What are the “Purposes” of Buddhist Sūtras? From Vasubandhu’s Logic of Exegesis (Vyākhyāyukti)

As its name implies, Vasubandhu’s Vyākhyāyukti (VyY) explains the logic or methodology (yukti) of exegesis or sūtra interpretation (vyākhyā) and only survives in a Tibetan translation. In recent years, research on this treatise has been gradually accumulating. However, due to the difficulty of the Tibetan translation, some of the arguments therein have been misunderstood. In this article, after reviewing the general framework of Vasubandhu’s method of interpreting the sūtras, I will present a newly discovered parallel regarding his discussion of the “purpose, prayojana” of the sūtras and reread it through a close philological examination of various sources in Sanskrit and Tibetan. Thus, this article will first elucidate the details of Vasubandhu’s explanation of the purpose of the Buddha’s preaching using synonyms and will clarify an aspect of his views on the Buddha’s word. In addition, concerning this “purpose,” I will elucidate the characteristics of the final chapter of this text, which provides stories about hearing the Buddha’s words with respect. By doing so, I would like to reveal the characteristics of the VyY as a manual for vivid preaching.


Introduction
The Vyākhyāyukti (VyY) by Vasubandhu (India,, 1 which explains the methodology of sūtra interpretation, only survives in a Tibetan translation, along with a voluminous commentary by Guṅamati, the Vyākhyāyuktiṭīkā (VyYT). Two articles by Yamaguchi Susumu were the first to focus on this treatise in the modern academic world. Articles by Matsuda Kazunobu, Honjō Yoshifumi, José Ignacio Cabezón, Peter Skilling, Jong Choel Lee, Peter Verhagen, and others followed. Specifically, Skilling's (2000) detailed overview is still informative for scholars. 2 In the last decade or so, it is worth noting that Chapters 1, 2, 4, and 5 have been translated (sometimes with critical editions) by Richard Nance, Ueno Makio, and the author of the present study. In other words, especially in the last decade or so, we have moved from the stage of a summary introduction to a previously unknown work to the stage of being able to perform a close reading of the text based on a textual critique of that same text. However, due to the material limitations of this treatise, namely, the original Sanskrit has been lost and only a Tibetan translation remains, there are often misunderstandings and points that have not been fully clarified. Furthermore, we need to examine each chapter organically in the context of the whole VyY and within the broader tradition of sūtra commentaries: the VyY consists of five chapters, each of which is closely related to the others, and so it would be inappropriate to present Vasubandhu's method of interpreting the sūtras on a piecemeal basis without understanding its structure as a whole. It should also be noted that the exegetical method of the VyY is not a purely original invention, but rather an extension of the Yogācāra school or even of the Abhidharma tradition. 3 This paper presents an overview of the chapters of the VyY and discusses the first of the five aspects of sūtra interpretation, "purpose, prayojana." I am revisiting this 1 For details, see Deleanu (2019). There are many issues concerning Vasubandhu, such as his chronology and the theory of two or several Vasubandhus, but they are not relevant to this paper. At the very least, we wish to confirm that Vasubandhu in this article refers to the person who wrote the VyY and the subsequent PSVy, on both of which Guṅamati wrote commentaries, and all of them are preserved in Tibetan translation. 2 For some of the recent studies on the VyY, see Ueno (2021b). 3 Cf. Ueno (2009). In this connection, I would like to comment on Hanner (2020), which deals with "scripture and scepticism in Vasubandhu's exegetical method," as his title shows. He first raises a question-"Is there a place, according to Vasubandhu, for scepticism in scriptural interpretation made by the religious tradition itself"-and deals with "religious scepticism" (p. 131). However, first, the relevance of scepticism to the study of Vasubandhu is questionable. Moreover, much of the discussion of "Vasubandhu's exegetical method" is not limited to him but can be traced back to the Yogācāra school, Abhidharma literature, or even to early Buddhist scriptures. Let me give only one example in this regard. On p. 137, concerning the VyY (=VY), Hanner writes as follows: "There is another exposition in the VY which displays this approach, in which Vasubandhu enumerates five benefits that come about from devotedly listening to the Buddha's teachings. These are: (1) hearing the unheard, ..." fn. 6: "VY 116b4-5, p. 257: ma thos pa thos par 'gyur ba dang|" However, this phrase is immediately preceded by: "bcom ldan 'das kyis chos mnyan pa la phan yon lnga gsungs te| (The Bhagavat preached the five benefits of hearing the Dharma [as follows])." In other words, this is merely Vasubandhu quoting an āgama. Therefore, it is inappropriate to discuss this as a position unique to Vasubandhu. One should instead say: "Bhagavat enumerates ..." section because I have found new parallel material written in Sanskrit concerning some of these passages. This finding will allow us to reread the VyY and, by extension, clarify an aspect of Vasubandhu's views on the Buddha's word. This point of "purpose" is also mentioned again in Chapter 5. Since some researchers have misunderstood the position of Chapter 5, we will read the relevant section. Thus, this paper aims to reconsider Vasubandhu's discussion of the "purpose" of the sūtras, reconfirm its position within the framework of sūtra interpretation in the VyY, and present the characteristics of the VyY as a lively manual for preaching.
What is the Purpose of Buddhist Sūtras?

Exegesis Based on Five Aspects
In the first part of the first chapter of the VyY, Vasubandhu says that a qualified sūtra commentator should comment on the sūtras in accordance with five aspects. VyY, D30b3, P33b5-6: mdo don smra ba dag gis ni|| dgos pa bsdus pa'i don bcas dang|| tshig don bcas dang mtshams sbyar bcas|| brgal lan bcas par bsnyad par bya|| Although the previously noted parallel clause in the AAA 4 did not fully match the original clause in the VyY, Tomabechi has reported that a recently discovered text, Abhayākaragupta's Āmnāyamañjarī, provides an entirely consistent Sanskrit version.
As we will see in Section "The structure of the VyY in relation to the methodology of sūtra interpretation", the VyY is itself structured to explain these five aspects by citing specific examples. Among them, the purpose is placed first, and its reasons are summarized in verse by Vasubandhu himself. As Skilling (2000, p. 331) has pointed out, this verse appears in Chapters 1 and 5 (see Section "The relation between "purpose" and hearing the Buddha's word respectfully") and is cited by the Arthaviniścayasūtranibandhana (AVSN).

