃ smaryate | yathā cānāgataṃ vartamānaṃ bhaviṣyati tathā buddhyā gṛhyate | yadi ca tattathaivāsti vartamānaṃ prāpnoti | atha nāsti | asad apy ālambanaṃ bhavatīti siddham.
76. It will not benefit the opponent to claim that the unicorn has its existence as an experience. Vasubandhu does not dispute the existence of the experience, only the object of the experience.
77. See II.62, where Vasubandhu discusses the ālambana-pratyaya, the mental-object condition, and also my section on I.41–42 in chapter 3.
78. In this way, the apparent “perception” of temporality is similar to the apparent “perception” of destruction, as argued in Vasubandhu’s proof of momentariness. See chapter 4.
79. AKBh 295.20–296.1: phalāt [I.25c] | yadi cātītaṃ na syāt śubhāśubhasya karmaṇaḥ phalam āyatyāṃ kathaṃ syāt | na hi phalotpattikāle [296] varttamāno vipākahetur astīti | tasmād asty evātītānāgatam iti vaibhāṣikāḥ.
82. AKBh 300.19–21: yad apy uktaṃ phalād iti | naiva hi sautrāntikā atītāt karmaṇaḥ phalotpattiṃ varṇayanti | kiṃ tarhi | tatpūrvakāt saṃtānaviśeṣād ity ātmavādapratiṣedhe saṃpravedayiṣyāmaḥ.
83. This is Kapstein’s elegant translation (2001a:374) of AKBh 477.9–20. I see no reason to change it.
84. AKBh 477.11–12, 477.15.
3. MERELY CAUSE AND EFFECT: THE IMAGINED SELF AND THE LITERALISTIC MIND
1. See Kapstein 2001a for a translation and Gold 2011 for a summary. Duerlinger 2003 is a book-length translation and study.
2. AKBh III.17 (129.5–21).
3. As Vasubandhu says when he cites this same passage in book IX, it is from the Paramārthaśūnyatā.
4. AKBh 129.5–11: atredānīṃ bāhyakā ātmavādaṃ parigṛhyottiṣṭante | yadi sattvo lokāntaraṃ saṃcaratīti pratijñāyate siddha ātmā bhavatīti | sa eṣa pratiṣidhyate nātmāsti [III.18a1] kīdṛśa ātmā ya imān nikṣipaty anyāṃś ca skandhān pratisaṃdadhātīti parikalpyate | sa tādṛśo nāsty antarvyāpārapuruṣaḥ | evaṃ tūktaṃ bhagavatā asti karmāsti vipākaḥ kārakas tu nopalabhyate ya imāṃś ca skandhān nikṣipati anyāṃś ca skandhān pratisaṃdadhāty anyatra dharmasaṃketāt | tatrāyaṃ dharmasaṃketo yadutāsmin satīdaṃ bhavatīti vistareṇa pratītyasamutpādaḥ.
5. Here Śāstri (2008) has an opposed reading: etad saty apy ātmani instead of etad asaty apy ātmani.
6. AKBh 129.12–21: kīdṛśas tarhy ātmā na pratiṣidhyate | skandhamātraṃ tu [III.18a2] yadi tu skandhamātram evātmeti upacaryate tasyāpratiṣedhaḥ | evaṃ tarhi skandhā eva lokāntaraṃ saṃcarantīti prāptaṃ skandhamātraṃ tu nātra saṃcaratīti | kleśakarmābhisaṃskṛtam | antarābhavasaṃtatyā kukṣimeti pradīpavat || (18) [III.18b–d] kṣaṇikā hi skandhās teṣāṃ saṃcarituṃ nāsti śaktiḥ | kleśais tu paribhāvitaṃ karmabhiś ca kleśamātram antarābhavasaṃjñikayā saṃtatyā mātuḥ kukṣimāyāti | tadyathā pradīpaḥ kṣaṇiko ’pi saṃtatyā deśāntaram iti nāsty eṣa doṣaḥ | tasmāt siddham etad asaty apy ātmani kleśakarmābhisaṃskṛtaḥ skandhānāṃ saṃtāno mātuḥ kukṣim āpadyata iti.
7. Dhammajoti 2009:246–252.
8. Vasubandhu stages an argument against the Vaibhāṣikas about the relationship between avidyā and prajñā at III.29 (141–142). The opponent (probably Vasubandhu’s own view) proposes that ignorance just is “wrong view,” but the Vaibhāṣikas reject this position and hold that ignorance is its own, distinct dharma. Either way, ignorance and “wrong views” are both causally efficacious.
9. One of the characteristic doctrines of Buddhist epistemology, known from Dignāga on (but already evident in the AKBh passage to be studied here), is that it is a mistake to distinguish between the “agent” and the “action” of knowing. Whereas other traditions distinguish between the knower (pramātṛ) and the action of knowing (pramiti), the Buddhist epistemologists see the action itself as both, as Dunne explains (2004:49–50). Two recent, somewhat divergent, interpretations of Buddhist thought as responding to the issues in Donald Davidson’s famous article “Actions, Reasons, and Causes” (1963) are Arnold 2012 and Coseru 2012. I will reflect upon this issue briefly—from what I take to be Vasubandhu’s perspective—in chapters 4 and 5, but for now it should be stated that Vasubandhu does not deny the conceptual utility of the distinction between reasons and causes. The important point is that the strict delineation between a realm of causes and a realm of reasons, as Davidson and others (such as Sellars 1997) have required, is a manifestation of the delusion of sentient beings. The realms interconnect in ways we do not ordinarily see. In a Buddhist view, we take reasons to be the basis for our beliefs and actions, but we are often caused, karmically and through ignorance, to only apparently “reason” our way to our mistaken perspectives.
10. See Arnold 2012:34–35 for a brief discussion of this issue as it is taken up by Dharmakīrti.
11. AKBh V.7 (281.19–282.2): ātmadṛṣṭir ātmīyadṛṣṭir vā satkāyadṛṣṭiḥ | sīdatīti sat | cayaḥ kāyaḥ saṅghātaḥ skandha ity arthaḥ | saccāyaṃ kāyaś ceti satkāyaḥ pañcopādānaskandhāḥ | nityasaṃjñāṃ piṇḍasaṃjñāṃ ca tyājayitum evaṃ dyotitā | etatpūrvako hi teṣv ātmagrahaḥ | satkāye dṛṣṭiḥ satkāyadṛṣṭiḥ | sarvaiva sāsravālambanādṛṣṭiḥ satkāye | ātmātmīyadṛṣṭir eva tu satkāyadṛṣṭir uktā | yathā gamyeta [282] satkāyadṛṣṭir iyaṃ nātmani nātmīye veti | yathoktaṃ ye kecid bhikṣavaḥ śramaṇā vā brāhmaṇā vā ātmeti samanupaśyantaḥ samanupaśyanti sarve ta imān eva pañcopādānaskandhān iti.
12. AKBh 282.2–3: tasyaivātmābhimatasya vastuno dhruvadṛṣṭir ucchedadṛṣṭir vā ’ntagrāhaḍṛṣtiḥ | śāśvatocchedāntagrahaṇāt.
13. I write of “substratum” here as a shorthand for the entities upon which the false appearance is mistakenly superimposed, but I hope the reader will keep in mind that there is never any intent, within Vasubandhu’s writings, to suggest that the “substratum” is a solid, essential thing. That would negate the whole point of denying the apparent unity which is imposed upon the substratum. The paradigmatic “substratum” is the five ever-changing aggregates.