Purpose
In Chapter 1 of the VyY, on the first of the methods of sūtra interpretation, the purpose is stated as follows: VyY, D31a5-6, P34b3-4: de la mdo sde'i dgos pa ni bsdu na rnam pa bzhir rig par bya ste| (1) kun tu rmongs pa rnams la yang dag par bstan pa dang| (2) bag med pa rnams la yang dag par len du gzhug pa dang| (3) kun tu zhum pa rnams la yang dag par gzengs (gzengs] D; gzeng P) bstod pa dang| (4) yang dag par zhugs pa rnams la yang dag par dga' bar bya ste| My translation of this passage is almost identical to those of Honjō (2001) and Ueno (2017). 7 Of them (among the five), in brief, the purpose of a sūtra should be known to be of four kinds: (1) to expound [the teaching] to the stupid [trainees]; (2) to 6 Ms, G, etc., are abbreviations for Sanskrit manuscripts of the AVSN. A brief history of the study of this text is as follows: Samtani published his edition in 1971, de Jong wrote a review with some suggested corrections, Honjō (1989) provided a Japanese translation and further suggested corrections, and Samtani published his English translation in 2002 without reference to Honjō's work. For details, see Horiuchi (2021). 7 It is important to note that Ueno pointed out the similarity of this paragraph to ASBh, 147.23-25. This is because at least the words describing the trainees-mūḍha, līna, and samyakpratipanna-are obtained therein as parallel phrases of the VyY. Based on the presence of prefixes such as kun tu, *sam-, etc., in the VyY, Ueno's conjectures for the four words for trainee are saṃmūḍha, pramatta, saṃlīna, and samyakpratipanna. His conjectures are based on and also supported by the AAA, which will be listed below. Cf. Nance (2012, p. 134). Nance (ibid.,  The words saṃdarśayati, samādāpayati, samuttejayati, and saṃpraharṣayati (expound, motivate, inspire, and delight) appear frequently in the sūtras. 9 Vasubandhu interprets these four words as the purpose of the sūtras that guide the four types of trainees. However, this is not a typology of human beings. As Vasubandhu says, even the same person may fit more than one category.
VyY, D30a-31b1, P34b6-7: gang zag kha cig gcig (gcig] D; cig P) pu yin yang don kha cig la kun tu rmongs la| kha cig la zhes bya ba nas yang dag par zhugs pa'i bar du yin pas gcig la yang bzhi 'thad do|| Since some people, even if it is one and the same person, are (1) stupid with respect to some objects and [are (2) lazy] with respect to [some objects], to (4) practice [some objects] correctly, it is appropriate to apply the four [purposes] to one person. 10 Then, Vasubandhu specifically describes the four trainees who correspond to the four goals as follows: VyY, D31b1-3, P34b7-35a1: de la (1) kun tu rmongs pa rnams ni (ni] D; na P) mi shes pa dang the tshom dang log par shes pa dang ldan pa'i phyir ro|| (2) bag med pa rnams ni le lo dang 'dun pa med pa dang g-yen spyo ba dang ldan pa'i phyir ro|| (3) kun tu zhum pa rnams ni dman pa la mos pa dang bdag nyid mi nus par sems pa dang 'gyod pa dang bya ba mang pos 'jigs pa dang ldan pa'i phyir ro|| (4) yang dag par zhugs pa rnams ni phyin ci ma log pa dang mnyam pa dang chog mi shes pa'i brtson 'grus sgrub (sgrub] VyYT DP (Ueno 2017, n. 15); bsgrubs DP) pa dang ldan pa'i phyir ro|| Scholars have pointed out that the AAA sometimes quotes the VyY. 11 However, I recently found that the AAA also quotes the VyY above. This allows us to recover much of the original Sanskrit of this passage and to re-read the relevant sections of the VyY and the AAA. 12 8 Cf. Nance, ibid.: "2. For those who are careless, a sūtra provides guidelines for what is genuinely acceptable." 9 Incidentally, Allon (2022), in discussing "[t]he highly structured, carefully crafted nature" of early Buddhist texts, refers to this phrase with the Buddha as a subject and writes: "his act of teaching is expressed through four semi-synonymous verbs rather than one. Further, the component units of these structures or strings are normally arranged according to a waxing number of syllables, that is, the first unit has fewer syllables than the last (or at least their count does not decrease); for example, the syllable pattern of the above four verbs sandassesi samādapesi samuttejesi sampahaṃsesi is 4+5+5+5." In the Sanskrit equivalents, too, the syllable pattern is retained, with an increase of one syllable: 5+6+6+6. Furthermore, Allon says in fn. 7: "[t]he phenomenon of arranging such material according to syllable length is well known in many fields," and he "coined the phrase Waxing Syllable Principle (WSP)" in Allon (1997, p. 191. Cf. fn. 18). 10 The translations by Honjō and Ueno are appropriate, but Nance's understanding of the syntax is not satisfactory. Moreover, Nance's continuation of this sentence to the next sentence is also inappropriate. Nance (2012, p. 134): "Each aspect references a specific [kind of] person and a certain form of ignorance concerning a particular object. Each single kind of person also encompasses four [kinds]:" 11 Skilling (2000, pp. 331-332), etc. AAA, 289.11-16 (cf. AAA(t), D133b1-4): (1) ajñānasaṃśayamithyājñānayogān mūḍhaṃ* samyagarthakathanāt sandarśayet| (2) kauśīdyācchandikatāvyāsaṅgayogāt pramattaṃ kuśalārthaṃ pravartanāt samādāpayet| (3) hīnādhimuktyaśaktātmasambhāvanākaukṛtyabahukṛtyabhīrutāyogāt** saṃlīnaṃ viśiṣṭavīryānuśaṃsakathanāt samuttejayet| (4) aviparītasamāsaṃtuṣṭivīryapratipattiyogāt*** samyakpratipannaṃ bhūtaguṇābhinandanāt saṃpraharṣayet| *: mūḍhaṃ] Ms; mūḍhaṃ prati W[ogihara] **: -aśaktātma-] Ms; -aśaktyātma-W ***: -samāsaṃtuṣṭi-] em (cf. Tib: mnyam pa dang chog mi shes pa);samādhisaṃtuṣṭi-Ms, -samādhyasaṃtuṣṭi-W 13 Thus, the translation of the passage is as follows, with a few corrections to Honjō's and Ueno's translations: Among them, (1) "the stupid" are [known as such] because they possess ignorance, doubt, and misapprehension; (2) "the lazy" are [known as such] because they possess idleness, lethargy, and distraction; 14 (3) "the dispirited" are [known as such] because they have inferior resolve, 15 a feeling of inability, regret, 16 and fear of the many things that need to be done; (4) "those who are practicing properly" are [known as such] because they possess vigor that is without error, balanced, and insatiable. 