14. Clearly this refers to the five consciousnesses that accompany the five sensory organs and their objects—that is, all of the consciousnesses except the mental consciousness.
15. AKBh 29.20–30.3: atha kasmāl laukikī samyagdṛṣṭir manovijñānasaṃprayuktaivocyate | yasmāt pañcavijñānasahajā dhīr na dṛṣṭir atīraṇāt || (41) [I.41c–d] santīrikā hi dṛṣṭir upadhyānapravṛttatvāt | na caivaṃ pañcavijñānasahajā prajñā | tasmād asau na (30) dṛṣṭḥ | ata eva cānyā ’pi kliṣṭā ’kliṣṭā vā prajñā na dṛṣṭiḥ | cakṣur idānīm asantīrakatve kathaṃ dṛṣṭiḥ | rūpālocanārthena | yasmāt cakṣuḥ paśyati rūpāṇi [I.42a].
16. See chapter 4, where I discuss Vasubandhu’s patterns of scriptural interpretation.
17. Taken as a dharma that is a particular modality of prajñā, dṛṣṭi should have only one essential nature (svabhāva), so to admit that this error is only “sloppy” would be already to move toward a figurative reading of the term and to reject it as a dharma.
18. Vasubandhu even quotes the Vaibhāṣika’s source text, the Vibhāṣā, to support one of his points, although of course he does not rely upon this text as chief witness.
19. Although the AKK summarizes the position of the Vaibhāṣikas of Kashmir, the word kila, “so they say,” appears eight times in the verses, and six of them provide telling hints that the author of the verses was already harboring the exact questioning, critical view that dominates in the AKBh. Butön’s (2013:243) biography of Vasubandhu even contains an episode in which AKK is said by the Vaibhāṣika followers of Saṃghabhadra to “insult our philosophical system” by virtue of its repeated inclusion of this term.
Overall, the eight instances of kila are, I believe, a significant indication that the author of the verses was of the same view as the author of the AKBh, namely, a skeptical reader of the system who nonetheless was highly motivated to expend a tremendous effort to examine its every detail. (1) At I.3 (2.22), the verse implies with its kila that Abhidharma is not an authority, but “they say” it is. This in itself tells us that the author of the verses considered that the ultimate authority must be the sūtras, not the Vibhāṣā—which is to say, he was a Sautrāntika. (2) At I.28 (18.14), the point of the kila is that “they say” the void of the space element is light and darkness, day or night. Vasubandhu’s commentary denies the reality of the dharma called “space.” (3) As we are currently discussing, at I.42 (30.10), “they say” that consciousness cannot be said to “see,” because vision is blocked by intervening matter, whereas consciousness is not—a position against which the commentary argues forcefully. (4) At II.1 (38.11), the verse which has a “so they say” in it is about the “predominating influence” of the sensory organs (indriya)—which is just senseless from the Sautrāntika perspective, since this way of speaking affirms the reality of something that is not at all causally real. Consciousness really causes things to happen (as is claimed at II.2). (5) At III.25 (133.10), the verse says that “they” advocate the static, not the other (less literal-minded) interpretations of dependent arising (pratītyasamutpāda). I will discuss some of the diverse interpretations of dependent arising allowed in the AKBh in chapter 6. (6) At IV.27 (213.1), “they say” that you only have “undiscipline” for a day and a night; but for the Sautrāntika, this is wrong: you do not take anti-vows. A lack of discipline is just something that happens, or that you do. As a “lack,” this issue parallels the arguments against nonexistent entities that I will discuss in chapter 4. So, the commentary says, undiscipline is not a special entity, it is just a form of volition. But of course, you could decide to be undisciplined for as long as you like. (7) At IV.31 (215.16), “they declare” a position on the meaning of partial Upāsaka vows which the commentary criticizes based on scriptural analysis. (8) At V.37 (307.6), “they” have specific meanings for the word āsrava. Perhaps this is mistaken because it is too literal-minded, but I am not sure what the problem is.
21. AKBh 30.4–6: yadi cakṣuḥ paṣyed anyavijñānasamaṅgino ’pi paśyet | na vai sarvaṃ cakṣuḥ paśyati | kiṃ tarhi sabhāgaṃ [I.42b1] savijñānakaṃ yadā bhavati tadā paśyaty anyadā neti | evaṃ tarhi tad eva cakṣurāśritaṃ vijñānaṃ paśyatīty astu.
22. The word sabhāga, which I translate as “one with a corresponding [consciousness]” is a technical term in the Vaibhāṣika system for those dharmas that are active (i.e., present tense) objects of consciousness. AKBh I.39 (27.19–20).
23. This argument establishes that the notion of a “view” is caused by the consciousness, in accord with the logic of causality and knowledge articulated at the start of AKBh IX. As I will explain further in chapter 4, Vasubandhu believes that in order for something to be called real it must be a necessary condition for some causal result.
24. AKBh 30.7–12: na tadāśritam | vijñānaṃ [I.42b2–c1] paśyatīti śakyam avijñātum | kiṃ kāraṇam | dṛśyate rūpaṃ na kilāntaritaṃ yataḥ || (42) [I.42c2–d] yasmāt kila rūpaṃ kuḍyādivyavahitaṃ na dṛśyate | yadi hi vijñānaṃ paśyet tasyāpratighatvāt kuḍādiṣu pratighāto nāsti ity āvṛtam api rūpaṃ paśyet.
25. It seems to me that there is no need to appeal to a Vijñānavāda explanation here, as Yaśomitra suggests (Śāstrī 2008:91–92), since the point is that the sensory organ provides the support (āśraya) for the consciousness—which is how the senses are said to operate at I.9c–d. But the argument is clearly tending toward a Vijñānavāda perspective.
26. AKBh 30.12–18: naiva hy āvṛte cakṣurvijñānam utpadyata ity anutpannaṃ kathaṃ drakṣyati | kiṃ khalu notpadyate | yasya tu cakṣuḥ paśyati tasya cakṣuṣaḥ sapratighatvād vyavahite vṛtty abhāva iti vijñānasyāpy anutpattir āśrayeṇaikaviṣayapravṛttatvāt yujyate | kiṃ nu vai cakṣuḥ prāptaviṣayaṃ kāyendriyavat yata āvṛtaṃ na paśyet | sapratighatvāt | kācābhrapaṭṭāmbubhiś cāntaritaṃ kathaṃ dṛśyate | tasmān na sapratighatvāc cakṣuṣa āvṛtasya rūpasyādarśanam | kiṃ tarhi | yatrālokasyāpratibandha āvṛte ’pi rūpe tatropapadyata eva cakṣurvijñānam | yatra tu pratibandhas tatra notpadyata ity anutpannatvād āvṛtaṃ nekṣate.
27. AKBh I.43d (32–33) clarifies that it is when in contact (spṛṣṭa)—that is, with no interval—that the organs perceive their objects.
28. Here we see Vasubandhu using the hermeneutic strategy of requiring consistency of purpose in texts, which McCrae (2000:433ff) writes about in the case of the Mīmāṃsā.