17 After this, Vasubandhu refers to something that he calls *prayojanaprayojana, the final goal in his later work, PSVy. While the former introduced the general goal, the latter differs from one sūtra to another (Ueno 2017, p. [53]). This means that various sūtras have their own specific final goal. 13 I have emended the current edition by Wogihara (W) on the basis of the Tibetan translation. The VyY, AAA's source here, also supports this emendation. VyY: mnyam pa (*sama) dang chog mi shes pa (*asaṃtuṣṭi). Note that the conjecture atṛpta for chog mi shes pa in Ueno 2007, p. (49) must be corrected to asaṃtuṣṭi on the basis of this parallel. Further evidence is as follows: the various kinds of vigor are given in MSA,. Of them, atuṣṭivīrya in the verse portion (MSA) is asaṃtuṣṭivīrya in the prose portion (MSABh. Nagao 2009, pp. 86-87). Incidentally, the AAA manuscripts had samādhisaṃtuṣṭi, but Wogihara corrected dhi to dhya, probably because of the need for a negative clause (a-) based on the Tibetan translation (samādhy-asaṃtuṣṭi). However, according to the above discussion, it could be surmised that the scribe at some stage read the word samādhi into the original *samāsaṃtuṣṭi (=samaasaṃtuṣṭi) and added dhi to create the current form samā\dhi[saṃtuṣṭi. 14 Nance translates g-yen spyo ba as "(those who) calumniate," and Ueno translates it as dòngyáo [agitation] (*vyākṣepa). We have another source: vyāsaṅga in the AAA and rnam par g-yeng ba in the AAA(t). Indeed, Negi has an entry for g-yen spyo ba and one of the corresponding Sanskrit equivalents is vyākṣepa. On the other hand, vyāsaṅga has the meaning of distraction (MW, s.v., vyāsaṅga) and fits this context. Therefore, I assume vyāsaṅga to be the Sanskrit original of g-yen spyo ba in the VyY. 15 Honjō's and Ueno's translation of dman pa la mos pa as "the preconception (*adhimukti) that [myself is] inferior (*hīna)" is unnatural. Nance's translation "who aim for what is trifling" is appropriate. Cf. Conze 1967, s.v., hīnādhimuktika: "of inferior resolve, one who has inferior intentions." 16 Since the VyYT (D143a6) regards 'gyod pa (regret) as a separate item ('gyod pa dang ldan pa'i phyir dang|), Nance's translation "those who are regretful of their own incapacities," which connects it to the previous item, is inappropriate. 17 Nance's translation for (4) is unique: "among those who are set out [should be counted also] those whose understanding is nonerroneous, those who possess equanimity, and those who are driven, lacking contentment."

Why Did the Buddha Teach Synonyms? Vasubandhu's View of the Buddha's Words
Overview Furthermore, Vasubandhu establishes one question and answer to further discuss this "purpose." In many sūtras, four similar words-samādāpayati, vinayati, niveśayati, and pratiṣṭhāpayati ([the Buddha] incites, trains, causes to settle in, and establishes)-are found and are also related to the purpose of the sūtras. If they were synonymous, then they would be meaningless repetitions. If, on the other hand, they have different meanings, it would undesirably follow that sutras have more than four purposes, 18 contradicting the previous statement that there are four types of purpose, such as saṃdarśayati (see Section "Purpose"). Vasubandhu responds that there is no problem even if they are synonymous since the Buddha has purposefully taught them. He also provides another option: even if they are heteronymous, there is no problem since they eventually converge on the four preceding phrases, such as saṃdarśayati. 19 In the broader context of Indian philosophy, what is being discussed here is the following: In Indian philosophy, repeating the same word or meaning, especially in debate, is known as punarukta, one of the points of defeat (nigrahasthāna) during an argument (Nyāyasūtra, 5.2.14). Merely repeating the same argument in response to a challenge from the opponent is considered a defeat. On the other hand, the listing of synonyms in Buddhist scriptures is slightly different from the above, but Buddhists also seem to have regarded the useless repetition of words and meanings as problematic. 20 In response to this problem, the VyY states that there are eight purposes for which the Buddha taught the Dharma using synonyms. In his later work, the PSVy, on the other hand, Vasubandhu presents this argument in connection with the enumeration of the synonyms for avidyā, such as ajñānam, adarśanam, etc., that the Pratītyasamutpādasūtra lists.
As has been pointed out by previous studies, 21 several treatises that came after the VyY deal with this topic: Vasubandhu's own work, the PSVy, and its commentary by Guṅamati, the PSVyT, as well as Asvabhāva's 18 As the VyYT (D144a2) states, in that case there would be seven purposes (dgos pa bdun du 'gyur ro); since samādāpayati has also appeared in the previous four purposes, excluding it and adding three (i.e., vinayati, niveśayati, and pratiṣṭhāpayati) makes seven. Incidentally, although not directly relevant to the discussion in this paper, the number of syllables here is 6+4+5+6, and so Allon's WSP (cf. fn. 9) does not seem to be valid. Allon (1997, p. 191), however, states that "[t]he exception to this general definition is where sequences can or must be divided into groups according to meaning or grammatical or morphological form." In this case, since samādāpayati (6 syllables) occupies a distinct position as one of the four purposes, the above can be divided into 6, 4+5+6, and his schema of WSP is also tenable here. Incidentally, Yamaguchi ([1959]1973, p. 168). n. 1 notes that the parallel in Pali has three rather than four phrases. The schema is the same there; Aṅguttaranikāya, 2.43: samādapeti niveseti patiṭṭhāpeti (5, 4+5). 19 On this point, see Ueno (2017).
However, I discovered that the same argument is also found in almost its entirety in Haribhadra's Abhisamayālaṃkārālokā (AAA). Furthermore, I also noticed that two of the eight purposes are found in the AVSN and not only one as previously thought. These findings will allow for a more precise reading, and even for a radical rereading, of the passage in question, not only in the VyY but also in related texts. In this section, the discussion clarifies how Vasubandhu thought that the Buddha's words themselves had the function of educating and training trainees and dharma preachers.
Here, specifically, a situation is assumed in which the Bhagavat (the Exalted One) or the Buddha 22 preached with word A followed by synonyms B, C, etc. Therefore, I will use the model of A, B, and C, etc., and analyze the argument in the relevant texts.