29. AKBh 30.18–24: yat tarhi sūtra uktaṃ cakṣuṣā rūpāṇi dṛṣṭveti | tenāśrayeṇety ayam atrābhisandhir yathā manasā dharmān vijñāyety āha | na ca mano dharmān vijānāti | atītatvāt | kiṃ tarhi | manovijñānam | āśritakarma vā āśrayasyopacaryate | yathā mañcāḥ krośantīti | yathā ca sūtra uktaṃ cakṣurjñeyāni rūpāṇīṣṭāni kāntānīti | na ca tāni cakṣuṣā vijñāyante | uktaṃ ca sūtre cakṣur brāhmaṇa dvāraṃ yāvad eva rūpāṇāṃ darśanāya ity ato gamyate cakṣuṣā dvāreṇa vijñānaṃ paśyatīti | darśane tatra dvārākhyā | na hy etad yujyate | cakṣur darśanaṃ rūpāṇāṃ darśanāyeti.
30. AKBh I.17 (11.18–12.3).
31. An alternate reading adopted by Śāstrī (2008:92.9) and Pāsādika (1989:26) is “to be cognized by the eye,” cakṣurvijñeyāni, which makes the figurative nature of the statement even more evident. The eye might conceivably have its own kind of knowledge, but not its own “cognition.”
32. See Gold 2007, Tzohar 2011, and chapter 4.
33. AKBh 31.1–3: yadi vijñānaṃ paśyati ko vijānāti kaś cānayor viśeṣaḥ | yad eva hi rūpasya vijñānaṃ tad evāsya darśanam iti | tadyathā kācit prajñā paśyatīty apy ucyate prajānātīty apy evaṃ kiṃcid vijñānaṃ paśyatīty apy ucyate vijānātīty api.
34. Vasubandhu will also pick up this argument in Viṃś 16. See chapter 5.
35. Vasubandhu rejects the question.
36. AKBh 31.3–10: anye punar āhuḥ | yadi cakṣuḥ paśyati kartṛbhūtasya cakṣuṣaḥ kā ’nyā dṛśikriyeti vaktavyam | tad etad acodyam | yadi hi vijñānaṃ vijānātītīṣyate | na ca tatra kartṛkriyābhedaḥ | evam atrāpi | apare punar bruvate | cakṣurvijñānaṃ darśanaṃ tasyāśrayabhāvāc cakṣuḥ paśyatīty ucyate | yathā nādasyāśrayabhāvāt ghaṇṭā nadatīty ucyata iti | nanu caivaṃ vijñānasyāśrayabhāvāc cakṣur vijānātīti prāpnoti | na prāpnoti | tad vijñānaṃ darśanam iti rūḍhaṃ loke | tathā hi tasminn utpanne rūpaṃ dṛṣṭam ity ucyate na vijñātam | vibhāṣāyām apy uktaṃ cakṣuḥ saṃprāptaṃ cakṣurvijñānānubhūtaṃ dṛṣṭam ity ucyata iti | tasmāc cakṣuḥ paśyatīty evocyate na vijānātīti.
37. What Vasubandhu says is that the distinction between an eye that sees and an eye that does not see is simply whether the eye brings about an event of visual consciousness. There is no activity for the eye that is called “seeing” distinct from the eye simply being an eye, or doing what eyes do by their very nature. But if there is an eye consciousness caused by that eye, we say, figuratively, that that eye “sees,” because it is the support for that conscious awareness. Arnold (2012:168–169) characterizes Dignāga as arguing, in a similar passage, that we can only take an awareness to be “pramāṇa” because of its “being contentful (viṣayākāratā).” This seems to differ from what Vasubandhu would say. For Vasubandhu, the point here is that there is never really any content (the content as it appears is an impossibility, since it appears present when it is past; see next note)—but we speak figuratively as though there is, since this is how it appears to mistaken awareness. But perhaps this is not a real disagreement, since Dignāga also says in this context that awareness is pramāṇa only figuratively.
38. This argument presages Vasubandhu’s interpretation of mental “nonduality” in his Yogācāra works. There he explains that mind and mental objects appear to be distinct entities, whereas in fact mental objects are mere appearances within mental events, without which the mental events would not exist. This is why mental events are called “nondual.” Here we see that even in the AKBh, Vasubandhu foregrounded the fact that the object of cognition does not even exist at the same time as the mental event of which it is an object.
39. This critique of the distinction between action and agent resembles, though not perfectly, the “duality” we see criticized in Yogācāra texts such as the opening of the DhDhVV—as we will discuss in chapter 5. In the DhDhVV, the linguistic divisions among entities are said, similarly, to be false. Yet the specifics denied are not action and agent, but action and object. This extension reflects the denial of the separate existence of external objects for Yogācāra-Vijñānavāda.
40. The context here makes it clear that the mention of conventional terms used “by choice” or “as one pleases” (chandataḥ) refers to the Buddha’s own figurative use of conventional terms.
41. AKBh 31.10–16: vijñānaṃ tu sānnidhyamātreṇa rūpaṃ vijānātīty ucyate | yathā sūryo divasakara iti | atra sautrāntikā āhuḥ | kim idam ākāśaṃ khādyate | cakṣur hi pratītya rūpāṇi cotpadyate cakṣurvijñānam | tatra kaḥ paśyati ko vā dṛśyate | nirvyāpāraṃ hīdaṃ dharmamātraṃ hetuphalamātraṃ ca | tatra vyavahārārthaṃ cchandata upacārāḥ kriyante | cakṣuḥ paśyati vijñānaṃ vijānātīti nātrābhiniveṣṭavyam | uktaṃ hi bhagavatā janapadaniruktiṃ nābhiniviśeta saṃjñāṃ ca lokasya nābhidhāved iti | eṣa tu kāśmīravaibhāṣikāṇāṃ siddhāntaḥ | cakṣuḥ paśyati śrotraṃ śṛṇoti ghrāṇaṃ jighrati jihvā āsvādayati kāyaḥ spṛśati mano vijānātīti.
42. Here we see the Vaibhāṣika view that consciousness must play the role of a kind of subjective (Cartesian?) “seer” in cognition set up as a counterargument, refuted by the strict view that all things are “only cause and effect,” not agent-and-object. See Williams (1998b:233–235) for an argument from the late Indian Buddhist philosopher Śāntarakṣita that resembles this one, used to defend the controversial doctrine of self-cognition (svasaṃvitti). At VII.18, Vasubandhu denies one of the foundational arguments for self-cognition—that for a Buddha’s omniscience to work, a single cognition must be able to cognize all cognizables, including itself. Yao (2005:44–49) provides a study of this and related arguments and sees Vasubandhu as essentially accepting the Sarvāstivāda view that the Buddha’s omniscience is not taken to see all things simultaneously. We may say that Vasubandhu goes further, though, by denying the implicit structure of consciousness that would distinguish, ontologically, between cognition and what is cognized.
43. “Where is this authority established? In the treatise. We have scriptures as means of knowledge; we do not have treatises as means of knowledge.” AKBh 146.3–4: kva caiṣa niyamaḥ siddhaḥ | śāstre | sūtrapramāṇakā vayaṃ na śāstrapramāṇakā.
44. In his comment on the opening of Vasubandhu’s Triṃś, Sthiramati explains: “[A word] is used figuratively with regard to something which is not there, as when [one calls] a Bahīkan [person] an ox.” Lévi 1925:16: yac ca yatra nāsti tat tatropacaryate | tadyathā bahīke gauḥ. See Gold 2007:142ff for a brief discussion of this passage. For an in-depth study of upacāra that places Vasubandhu’s usage in its proper context, see Tzohar 2011.
45. AKBh II.46 (76.20–21): tad etad ākāṣaṃ pāṭyata iti sautrāntikāḥ. The discussion of the characteristics of conditioned things is AKBh II.45c–46.