Investigation of Each Item
The relevant section begins as follows: VyY: dgos pa dag ni brgyad (brgyad] D; brgyud P) de\|[ rnam grangs gsungs pa ni 'dul ba tha dad pa'i phyir te| There are eight purposes. The teaching of synonyms is because of the variety of trainees (*vineya). 23 The first purpose of preaching using synonyms is as follows: (1) VyY: (I) de'i tshe dang phyi ma'i tshe (II) kha cig la (la] P; la la D) las (III) khong du chud par bya ba'i phyir dang| Nance: "[1.] In order to cause [a particular person] to grasp [the teaching] in this life or in a subsequent life." Ueno: "1. In order to cause a person to grasp [the Bhagavat's statement] at that time or at a subsequent time," AVSN: (II) kasyacit kathaṃcit (III) tadarthāvabodhārthaṃ AAA: (I) tadā cāyatyāṃ ca (II) kasyacit kathaṃcid (III) arthāvabodhārtham| AAA(t): (I) de'i tshe dang phyis (II) 'ga' zhig gis 'ga' zhig rnam pa 'ga' zhig ltar (III) don rtogs par bya ba'i phyir dang| PSVy: (II) 'ga' zhig ji ltar yang (III) de'i don khong du chud par bya ba'i phyir dang| Nett-a: desanākāle āyatiñ ca kassaci kathañci tadatthapaṭibodho siyāti pariyāyavacanaṃ| I have divided this passage into three parts for a clear discussion: (I) to (III). First, since kasyacit kathaṃcit is assumed to be the original word in (II) of the VyY compared to parallel texts, VyY, P should be adopted (kha cig, la las). The translation of this would then be "someone, somehow." Second, we should add an object of understanding to (III) of the VyY. Here, since the PSVy by Vasubandhu himself has de'i don, *tad-artha-(cf. AVSN: tadartha-, AAA: artha-), this should be added. Third, regarding (I), Nance's translation "in this life or in a subsequent life" is not satisfactory. To begin with, in Sanskrit the pair iha and amutra means in this life and in the life to come (cf. Apte, s.v., amutra), but the original words from the AAA are tadā and āyati, not this pair. In this regard, it is interesting to note that Ueno 2017, n. 47 points out that "that time" would refer to the time when the Bhagavat preached the word (A) and that "later time" would refer to the time when he preached the second and subsequent synonyms (B, C, etc.). However, according to the PSVyT by Guṅamati, kasyacit kathaṃcit seems to fit the connotation found in (II). Let us consider (I) further.
In the Nett-a, instead of tadā, we find desanākāle (at the time of exposition). Moreover, the Nett-t, its commentary, annotates the word āyatin as "Āyatin ti paccavekkhaṇakāle" ("In the future" means at the time of consideration). Furthermore, since this context pertains to the purpose of preaching B, etc., "at the time of preaching" must refer to the time when B, etc., not A, are taught. If so, then this (I) would mean when B is heard and (or?) when one considers B after the preaching takes place. In any case, there is no reference to understanding in the next life. In addition, since Vasubandhu does not mention this (I) in his later work, the PSVy, he may not have emphasized this (I).
Let us then look at Guṅamati's comments on (II). Although he does not annotate this in the VyYT, he does in the PSVyT.
PSVyT: 'ga' zhig ji ltar yang de'i don khong du chud par bya ba'i phyir dang zhes bya ba ni nyan pa po 'ga' zhig ji ltar yang rnam grangs tha dad pa dag gis kyang de'i don rtogs par bya ba'i phyir te| kha cig la ni tshig gi don kha cig grags pa yin no|| "In order for somebody to somehow understand its meaning" means in order that "someone", i.e., some hearer, "somehow", i.e., even through different synonyms, can understand its meaning. That is, a certain meaning of a word is [at least] well known to a particular person.
The meaning may be as follows: A person may not understand the meaning of A but may be able to understand B or C. Hence, the Bhagavat preaches B, C, etc., which are synonyms of A. Thus, (I) and (II) can be summarized as follows: (I) tadā cāyatyāṃ ca (that time and later): When the hearer hears synonym B and/or when the hearer reflects on B after preaching. (II) kasyacit kathaṃcid (somebody, somehow): Somebody who cannot understand A may understand B, and somebody who cannot understand B may understand C, etc.
Then, the translation of the relevant passage would be as follows: Alternative: In order for somebody to somehow (*kathaṃcit) grasp [the meaning of A] at that time (when hearing B, etc.) and/or at a subsequent time (when considering the meaning of B, etc.).
(2) VyY: de'i tshe rnam par g-yengs pa rnams la brjod pa de nyid kyis (kyis] P; kyi D) ni gzhan dag gis (gis] D; gi P) smad par 'gyur bas rnam grangs kyis de'i don bstan par bya ba'i phyir dang| Nance: "[2.] In order to teach the meaning of that [teaching], via discursive strategies, to those who are distracted in this life-since others might disparage a statement as to what is in fact the case." Ueno: "2. In order to teach the [same] meaning (*artha) of that [statement] using synonyms, by [repeating] the same statement to those who are distracted at that time, since others [who are concentrating] might disparage [the Bhagavat for repeating the same statement in vain]," Nett-a: tasmiṃ khaṇe vikkhittacittānaṃ aññavihitānaṃ aññena pariyāyena tadatthāvabodhanatthaṃ pariyāyavacanaṃteneva padena punavacane tadaññesaṃ tattha adhigatatā siyāti| Since the PSVyT clearly states that "at that time" means "at the time of hearing" (nyan pa'i tshe), there is no reason to understand it as "in this life," as in the case of Nance. Additionally, his translation of the second half is not appropriate. Although Nance's understanding differs, my translation is almost identical to Honjō's and Ueno's.
Alternative: In order to teach the meaning of it (A) using synonyms (B, etc.) for those who were distracted at that time (i.e., when the Bhagavat was preaching A) since others (i.e., who could understand A) would disparage [the Bhagavat] for using the same expression (A).
On the other hand, let us examine the AVSN. Ueno pointed out the parallel between the above AVSN and the VyY. However, there are problems with the previous reading of the text due to a misreading of the manuscripts.
Concerning the word marked with an asterisk, Samtani reads avaśītaṃ(?) in the 1971 edition and translates it as "clear" in the English translation published in 2002. Honjō 1989, on the other hand, offers the correction avagataṃ (understood) in his Japanese translation, which Ueno also follows in his Japanese translation. 24 However, when I checked the manuscripts, I found avagītaṃ in a manuscript named G. Other manuscripts indeed have ś instead of g. However, g and ś are easily confused because of the similarities of their shape in Sanskrit manuscripts. The word avagītam is a noun that means "reproach, blame" (Apte). Moreover, this corresponds precisely to smad pa in the VyY.
Thus, my translation of (2) in the AVSN, including the phrases in parentheses, is as follows: (2) To let those whose minds are distracted at that time (when he heard A) hear its meaning through synonyms (B, C, etc.)-for by that same expression (A) [the Bhagavat] would be disparaged (avagīta) by others (and so on. I will not state all of the types of purpose for fear of making the text [i.e., the AVSN itself] voluminous).
The above discussion also allows for a rereading of the PSVy.