47. As Roy Tzohar has pointed out, the implications of the interactions among these various elements were only cashed out explicitly by Vasubandhu’s great commentator, Sthiramati. But I hope to have shown that a careful reading even of this early passage displays Vasubandhu’s intertwined interests in hermeneutics and philosophy of mind. Tzohar 2011:200: “While all these elements—upacaras, the non-existence of objectified phenomena, and their underlying causal reality—are present in the AKBh (often even in a single passage, as in AKBh 1.42), only in Sthiramati’s TriṃsBh are they explicitly tied together theoretically, joining to form a cohesive account of the relation between language and reality.”
48. Meyers (2010) studies Vasubandhu’s discussion of seed-to-sprout causality as a model of “organic” as opposed to “mechanistic” causality. This is an important distinction for modern interpreters, who too readily think that a causal story is one in which results are determined, inevitably, by causes. I will discuss the issue of freedom and determinism in chapter 6. Here the issue is not the mechanistic vs. organic, but linear vs. part/whole causality.
49. For a translation of the argument as it appears in the Viṃś, see Kapstein 2001b. I provide a brief summary in chapter 5.
50. As I gather, this does not mean that all beings manage to achieve meditative absorption during these intermediate periods. Rather, only beings in absorption survive in this universe, but countless other universes are at different stages in the cycle of development and destruction, so there are ample opportunities for rebirth in one’s appropriate form.
51. AKBh 188.24–189.6: atha katīmāḥ saṃvartanyaḥ | [189] saṃvartanyaḥ punas tistro bhavanty agnyambuvāyubhiḥ | [III.100ab] ekatra dhyāne sattvāḥ samaṃ saṃvartante etasyām iti saṃvartanī | saptabhiḥ sūryais tejaḥsaṃvartanī bhavati varṣodakenāpsaṃvartanī vāyuprakopād vāyusaṃvartanī | tābhiś ca bhājanānāṃ sūkṣmo ’py avayavo nāvaśiṣyate | atra tu kecit tīrthakarā icchanti | paramāṇavo nityās te tadānīṃ śiṣyanta iti | kasmāt ta evam icchanti | mā bhūd abījakaḥ sthūlānāṃ prādurbhāva iti.
52. AKBh I.43 (33.2–3): na spṛśanti nirantare tu spṛṣṭasaṃjñeti bhadantaḥ | bhadantamataṃ caiṣṭavyam.
53. Though to be more accurate it seems that they in fact used the same metaphorical notion of the eye—meaning the mind—that Vasubandhu says the Buddha adopts, when they say that the image is constructed in the “eye” of the beholder.
54. AKBh 189.6–8: nanu ca sattvānāṃ karmajaḥ prabhāvaviśiṣṭo vāyurbījam uktam | saṃvartanīśīrṣavāyur vā tasya nimittaṃ bhaviṣyati | vāyunā lokāntarebhyo bījānyāhriyanta iti mahīśāsakāḥ sūtre paṭhanti.
55. Some Indian creation stories do, however, have the universe proceeding from a cosmic egg.
56. He does this in his disproof of a creator god as well. See Gold 2011.
57. AKBh 189.8–13: evam api na te bījādibhyo ’ṅkurādīnām utpattim icchanti | kiṃ tarhi | svebhya evāvayavebhyas teṣām api svebhyaḥ | evaṃ yāvat paramāṇubhyaḥ | kim idaṃ bījādīnām aṅkurādiṣu sāmarthyam | na kiṃcid anyatra tatparamāṇūpasarpaṇāt | kiṃ punaḥ kāraṇaṃ ta evam icchanti | na hi vijātīyāt saṃbhavo yukta iti | kasmān na yuktaḥ | aniyamo hi syāt | śaktiniyamān naivaṃ bhaviṣyati | śabdapākajotpattivat | citro hi guṇadharmo dravyaṃ tu naivam | samānajātīyebhyaḥ eva hi dravyebhyaḥ samānajātīyānāṃ dṛṣṭa utpādas tadyathā vīraṇebhyaḥ kaṭasya tantubhyaḥ paṭasyeti.
58. See my conclusion for an elaboration on a reading of Vasubandhu’s view as a way of structuring relations among scientific theories.
59. In AKBh, the question of just what kind of cause can produce just what kind of result is debated in the context of the discussion of an “equal and immediately preceding condition,” samanantarapratyaya, at II.62. See Griffiths 1991.
60. On the contrary, he adopts this rule as part of his causally based definition of a logical reason, a “cause” (hetu; gtan tshig). See chapter 4.
61. AKBh 189.14–18: idam ayuktaṃ vartate | kim atrāyuktam | yad asiddhaṃ sādhanāyodāhriyate | kim atrāsiddham | anyo vīraṇebhyaḥ kaṭo ’nyaśca tantubhyaḥ paṭa iti | ta eva hi te yathāsaṃniviṣṭās tāṃ tāṃ saṃjñāṃ labhante | pipīlikāpaṅktivat | kathaṃ gamyeta | ekatantusaṃyoge paṭasyānupalambhāt | ko hi tadā sataḥ paṭasyopalabdhau pratibandhaḥ | akṛṛttau paṭabhāgo ’tra syān na paṭaḥ | samūhamātraṃ ca paṭaḥ syāt | kaś ca tantubhyo ’nyaḥ paṭabhāgaḥ.
62. AKBh 189.18–21: anekāśrayasaṃyogāpekṣaṇe daśāmātrasaṃyoge paṭopalabdhiḥ syān na vā kadācit | madhyaparabhāgānām indriyeṇāsaṃnikarṣāt | kramasaṃnikarṣe cāvayavānāṃ cakṣuḥsparśanābhyām avayavavijñānaṃ na syāt | tasmād krameṇa saṃnikarṣād avayavivyavasāyād avayaveṣv eva tad buddhir alātacakravat.
63. This refers back to where the opponent said that there can be no birth from a different class. If a variegated form arises from threads of individual colors, then this rule is violated.
64. The point seems to be that if one sees an individual thread, one must be able to distinguish different shades or colors—which means one sees a variegated form even when the thread is all the same color.
65. AKBh 189.21–24: bhinnarūpajātikriyeṣu tantuṣu paṭasya rūpādyasaṃbhavāt | citrarūpāditve vijātīyārambho ’pi syāt acitre ca pārśvāntare paṭasyādarśanaṃ citradarśanaṃ vā | kriyā ’pi citrety aticitram | tāpaprakāśabhede vā ’gniprabhāyā ādimadhyānte tadrūpasparśayor anupapattiḥ.
66. This view of perception as always a false construction will be useful to keep in mind in the upcoming chapters, as we see Vasubandhu advancing the utility of perceptually-based reasoning even while allowing that perception employs false constructions.
67. Things arise from multiple causes coming together.
68. AKBh 189.24–190.2: paramāṇvatīndriyatve ’pi samastānāṃ pratyakṣatvaṃ [190] yathā teṣāṃ kāryārambhakatvaṃ cakṣurādīnāṃ ca taimirikāṇāṃ ca vikīrṇakeśopalabdhiḥ | teṣāṃ paramāṇuvad ekaḥ keśo ’tīndriyaḥ.
69. A more literal reading here would be “there is no thought of them with respect to burnt wool, cotton, safflower, and saffron, etc.”
70. AKBh 190.2–190.8: rūpādiṣv eva ca paramāṇusaṃjñāniveśāt tadvināśe siddhaḥ paramāṇuvināśaḥ | dravyaṃ hi paramāṇur anyac ca rūpādibhyo dravyam iti na teṣāṃ vināśe tadvināśaḥ siddhyati | ayuktam asyānyatvam yāvatā na nirdhāryate kenacit imāni pṛthivyaptejāṃsi ima eṣāṃ rūpādaya iti | cakṣuḥsparśanagrāhyāṇi ca pratijñāyante dagdheṣu corṇākarpāsakusumbhakuṅkumādiṣu tadbuddhyabhāvād rūpādibhedeṣv eva tadbuddhiḥ | pākajotpattau ghaṭaparijñānaṃ saṃsthānasāmānyāt paṅktivat | cihnam apaśyataḥ parijñānābhāvāt | ko vā bālapralāpeṣvādaraḥ iti tiṣṭhatu tāvad evāpratiṣedhaḥ.
71. Against Schmithausen (1967), who sees unity between the AKBh, the Viṃś, and the Triṃś, but holds with Frauwallner’s “two Vasubandhus” thesis in distinguishing, doctrinally, the Asaṅga commentaries, I believe that the MAVBh, the MSBh, and the TSN all cohere with the causal linearity thesis advanced here. It is, of course, in principle possible that followers or imitators of Vasubandhu adopted his arguments and his approach. My point is to show the conceptual unity of these texts, which display a consistent intellectual agenda.
4. KNOWLEDGE, LANGUAGE, AND THE INTERPRETATION OF SCRIPTURE: VASUBANDHU’S OPENING TO THE MAHĀYĀNA
1. AKBh 299.1–2: vayam api brūmo ’sty atītānāgatam iti | atītaṃ tu yad bhūtapūrvam | anāgataṃ yat sati hetau bhaviṣyati | evaṃ ca kṛtvā ’stīty ucyate na tu punar dravyataḥ.
2. The question of the relation between Buddhist epistemology and the path to liberation is complex and contested. See Kapstein 2001c, Jackson 1995, and works cited therein for a range of problematics. My approach here is to suggest that the basic concepts in Buddhist epistemology are an expression of Vasubandhu’s understanding of how knowledge operates in relation to basic Buddhist doctrine and its sources in the sūtras. If this is so, we have a full justification of the importance of Buddhist epistemology not as a tool for attaining higher knowledge, but as a working out of intellectual problems that stem from the interpretation of Buddhist scripture.
3. NBh pp. 23.14–24.2: yajjātīyasyārthasya sannikarṣāt sukham ātmopalabdhavān tajjātīyam evārthaṃ paśyann upādātum icchati seyam ādātum icchā ekasyānekārthadarśino darśanapratisandhānād [24] bhavantī liṅgam ātmanaḥ | niyataviṣaye hi buddhibhedamātre na sambhavati dehāntaravad iti. Kapstein (2001:378ff) provides a translation of this passage, and Uddyotakara’s brilliant, innovative subcommentary, also studied in juxtaposition with AKBh IX.
4. Here we see the assumption of newness as intrinsic to knowledge that is central to Buddhist epistemologists later on as well. When knowledge is understood as an event rather than a disposition, it is crucial that something be previously unknown for it to be considered an event of “knowing” (jñāna or pramāṇa). This suggests that a better translation for these terms might be “realizing” or “dawning.”
5. On the issue of continuity among Buddhists, see Griffiths 1991 and other sources noted in chapter 2, note 69. This passage, and Vasubandhu’s attention to it, show that the issue was of great importance to Buddhist defenses against non-Buddhists. If the Yogācāra-vijñānavāda represented a stable resolution to this problem, then it makes sense to see this tradition’s emergence as not just an internal development of Buddhist doctrine, but at least in part a reformation in response to external critiques.
6. See Kapstein 2001a:367. I am not claiming that the chapter as a whole is intended as a response to this passage, but Vasubandhu is clearly aware of the content of the argument—whether in Vātsyāyana’s words or not, I cannot say.
7. AKBh 472.15: smṛṣayasaṃjñānvayāc cittaviśeṣāt.
8. AKBh 472.20–22: na hi tayoḥ saṃbandho ’sti akāryakāraṇabhāvād yathaikasaṃtānikayoḥ | na ca bhrūmo ’nyena cetasā dṛṣṭam anyat smaratīti | api tu darśanacittāt smṛticittam anyad utpadyate.
9. This refers to anything outside of the Buddha’s teachings, and shows the final chapter to be in fact an extended commentary on the final verse of chapter VIII.
10. AKBh 461.2–4: kiṃ khalvato ’nyatra mokṣo nāsti | nāsti | kim kāraṇam | vitathātmadṛṣṭṣṭatvāt | nahi te skandhasaṃtāna evātmaprajñaptiṃ vyavasyanti | kim tarhi | dravyāntaram evātmānaṃ parikalpayanti ātmagrāhaprabhavāś ca sarvakleśā iti.
11. As Franco (2001:291, 296) points out, Dharmakīrti essentially reduces the Buddha’s teachings to the anātman doctrine. This is a natural extension of Vasubandhu’s emphasis.
12. I call this a formal proof, but I do not intend to suggest that he has adopted the structure of a prāmāṇika syllogism. I say “means of knowledge” for pramāṇa for ease of comprehension (relative ease). I do not have a strong attachment to this translation.
13. This introduces a second example of inference, this time to prove the point at hand.
14. Vasubandhu uses the term niścaya, “determination” only very occasionally, and always to refer to the elimination of doubt. In a couple of instances it refers to the stages on the path in which various kinds of doubt are eliminated, and in others to ordinary certainty (for instance, a lack of doubt that the one you are shooting at is a living being).
15. AKBh 461.4–20: kathaṃ punar idaṃ gamyate skandhasaṃtāna evedam ātmābhidhānaṃ vartate nānyasminn abhidheya iti | pratyakṣānumānābhāvāt | ye hi dharmāḥ santi teṣāṃ pratyakṣam upalabdhir bhavaty asaty antarāye | tadyathā ṣaṇṇāṃ viṣayāṇāṃ manasaś ca | anumānaṃ ca | tadyathā pañcānāṃ indriyāṇām | tatredam anumānam sati kāraṇe kāraṇāntarasyābhāve kāryasyābhāvo dṛṣṭo bhāve ca punarbhavas tadyathāṅkurasya | saty eva vābhāsaprāpte viṣaye manaskāre ca kāraṇe viṣayagrahasyābhāvo dṛṣṭaḥ punaś ca bhāvo ’ndhabadhirādīnām anandhābadhirādīnāṃ ca | atas tatrāpi kāraṇāntarasyābhāvo bhāvaś ca niścītyate | yac ca tatkāraṇāntaraṃ tadindriyam ity etad anumānam | na caivam ātmano ’stīti nāstyātmā. More-or-less the same argument is made in much shorter form with reference to the unreality of the unconditioned at AKBh II.55d.
16. The fragments that exist are sufficient to show that Vasubandhu’s work was a strong precedent to Dignāga in Buddhist pramāṇa studies. Anacker (1998:31) credits Vasubandhu’s writings with “the dawn of Indian formal logic.” The main fragments are the sections from his Vādavidhi, cited by Dignāga and Jinendrabuddhi extant in Tibetan translation, collected in Frauwallner 1957, and introduced and translated in Anacker 1998:31–48. I will discuss one fragment below. It is of course possible to doubt that these fragments were written by the person who composed the AKBh, but as my first chapter explains, I am pursuing the assumption of continuity among these works. Verhagen (2008:253–258) provides an edition of the section from the VyY on the three pramāṇas.