PSVy: de'i tshe rnam par g-yengs pa rnams de'i don thos par bya ba'i phyir ro|| de nyid brjod na ni gzhan dag gis dpyas* (dpyas] P; dpyad D) par 'gyur ro|| Regarding the word marked with an asterisk, Ueno (2017, p. 93) and (2021b, p. 110) adopted D dpyad and translated it as lǐjiě [understanding]. Hence, his syntactic understanding of the whole sentence differs from mine. P, on the other hand, has dpyas here. This dpyas (pa) is the past tense of the verb dpya ba, which, according to Das, has the same meaning as 'phya ba. The meaning of 'phya ba, according to Jäschke, is as follows: "to blame, censure, chide; the context, however, in which the word occurs, seems to suggest the meaning: to scoff, to deride." This perfectly matches the context and smad pa in the VyY (cf. AVSN: avagītaṃ, AAA: avagītatā, AAA(t): smad pa. On the other hand, the Nett-a has adhigatatā, but the context requires the meaning "disparagement," and so a textual problem is suspected). Thus, my translation is as follows: PSVy: In order for those who are distracted at that time (when A was uttered) to hear its meaning. For if [the Buddha] stated the same word (A), he would be disparaged (dpyas pa, *avagīta) by others.
This investigation has thus first clarified that the PSVy and AVSN faithfully carried over the argument in the VyY.
On the other hand, the AAA has been somewhat reworded and seems to envisage a somewhat different situation. 25 AAA: tenaiva cārthābhidhāne pūrvaśrutānām evāvagītatā syād iti taddoṣaparihāreṇa pūrvakālavikṣiptānāṃ* paścādāgatānāṃ ca tadarthaśravaṇārtham| *: -kāla-] AAA; -kālaṃ Ms For when [the Bhagavat] stated the meaning by that same [word], there will be contempt from those who had heard [that same word] before. Therefore, [the Bhagavat has taught synonyms] so that (i) those who were distracted previously and (ii) those who come later will hear the meaning, removing that fault [of contempt of repetition].
The key point of the above is as follows: The Bhagavat preaches with synonyms so that (i) those who were inattentive when the Bhagavat preached A and (ii) those who were not present when the Bhagavat preached A might understand the meaning of A through the use of synonyms B, C, etc., while the Bhagavat avoids the criticism of meaningless repetition. We must imagine that, in the period when these texts were written, there were no recordings as there are today.
VyY: yid mi gzhungs pa rnams la yang dang yang du de'i don yang dag par mtshon pas mi brjed par bya ba'i phyir dang| Nance: "[3.] In order to characterize the meaning over and over again, so that inattentive persons will not forget it." Ueno: "3. In order that inattentive persons will not forget [the Bhagavat's statement] by setting forth the [same] meaning over and over again," AAA: durmedhasāṃ punaḥpunas tadarthalakṣaṇārtham| AAA(t): blo zhan pa rnams la don de yang dang yang du bzlas pas rtogs par bya ba'i phyir dang| PSVy: yid rtul po dag kyang yang nas yang du de'i don rtogs pas mi brjed par bya ba'i phyir dang| Nett-a: mandabuddhīnaṃ punappunaṃ tadatthasallakkhaṇe asammosanatthaṃ pariyāyavacanaṃ| Nance and Ueno translate yid mi gzhungs pa rnams as "inattentive persons," but these are the people described in (2). Since the assumed Sanskrit here is durmedhas (cf. AAA, Mvy, no. 2899: medhāvī, yid gzhungs pa, and Nett-a: mandabuddhi), the subject here is persons who are slow in understanding, who are different from those assumed in (2). The point here is that for those who are slow to understand, the Bhagavat teaches them one meaning through all possible means, i.e., through using various synonyms.
Alternative: 3. So that dull-witted persons (*durmedhas) will not forget [the meaning of A] by designating/defining the meaning [of A by synonyms B, C, etc.] over and over again, VyY: tshig gcig la don du ma byung bas don gzhan du rtog pa bsal ba'i phyir dang| Nance: "[4.] In order to eliminate ideas that bear on alternative meanings, in those cases in which a single phrase may possess multiple meanings." Ueno: "4. In order to eliminate ideas that bear on alternative meanings, or in cases in which a single phrase may possess multiple meanings," AAA: ekaśabdānekārthatayā 'rthāntarakalpanāvyudāsārtham| AAA(t): sgra gcig don du mar gyur pas (gyur pas] D; 'gyur bas P) don gzhan du rtogs pa bsal ba'i phyir dang| PSVy: sgra gcig la don du ma yod pas don tha dad par rtogs (cf. rtog in PSVyT) pa bsal ba'i phyir dang| Nett-a: aneke pi atthā samānabyañjanā hontīti yā atthantaraparikappanā siyā tassā parivajjanattham pi pariyāyavacanaṃ| There does not seem to be any major problem here with the preceding translations of the VyY. However, the following sentence from the PSVyT makes the situation more transparent. Moreover, this understanding is also related to the understanding of the next item: PSVyT: rnam grangs gnyis pa la sogs pas ni rnam grangs dang po'i don kho na shes par byed pa'i phyir ro|| Because the very meaning of the first synonym (A) is known by the second synonym, etc. (B, C, etc.) The PSVyT says that when A is polysemous, if followed by synonyms B and C, in which they act as determiners of A, excluding other semantic choices, A's meaning is determined to be one.
Alternative: To eliminate the supposing of another meaning since one word has many meanings: VyY: gzhan du ming de rnams kyis de'i don yang dag par bsgrub pa'i phyir sGra nges par sbyor ba lta bu dang| Nance: "[5.] In order to establish the meaning of that [teaching] via alternate phrasing, so as to use just the right words." Ueno: "5. In order to rightly comprehend the meaning [of the Bhagavat's statement] with other nouns, as in the Nighaṇṭu," AAA: anyatra nirghaṇṭuvat (-ṇṭu-] Ms; -ṇṭa-AAA) tābhiḥ saṃjñābhis tadarthasampratipattyartham| AAA(t): gzhan las ming bshad pa dang 'dra bar ming de dag gis don de legs par rtogs par bya ba'i phyir dang| PSVy: ming de dag gis gsung rab las byung ba'i don de rtogs par bya ba'i phyir te| sman gyi rnam grangs kyi ming bzhin no| Nett-a: anaññassa vacane anekāhi tāhi tāhi saññāhi tesaṃ tesaṃ atthānaṃ ñāpanattham pi pariyāyavacanaṃ seyyathāpi nighaṇṭusatthe| First of all, Ueno's assumption that sGra nges par sbyor ba is Nighaṇṭu is a wise one. He further points out the parallelism in the Nett-a, which has "seyyathāpi Nighaṇṭusatthe (as in the Nighaṇṭu treatise)." We can add to this another example of parallelism, Nirghaṇṭu, in the AAA. This Nighaṇṭu is a kind of Vedic lexicon. 26 Thus, (5) must mean: A's meaning is more accurately understood when accompanied by synonyms, such as in a lexicon.
However, although earlier translations do not translate it, the VyY has gzhan du and the AAA has anyatra at the beginning. This word seems to be significant in this context. According to Apte, anyatra can be defined as: "adv. 1. elsewhere (with abl.); 2. on another occasion (in comp.); 3. except; 4. otherwise." Since the Nighaṇṭu is a non-Buddhist text, anyatra may mean "in texts other than Buddhist texts." However, this possibility is unlikely since there is no ablative (abl.) here. If this is the case, it would be natural to assume that it means "otherwise," forming a pair with the previous item (4).
To illustrate this, I will preemptively summarize the eight beneficiaries for whom the Bhagavat preaches synonyms. In our established notation, (1) is a person who cannot understand a particular word, (2) is a person who did not hear a word due to being distracted, and (3) is a person who is slow to understand. On the other hand, (6) is the preacher, (7) is the Bhagavat himself, and (8) is a future preacher. However, (4) and (5) are listeners with normal comprehension and levels of attention. Moreover, the situation envisaged in (4) is that the words are polysemous and difficult to determine in a single sense. We can regard (5) as being a similar situation.
If this is the case, then anyatra in (5) is meant to contrast the previous item (4) and (5), meaning "otherwise" or "or else." In short, we can envisage the situation as follows: in the case of a polysemous word, the meaning of the word in question is uniquely determined by listing several synonyms in (4), while in (5) the meaning of the polysemous word can be precisely (sam=samyak) understood (yang dag par bsgrub pa, *sampratipatti) through synonyms as in the case of a lexicon.
Alternatively, (5) may indicate that a synonym is used to help us to better understand a difficult word. Thus, based on the word anyatra, my understanding of the VyY above is that 4 and 5 form a pair. Alternative: Otherwise/or else (anyatra, if it is not in the situation of (4)), so [the hearer] will correctly understand its [A's] meaning by means of those terms [B, C, etc.], as in the case of the Nighaṇṭu.
However, I have to admit that unlike the VyY by the same author, the PSVy has no word *anyatra. Thus, the PSVy can be translated as "to understand its meaning that is found in the scriptures by those terms. For example, as in the terms of synonyms of medicine." As is indicated by the wavy underlines, instead of anyatra, as seen in the VyY, the PSVy has "gsung rab las byung ba'i, that which is found in the scriptures." Since the Tibetan translators of the PSVy and VyY are different and because I am of the impression that the quality of the translation of the PSVy is inferior to that of the VyY, we cannot rule out the possibility of a mistranslation here. However, this is not very likely. Nevertheless, the above discussion would still allow us to agree that (4) and (5) form a pair.
Incidentally, the Mvy mentions kuśala as the original Sanskrit for mkhas pa in several entries (nos. 798, 817, etc.). This mkhas pa nyid is thus presumably a translation of the noun form, kauśalya, which is also found in the AAA (cf. mkhas pa[r bya ba] in AAA(t)). This understanding is also appropriate for the context. However, the PSVy has the phrasing rjes su mthun bar bya ba. Negi records the correspondence rjes su mthun par byed, anulomayati, etc., which may mean "conducive to," etc. In any case, there is undoubtedly no upāyakauśalya here either.
Hence, I have removed thabs from the text of the VyY and rendered it as mkhas pa nyid, which must be a translation of kauśalya, skill. 27 I assume that over the course of the transmission of the Tibetan translation, somebody added thabs to mkhas pa nyid through association with the frequent phrase thabs la mkhas pa. Thus, my translation is as follows.
Alternative: In order to produce in the dharma preachers (*dhārmakathikas) the skill (mkhas pa nyid, *kauśalya) to describe meaning and the acquisition [of meaning].
(7) VyY: nyid la chos so so (so] P; sor D) yang dag par rig pa mnga' bar bstan pa'i phyir dang| Nance: "[7.] In order to demonstrate that one possesses the discrimination of dharma (chos so so yang dag par rig pa, *dharmapratisaṃvid)." Ueno: "7. In order to demonstrate that [the Bhagavat him]self possesses the discrimination of the Dharma (*dharmapratisaṃvid)," Alternative: In order to proclaim the special knowledge of dharma (*dharmapratisaṃvid) that [the Bhagavat] himself has. AAA: ātmano dharmapratisaṃvidudbhāvanārtham| AAA(t): bdag nyid chos so so yang dag par rig pa la mnga' brnyes par bstan pa'i phyir dang| PSVy: bdag nyid chos so so yang dag par rig pa yin par brjod par bya ba'i phyir dang| Nett-a: attano dhammaniruttipaṭisambhidāppattiyā vibhāvanatthaṃ| There are no significant problems with earlier translations. The underlying Sanskrit for bstan and brjod in the VyY and PSVy can be assumed to be udbhāvanā from the AAA and the PSVyT below. The BHSD has the entry "udbhāvana, nt.°nā, f." This suggests the meanings "(laudatory) manifestation, making them known, declaration." Furthermore, the BHSD explains the meaning of this in the compound form as follows: "guṇodbhāvanā, rarely°na, manifestation or making known, the proclamation of virtues." This seventh item is unique because it seems to be for the Bhagavat himself, not for the hearer or the preacher. What is the significance of this? Prior studies have not considered this issue, but Guṅamati's comments provide some hints. In the VyYT, Guṅamati only explains the meaning of the words in the term dharmapratisaṃvid, 27 Cf. VyY (Chapter 2), D40a7, P45b4: 'di man chad ni gzhan dag la tshig gi don bshad pa la mkhas pa nyid bskyed pa'i phyir (Hereafter, in order to produce in others a skill (mkhas pa nyid, *kauśalya/kauśala) to explain the meaning of the words ...); ASBh, 17.20: vineyānāṃ samāsavyāsanirdeśakauśalyotpādanārtham, ASBh(t), D14a2: gdul bya rnams bsdus pa dang{|} rgyas par bstan pa la mkhas pa bskyed pa'i phyir te| but in the PSVyT he develops a more detailed discussion. Let us examine the relevant passage: PSVyT: [Q[uestion]] ci'i phyir bcom ldan 'das chos 'chad pa'i tshe bdag nyid chos so so yang dag par rig par brjod par mdzad ce na\| [ [A[nswer]] gdul bya rnams kyi dgos (dgos] D; dgongs P) pa bsgrub par bya ba'i phyir te| kha cig de'i sgo nas mngon par yid ches pa skye ba'i phyir ro|| ji skad du| cho 'phrul gang mdzad seng ge'i sgra|| rang gi yon tan gang sgrogs pa|| 'dod med rnam par 'gyur bral ba'i|| thugs rje'i 'byung khungs (khungs] D; khung P) yod pas yin|| (ṛddhir yā siṃhanādā ye svaguṇodbhāvanāś ca yāḥ| vāntecchopavicārasya kāruṇyanikaṣaḥ sa te|| Śatapañcāśatka ([Śpś], 63) zhes gsungs so|| Question: Why does the Bhagavat proclaim that he himself possesses the special knowledge of dharma when he preaches? Answer: It is in order to complete the trainees' purpose (dgos pa, *prayojana). Because some people develop trust [in the Bhagavat] through it. As is said [in the Ś pś]: "The magic, the Lion's Roars, the displays of your own qualities, these were the whetstone-rubbings from Pity's gold in you who had spewed out the activity of desire." It could be a verse from Mātṙcet˙a's Śpś that Guṅamati is quoting here (the translation of the verse is Bailey's [1951, p. 163]). The gist of the verse is as follows: just as gold is marked when struck against a touchstone, the performance of supernatural power, the manifestation of one's virtues, etc., are nothing more than traces of the Bhagavat's revelation of compassion. There is nothing personal there.