17. Vasubandhu appeals to the three means of knowledge in several passages in the AKBh, and also in the Vyākhyāyukti. In this he is in agreement with a number of Yogācāra texts, as observed by Tucci (Verhagen 2008:245). It would appear that Yogācāra philosophers before Vasubandhu were already interested in establishing a Buddhist approach to pramāṇa.
18. Though perhaps not. Vasubandhu identifies two categories of objects, one for each pramāṇa, and does not allow that a third type of object exists which might be unidentifiable by the two pramāṇas he mentions.
19. Thus, Dignāga was not the first to suggest that different pramāṇas—different means of knowledge—come to know different sets of objects.
21. This doctrine would appear to be a formalization of the understanding of the real as what has causal efficacy. Anacker (1998:32) cites Vasubandhu’s Vādavidhi as the first Indian text to suggest that a proof requires invariable concomitance (avinābhāva) between the reason and the proven property. For Vasubandhu, like Dharmakīrti, that necessary concomitance is rooted in causality, understood in this simple, manipulationist mode. Wiggle one thing and the other changes. This is what Vasubandhu says is lacking in the quest for a proof of a self. This is simply an extension, for him, of saying that there is nothing that can only be explained as having been caused by a self.
Although I do see a direct line here from Vasubandhu to Dharmakīrti, I would emphasize that I do not intend to impose on Vasubandhu a reification of causal powers or an empiricist foundationalism, as are often attributed to Dharmakīrti. We will see shortly that Vasubandhu’s understanding of causal relations coheres with Woodward’s relativistic approach, according to which what is termed a “cause” is always relative to the investigative framework. Like any other imposition across dharmas, causality is a conceptual construction, and not substantial. This is true even though it is only through appeals to causality that inference provides us the means for calling some entities substantially real.
23. As discussed in chapter 2, Vasubandhu also refuses to accept that the future must exist as an object of a Buddha’s awareness. He says that Sautrāntikas hold that merely by intending it, the Buddha knows whatever he chooses to know, but they do not base this on any particular causal story of how a Buddha knows the future. Rather, they base this view on the scripture that says that the qualities of a Buddha’s awareness are incomprehensible. AKBh 99.10: tasmāt sarvam icchāmātreṇa bhagavān jānātīti sautrāntikāḥ | acintyo hi buddhānāṃ buddhiviṣaya ity uktaṃ bhagavatā. See below for discussion of the incomprehensible (acintya).
24. This is clear also in the proof of momentariness (see below). Yet Vasubandhu’s commonsense, relative preference for the evidence of perceiving something over inferring it (which often relies on the analysis of perceptions in any case) should not be taken to indicate a deeper claim to anything resembling empiricist foundationalism. Perception may be the best mode of knowledge, and it may be the source of a Buddha’s perfect knowledge, but it does not give us perfect, direct, or even accurate knowledge. As we have seen in chapter 3, Vasubandhu believes that we fail to see things as they are even through perception. If atomic dharmas are imperceptible, he argued, then perception cannot give you knowledge of dharmas. We do not perceive what we think we perceive. Nonetheless, as I read Vasubandhu, even if perceptual knowledge is shot through with false conceptualization, it is the basis for our knowledge of the world and its living beings, and we cannot pretend to do without it. Inference relies upon and builds upon perceptual knowledge, but only by moving yet a further step into conceptual construction and away from causal reality.
25. That is, it is the source of the line that extends through Dignāga to Dharmakīrti and on. It is not the first Buddhist text to present pramāṇa discussions, however. See Verhagen 2008:244–247 and sources cited therein.
26. Frauwallner 1957:135–136: de ’dra ba dang med na mi ’byung ba’i chos nye bar ston pa gtan tshigs so zhes pa ni | don gang sgra mi rtag pa nyid la sogs pa bsgrub par bya ba de ’dra ba ste | de’i rigs can med na don gang zhig ’gar yang ’byung ba ma yin pa ste | rtsol ba las byung ba nyid ni mi rtag pa nyid lta bu dang du ba ni me lta bu’o zhes pa de ni de ’dra ba med na mi ’byung ba’i chos can te | de nye bar ston pa ni ’dis nye bar [136] bstan par bya’o zhes pa’i tha tshig go || dper na brtsal ma thag tu byung ba nyid kyi phyir zhes pa ’di lta bu la sogs pa ste de ni gtan tshigs so || gang gis nye bar ston pa ma yin pa de ni gtan tshigs ma yin te | dper na sgra mi rtag ste mig gis gzung bya nyid kyi phyir zhes pa ’di lta bu la sogs pa’o.
27. It proves a philosophical problem for Buddhists to account for certitude under an epistemology that denies the reality of conceptual constructions. See Arnold 2005, 2012. Reading through Vasubandhu, I find the basic Buddhist view to be that concepts are not real because they are not causally viable except to the degree that they are made up, constructed, by mental events. This means that ideas themselves are not causes; the minds that hold them are. But this talk of “holding” imports the basic illusion of selfhood into our talk about concepts. There is no agent, no subject who “holds” an idea. The mental event that constructs an appearance is the “holding.” As I will explain in chapter 5, Vasubandhu argues that, like hairs that appear before diseased eyes, the appearance of “holding” an idea is a delusory by-product of the perceptual structure of awareness (see also Gold 2006 for more detail on this). Thus, there is no distinction between “reasons” and “causes,” because “reasons” are real only as mental events. There is, thus, no distinct “logical space of reasons” (Sellars, cited in Arnold 2005:54 and throughout Arnold 2012)—but until liberation, it unavoidably appears as though there is.
28. Cardona (1967–1968:352) shows that the grammarians used these concepts to establish a causal relationship between words or grammatical forms and the resultant arising of understanding of their meanings. Vasubandhu applies the process in a quite similar way to perception—establishing a causal relationship between its causes and the resultant awareness.
29. To put it another way, we could say that the requirement of a linear approach to causality transforms anvaya and vyatireka into the structure for a hypothetical manipulation. This is Vasubandhu’s understanding of the Buddhist doctrine of the impermanence of all conditioned things: if something is permanent, it cannot be engaged with the flow of causes, and therefore it cannot be a true, substantial existent. The converse is that every substantial existent can be detected in some hypothetical manipulation.
30. Let us postpone the discussion of the implications of the ontological status of causality itself as, ironically, a mere conceptual construct that nonetheless verifies the substantial nature of entities. Ultimately, Vasubandhu finds this paradox best addressed in the Yogācāra rejection of the very concept of “reality.” For the time being, it only means that to say that something is unreal or uncaused does not mean it does not occur, only that it fails to meet the criterion of substantiality. Things surely pass out of existence. But their “nonexistence” is not a substantial thing.
32. Williams (1998a:78–80) discusses the nonexistence of negation in the Tibetan interpretation of Śāntideva; we may move this idea back to Vasubandhu. Williams makes it clear that Śāntideva shares Vasubandhu’s view that negation is in the mind, even to the point of affirming the inherent conceptual connection between the negation and the negandum.