This citation elucidates the meaning of (7). The proclamation of the Bhagavat's possession of the special knowledge of dharma (*dharmapratisaṃvid) is due to compassion or, more precisely, to cause the hearer to trust him. Thus, this (7) is also understood as ultimately being for the benefit of the listeners. The importance of producing trust or faith in the Bhagavat when one hears the dharma will be touched upon again in Section "The structure of the VyY in relation to the methodology of sūtra interpretation".
VyY: gzhan dag la de'i sa bon bskyed pa'i phyir te| Nance: "[8.] In order to arouse the seeds of that [discrimination of dharma] in others." Ueno: "8. In order to arouse those seeds [for the discrimination of the Dharma] in others." AAA: pareṣāṃ ca tadbījādhānārtham AAA(t): gzhan rnams la de'i sa bon bskyed pa'i phyir te| PSVy: gzhan dag la de'i sa bon gzhag par bya ba'i phyir ro|| Nett-a: veneyyānaṃ tattha bījāvāpanatthaṃ vā pariyāyavacanaṃ bhagavā niddisati| The last portion is also about the hearer as a subject. There are no problems with earlier translations. However, regarding the referent of the word "it (tad-)," Guṅamati notes that it refers to special knowledge of dharma in the VyYT. But in the PSVyT he offers two interpretations of tad: the seed of a synonym and the seed of special knowledge of dharma. In the context, both would be possible. Additionally, in both cases this (8) can mean benefitting the hearer when s/he becomes a preacher in the future.
VyYT: nyan pa po gzhan dag la rnam grangs gsungs pa'i sa bon bskyed pa'i phyir ro zhes bya ba'i tha tshig go|| (Meaning (*ity arthaḥ): to place the seeds of the synonyms that are taught to other hearers.) PSVyT: nyan pa po rnams la rnam grangs kyi tshig gi sa bon gzhag (gzhag] D; bzhag P) pa'i phyir zhes bya ba'i tha tshig dang chos so so yang dag par rig pa'i sa bon gzhag (gzhag] D; bzhag P) pa'i phyir yang yin no|| (Meaning (*ity arthaḥ): to place the seeds of the synonyms in other hearers. It also means to place the seeds of a special knowledge of dharma [in them].) Alternative: In order to place the seeds of it (i.e., synonyms and/or special knowledge of dharma) in others.

Conclusion of this Section
The above discussion clarifies what was unclear in previous translations. First, the beneficiaries of the Bhagavat's preaching with synonyms can be classified as follows: (1)-(5) are the hearers; (6) is the preacher; (7) is the Bhagavat himself (however, see below); and (8) are hearers who will become preachers in the future.
At first glance, (7) did not seem to be consistent with the other items since the Bhagavat himself is the beneficiary. However, by referring to the PSVyT, we confirmed that the hearer is also the beneficiary here. Furthermore, by focusing on the words gzhan du and *anyatra in the VyY, we also showed that (4) and (5) form a pair.
If we further elaborate on the model using (word) A, (word) B, (word) C, etc., the eight purposes can be systematically understood as follows.
(1) Those who cannot understand by means of A can understand synonyms B, C, etc.; (2) those who were distracted and did not hear when the Bhagavat preached A can understand through synonyms such as B and C; (3) those who are slow to understand will not forget the meaning of A through the use of synonyms such as B and C; (4) to fix the meaning of a polysemic word A in one sense using B, C, etc.; (5) to clarify the meaning of A by enumerating synonyms B, C, etc., as is done in lexicons; (6) to benefit the preacher's preaching by enriching their vocabulary; (7) to show that the Bhagavat himself possesses a special knowledge of dharma (*dharmapratisaṃvid) (and by doing so, inducing the hearers to respect the Bhagavat and listen to his teachings attentively); (8) so that when the hearers become dharma preachers, they will be able to preach with synonymous words and/ or obtain special knowledge of dharma in the future.
Thus, one aspect of Vasubandhu's view of the Bhagavat's words has become clear: not a single word of the Bhagavat's was spoken in vain, and he preached with synonymous terms to benefit his hearers, who were expected to be in all kinds of situations.

The Structure of the VyY in Relation to the Methodology of Sūtra Interpretation
In this section, I would like to take up Vasubandhu's explanations of the organic connections between the five aspects of sūtra interpretation presented in Section "Exegesis based on five aspects" and the structure of the chapters of the VyY.

The Organic Relationship Between the Five Aspects of Sūtra Interpretation
Vasubandhu explains the organic relationship between the five aspects of sūtra interpretation in two (similar) ways. The following is the second interpretation.
Rigs pa dang snga phyi in (5) are the two components of *artha-codya (objections concerning the meaning). This is one of the two types of codya, objections, in the VyY.
don la brgal ba yang rnam pa gnyis te| snga phyi 'gal bar brgal ba dang| rigs pa dang 'gal bar brgal ba'o|| Objections concerning meaning are also of two kinds. Namely, objections [concerning the relation with statements] before and after* and objections concerning reasoning. (Cf. Hanner, 2020, fn. 21) *: Here "before and after" is not temporal. This is an objection concerning the [seeming] contradiction between what the Bhagavat has said in one sūtra and another sūtra.
Hanner (2020)'s understanding of the underlined section differs from mine. 28 (1) is known from (2), up to (4) is known from (5). This is what Vasubandhu is saying in the above. Furthermore, the intention of the above passage is to say that (1) through (5) must be explained organically.