33. AKBh 193.10–14: dṛṣṭo vai kāṣṭhādīnām agnyādisaṃyogād vināśaḥ | na ca dṛṣṭād gariṣṭaṃ pramāṇam astīti | na ca sarvasyākasmiko vināśaḥ | kathaṃ tāvat bhavān kāṣṭādīnām agnyādisaṃyogād vināśaṃ paśyāmīti manyate | teṣāṃ punar adarśanāt | saṃpradhāryaṃ tāvad etat | kim agnisaṃyogāt kāṣṭhādayo vinaṣṭā ato na dṛśyante utāho svayaṃ vinaṣṭā anye ca punar notpannā ato na dṛśyante | yathā vāyusaṃyogāt pradīpaḥ pāṇisaṃyogād dhaṇṭāśabda iti | tasmād anumānasādhyo ’yam arthaḥ.
34. AKBh II.55 (91.14–21). In this passage there is an even stronger point to be made about the connection between Vasubandhu’s denial of nonexistents and his view of the unitary causal line. For the reason the Sarvāstivādins give that nirvāṇa must be a cause is that it is the experiential object of the Buddha’s awareness. In answer, Vasubandhu makes the distinction, discussed in chapter 2 with regard to past and future experiential objects, between “producer conditions” and merely experiential objects. The latter, he says, are not properly causes. Here he uses slightly different terms: instead of distinguishing a “producer condition” (janakaḥ pratyayaḥ) from a “merely experiential object” (ālambanamātra), he quotes the Buddha’s claim that the conditions “for arising” (utpādāya) are impermanent, and says that this is not the same as saying that all conditions are impermanent. The point is the same: the Buddha did not intend to say that objects of awareness, which are named “experiential object conditions” of that awareness, must be producers of that awareness. Hence, they do not need to exist in order to be mental objects. Conversely, to return to the point at hand, nonexistent entities can be known, even by a Buddha, without being, thereby, substantial. This point, and its connection to the argument over the three times, is confirmed when, in answer to the objector’s concern, “If the unconditioned were simply nonexistent, then awareness of space and nirvāṇa would have nonexistent experiential objects,” Vasubandhu replies, “We will consider this in the context of considering the existence of past and future.” AKBh 93.13–14: yady asaṃskṛtam abhāvamātraṃ syād ākāśanirvāṇālambanavijñānam asadālambanaṃ syād | etad atītānāgatasyāstitvacintāyāṃ cintayiṣyāmaḥ.
35. Von Rospatt (1995) shows that the argument from the unreality of a nonexistent is Vasubandhu’s distinctive contribution to the history of the doctrine of momentariness.
36. It would be admissible to take the sixth case of yaś cābhāvas tasya kiṃ kartavyam as an instrumental, which would give us, “What result is to be made by that which is nonexistent?” or “What can that which is nonexistent do?” Presumably, however, this is not quite Vasubandhu’s point, since under such a reading the nonexistent would be the cause. I take the argument to be that the nonexistent is either the result in question or, possibly, that neither cause nor result are relevant to something that does not exist.
37. AKBh 193.4–9: syād etad eva yadi sarvasya kṣaṇikatvaṃ sidhyet | siddham evaitat viddhi | kutaḥ | saṃskṛāvaśyaṃ vyayāt || (2) [IV.2d2] ākasmiko hi bhāvānāṃ vināśaḥ | kiṃ kāraṇam | kāryasya hi kāraṇaṃ bhavati | vināśaś cābhāvaḥ | yaś cābhāvas tasya kiṃ kartavyam | so ’sāv ākasmiko vināśo yadi bhāvasyotpannamātrasya na syāt paścād api na syād bhāvasya tulyatvāt.
38. AKBh III.59d; see chapter 2.
39. This is true about this passage, even though, as mentioned above, Vasubandhu generally shares in the Yogācāra tendency of appealing to three pramāṇas. See Verhagen 2008:244–247.
40. Without Dignāga’s innovative terminology, on which see Arnold 2005:13–31.
41. Without Dharmakīrti’s innovative definitions, on which see Dreyfus 1997: esp. 65–67. This does not, however, make Vasubandhu into an empirical foundationalist, since for him causality itself is a useful, but ultimately false, conceptual construction.
42. I believe Dignāga’s apoha theory derives its impetus here as well, from Vasubandhu’s theory of the inferential construction of nonexistence, though to explain that complex theory would take us too far afield. For the current state of the art, see Siderits et al. 2011. The Nyāya, for their part, took the opposite path. They came to reify the nonexistent, to add it to their list of entities. For to admit to the conceptual construction of nonentities was to give in to the Buddhists, who (with Dignāga) used this to construct a nonrealist theory that supported the entire edifice of language and conceptualization.
46. This is Vasubandhu’s understanding of the Buddhist doctrine of the impermanence of all conditioned things: if something is permanent, it cannot conceivably be engaged with the flow of causes, and therefore it cannot be a true, substantial existent. The converse is that every substantial existent can be detected in some hypothetical manipulation.
The fact that causality is a dependence relation whose postulation requires only hypothetical, not actual, intervention helps to address a number of misunderstandings of Vasubandhu’s reductionist approach, especially as it applies to abstract entities. As just one example, Vasubandhu denies that a soul must exist to exercise “ownership” over experiences, or that “ownership” exists in the abstract, over and above the owner and the entity owned. He says, instead, that the relationship between owner and owned is causal—the owner is the one who milks the cow, and takes it to market, and so on.
Jonardon Ganeri’s recent book cites, favorably, later Nyāya philosophers who criticize this idea of Vasubandhu’s that ownership can be reduced to causal relations between owner and owned. As they say, ownership is not any particular action or set of actions, but rather a capacity that an owner has, to do what he likes with his property. To push the point, then, Ganeri writes, “A capacity might go unexercised, so there may be no causal connections corresponding to it” (2013:176). Now that we know that Vasubandhu sees causality as a dependence relation determinable through a merely hypothetical intervention, and not any actual set of events, the fact that the capacity of ownership may be unexercised is irrelevant to the validity of the claimed causal relation. Granted, we speak conventionally as though there is some “ownership” that Chaitra has, over and above Chaitra’s various relevant acts and decisions about his cow. But you cannot say that an appearance is a reality by appealing to a mere convention.
47. A particularly clear example is found in Madhava (Stoker 2004), but the point could be made just as well about the distinction between Mīmāṃsā and Vedānta, or about the distinction between āstikas and nāstikas. Scriptural interpretation, and the methods of justifying scriptures, are crucial issues that determine philosophical positions.
48. “The knower of the text is beloved of the gods, but is not a knower of meanings! The Lord says the meaning is the refuge.” AKBh 76.24–25: granthajño devānāṃ priyo na tv arthajñaḥ | arthaś ca pratiśaraṇam uktaṃ bhagavatā.
50. AKBh 143.17. Interestingly, he also accuses his opponents of doing the same (which they would, presumably deny), saying that they would accept an emendation of the text, or a figurative interpretation.
51. AKBh 64.7: lobhasyāvinodanaṃ.
52. AKBh 468.1–8; Kapstein 2001a:360–361, section 6.3. Pāsādika (1989:128) identifies the source in the Bhārahārasūtra.
53. This is a claim that I have begun to document using statistical analysis of scriptural citations and arguments in the AKBh, but it is not yet decisively proven. The sources to exemplify Vasubandhu’s approach to scriptures used in what follows are primarily drawn from the section of the AKBh with the longest, most densely packed series of scriptural citations, which is verses 4–44 of chapter III.
55. AKBh 91.23: sūtrāṇi ca bahūny antarhitānīti katham etan nirdhāryate nokta iti. This AKBh passage is proposed by an opponent; but Vasubandhu makes the same point himself in the KSP and the VyY.
59. AKBh 133.15. Cox (1992) reads this as only a principle of Vaibhāṣika scholasticism, but I take it that it coheres with Vasubandhu’s overall approach even as a Sautrāntika. Abhidharma is intended to determine the definitive truth—what is substantially real. This point describes the purpose of this type of text, quite aside from declarations of the authority of any given text of the genre.
60. AKBh IX (466.5–471.21) explains the Buddha’s reasons for famously having refused to address certain difficult topics, calling them “undeclared” (avyākṛta). See Kapstein 2001a:363–366 for a translation. There are, in fact, many reasons for the Buddha’s having refrained from speaking forthrightly. Nance (2012:135) provides a translation from Vasubandhu’s VyY of the “eight purposes for rhetorical strategies” in the Buddha’s discourses. The lesson from all of this is that when something is “unsaid,” it is for a specific reason: nonspeech has an intention just like speech does. Things that are not spoken of are not necessarily ineffable.
61. AKBh 115.4–5. As another example of this, two competing sūtras are in evidence at AKBh 143.2ff, where the Sautrāntikas have it out with the Sarvāstivādins about the nature of contact. One sūtra says that the coming together of the three kinds of dharmas for each sense (organ, object, and consciousness) just is the contact (143.17: ya eṣāṃ trayāṇāṃ saṃgatiḥ saṃnipātaḥ samavāyaḥ sa sparśaḥ), and the Sautrāntika takes this on its face as a definition of contact as just the activity of the distinct entities. The other sūtra passage (143.8–10) contains a list of “six sixes” (ṣaṭṣaṭko) which include the organs, objects, and consciousnesses along with the “six contacts” (ṣaṭśakāyāḥ), “six feelings,” and “six thirsts”—which the Sarvāstivādins take as evidence that these are distinct things. Vasubandhu’s reply is that if these are taken as a list of six separate things, then the six feelings and the six thirsts should be distinct from even the dharmāyatana, which is absurd.
63. Did the Buddha have to have “acted locally?” There is no reason, in principle, that a Buddha could not foresee that his words would be written down and misinterpreted by textual literalists in the future—which he did less than he perhaps could have to prevent. But the Buddha took it as his responsibility to focus on transforming the minds of his students, even if that meant confusion for his students’ students’ students’ students’ students. This may be a reasonable choice, if made on the assumption that the decline of the dharma is unavoidable.
65. VyY 227.9–18: gzhan yang theg pa chen po sangs rgyas kyi gsung ma yin no zhes zer na | sangs rgyas kyi gsung gi mtshan nyid ci yin zhes brgyal zhing brtag par bya’o || smras pa | sangs rgyas kyi gsung nyid du sde pa bco brgyad dag gis yongs su bzung ba gang yin pa’o zhe na | ’o na don dam pa stong pa nyid la sogs pa’i mdo sde bdag med pa dang ldan pa ’phags pa kun gyis bkur ba’i sde pa dag gis khas mi len pa dag dang | srid pa bar ma do dang ldan pa srid pa bdun zhes bya ba de lta bu la sogs pa sa ston pa’i sde la sogs pa rnams kyis khas mi len pa gang yin pa de dag sangs rgyas kyi gsung du mi ’gyur ro.
67. See Gold (forthcoming) for my discussion of the Yogācāra response to the interpretation of Mahāyāna scriptures, in which Vasubandhu played a key clarifying role.
68. This point is made by Cabezón (1992) in his pioneering study of the VyY.
69. VyY 236.18–237.3: nged ni gang zak kun rdzob tu yod kyi rdzas su ni ma yin te | phung po rnams la de’i ming gdags pa’i phyir ro || las dang rnam par smin pa dag ni kun rdzob tu rdzas su yod | don dam par ni med de | ’jig rten pa’i shes pa’i yul yin pa’i phyir ro | dam [237] pa ni ye shes ’jig rten las ’das pa yin te | de’i don yin pas don dam pa’o || de gnyis kyi rang gi mtshan nyid ni de’i yul ma yin te | de’i yul ni brjod du med pa’i phyi’i mtshan nyid yin ba’i phyir ro.
71. VyY 237.15–19: yang theg pa chen po pa kha cig thams cad rang gi mtshan nyid du ni med pa kho na yin la | kun rdzob tu ni bcom ldan ’das kyis chos rnams yod pa nyid du bstan to zhes sgra ji bzhin pa nyid kyi don yin par brjod pa gang yin pa de dag la ji skad bstan pa’i rtsod pa ’di yang ’byung bar ’gyur ro.
72. VyY 239.18–22: brjod du med pa’i mtshan nyid dag chos rnams thams cad yin na ni | byis pa rnams kyis kun tu brtags pa bzhin du brjod pas | rjod par byed pa de’i rjes so ’brangs nas ’phags pa rnams kyis kyang min med pa rnams ming gis yongs su brjod pa yin par rung ngo. This point is the culmination of Vasubandhu’s argument in this passage from the VyY, translated in appendix E. See also Cabezon 1992 on the VyY’s view of Buddhist scripture.
73. This sentence, from Vasubandhu’s VyY quotation, is different from SNS.
74. VyY 239.22–240.10: de ltar na dgongs pa nges par ’grel ba’i mdo las | rigs kyi bu ’dus byas zhes bya ba de ni ston pas btags pa’i tshig yin te | ston pas btags pa’i tshig gang yin pa de ni kun tu [240] rtog pa las byung ba tha snyad du brjod pa yin la | kun tu rtog pa las byung ba tha snyad du brjod pa gang yin pa de ni kun tu rtog pa sna tshogs kyi tha snyad du brjod pa gtan yongs su ma grub pa’i phyir ’dus byas ma yin no || ’dus byas ji lta ba bzhin du ’dus ma byas kyang de bzhin no zhes de bzhin du sbyar ro || ’dus byas dang ’dus ma byas su ma gtogs pa gang ci brjod kyang de yang de dang ’dra ba nyid du ’gyur ro || de yang brjod pa ni dngos po med pa can yang ma yin te | dngos po de yang gang zhe na | ’phags pa rnams kyis ’phags pa’i shes pa dang ’phags pa’i mthong bas brjod du med par mngon par rdzogs par sangs rgyas pa gang yin pa ste.
75. Siderits (2009) discusses a similar point regarding the difficulty of speaking of ultimate and conventional realities—within the same frame of reference or in the same sentence. I read Vasubandhu’s solution as something akin to what Siderits calls “panfictionalism” and attributes to Jñānaśrī—the view “that anything expressible is a conceptual fiction” (67). The difficulty with this solution, as Siderits notes, is that it must somehow still maintain distinctions within the conventional—a problem that I see Vasubandhu addressing through appeals to manipulationist causal framing and the pragmatics of scriptural interpretation.
76. Dunne 2004 and McClintock 2010. Dreyfus (1997) refers to “ascending scales of analysis.”
77. AKBh 126.7: ata eva coktaṃ bhagavatā acintyaḥ sattvānāṃ karmavipāka iti.
78. AKBh 121.5: acintyo hi dharmāṇāṃ śaktibhedaḥ.
79. AKBh 99.10: tasmāt sarvam icchāmātreṇa bhagavān jānātīti sautrāntikāḥ | acintyo hi buddhānāṃ buddhiviṣaya ity uktaṃ bhagavatā.
80. Kapstein 2001a:357–367.