The Structure of the VyY
Let us now provide an overview of the structure of the VyY. The correspondence between the first four chapters of the VyY and the five aspects of sūtra interpretation is as follows (the locations in D have been added in parentheses): The above simplifies the correspondence between the five aspects and their respective chapters (cf. VyY(L)), but the reality is more complex. Let me give one example. VyY Chapter 4 deals with the proof of the authenticity of the Mahāyāna. However, this is a response to the final objection in Chapter 3. Moreover, the objection itself is derived from Chapter 2. More specifically, Chapter 2 takes up 103 sūtra passages and explains the meaning of words (padārtha) therein. The last sūtra passage is about the twelvefold branches (dvādaśāṅga) of the Buddha's word, which is one of the ways of classifying the Buddha's word. Among them, Vasubandhu interprets the branch vaipulya to mean Mahāyāna. 29 In VyY Chapter 3, on the other hand, an opponent, presumably a non-Mahāyānist or Ś rāvakayāna, takes up this interpretation and raises an objection (codya) known as a "contradiction to scripture, *sūtravirodha," 30 and Vasubandhu responds (parihāra) to it throughout the entirety of Chapter 4. 28 With regard to the "objections and replies (*codya-parihāra)" in the VyY(VY), the fundamental topic of Hanner (2020), he (p. 142) writes as follows and cites the following text in the footnote (fn.).
"In the first appearance of the term in the VY, Vasubandhu explains that objections and replies are "non-contradiction with respect to reasoning (rigs pa; nyāya or yukti) and with respect to coherence (snga phyi, literally 'previous and next')."" fn. 18: VY 31a1, p. 7: brgal ba dang lan brjod pa las ni rigs pa dang snga phyi mi 'gal ba'o|| However, the above is not the "first appearance" of objections and replies since there is already a description of it in D30b5: rigs pa dang mi 'gal ba dang| snga phyi mi 'gal ba ni brgal ba dang lan las yin pas, which comes before the above (D31a1). Also, since there is ablative las, the above phrases cannot be connected by the copula "are" as Hanner does. If we applied Hanner's logic, it would follow that (1)=(2)= (3)=(4)=(5); however, this is not what Vasubandhu is saying here. 29 VyY, D82b, P97a8: shin tu rgyas pa'i sde ni theg pa chen po yin te| 30 VyY, D88a5, P103b2: shin tu rgyas pa'i sde ni theg pa chen po yin no zhes gang bshad pa yang lung (lung] D; rung P) dang 'gal ba yin no|| From this understanding of the structure of the VyY, Hanner's (2020, p. 138) translation of the last part of Chapter 4 is untenable: VyY, D114a6, P133a3: de bas na theg pa chen po sangs rgyas kyi gsung yin no zhes bya ba de ni mi 'gal lo|| de'i phyir shin tu rgyas pa'i sde theg pa chen po yin no zhes bya ba 'gal ba med do|| Hanner: "therefore, the claim that the Mahāyāna is the word of the Buddha is not contradictory. For that reason, the claim that the vaipulya [sūtras of the Mahāyāna] are the Mahāyāna is without contradiction." If we remove the brackets, we can understand the strangeness of the last sentence. " [T]he claim that the sūtras of the Mahāyāna are the Mahāyāna" is, of course, not a contradiction, but it is a strange statement and even seems to be a tautology. The background to this sentence is as follows: vaipulya is one of the twelve branches of the Buddha's word, and while the non-Mahāyāna or Ś rāvakayāna sects understand them to be non-Mahāyāna scriptures, the Mahāyāna side says that the Mahāyāna scriptures are the equivalent of this vaipulya. Additionally, since the Mahāyāna is the Buddha's word, there is no problem in understanding that the vaipulya is equivalent to the Mahāyāna. Hence, the last sentence can be translated as follows: Alternative: For that reason, the claim that the vaipulya (one branch of the Buddha's word that non-Mahāyānists or Ś rāvakayāna claim to be sūtras transmitted in the non-Mahāyānist or Ś rāvakayāna tradition) is the Mahāyāna [which I (i.e., Vasubandhu) have discussed in Chapter 2] is without contradiction (in spite of the opponent's objection in Chapter 3).
The Position of VyY Chapter 5 in Relation to the "Purpose" The Relation Between "Purpose" and Hearing the Buddha's Word Respectfully The above fourth chapter completes the explanation of the method of sūtra interpretation, which consists of five aspects. Chapter 5, the final chapter of the VyY, discusses a different topic: respectfully listening to the Buddha's word. What is the position of this chapter in relation to the previous four chapters? Recently, Hanner (2020) translated a section on this subject. However, in view of his understanding of the Tibetan and the structure of the VyY, his understanding seems odd. 31 Therefore, in elucidating Chapter 5, I will critically examine his understanding. For the sake of the discussion, I will divide this opening part of Chapter 5 into three sections, (I), (II), and (III), even though they are consecutive sections.
VyY, D114a7ff., P133a4ff.: (I) brgal ba dang lan gyi rnam pa yang bstan| ji ltar mdo sde bshad par bya ba'i rigs (rigs] D; rig P) pa yang bstan to|| gzhan yang chos smra ba pos thog ma kho nar mdo sde bkod nas brgal zhing brtag par bya ste| 'khor lan rnams la sred pa bskyed pa'i phyir ro|| 'on kyang sred pa chung ba yin pa rnams la yang rna blag tu gzhug pa'i phyir gus par mnyan pa dang ldan pa bshad par bya'o|| [Q] gus par mnyan pa dang ldan pa 'di ci zhig ce na| [A] gtam gang las gus par (par] D; pas P) mnyan pa bstan pa'o|| rna blags na dgos pa la sogs pa'i rim pas mdo sde bshad par bya'o|| [I, i.e., Vasubandhu] have also explained the aspect objection and answer (*codyaparihāra, the fifth of the five aspects of sūtra interpretation). [I have] also explained the method (*yukti) of how one should annotate the sūtras. Furthermore, the preacher should quote (bkod) the sūtras first and then ply with questions (brgal zhing brtag, *paryanu-√yuj It should be noted that the last section states that a story (*kathā) must be preached based on listening to it respectfully before the purpose (prayojana), the first of the five aspects of sūtra interpretation. The text continues: (II) [Q] brgal zhing brtag (brtag] D; brtags P) pa byas na yang brgal ba ci'i phyir smos she na| [A] brgal ba'i mjug thogs (mjug thogs] D; 'jug thog P) su lan rnams bde bar khong du chud par bya ba'i phyir ro|| [Q] sngar| mdo sde'i che ba nyid thos na|| mnyan pa dang ni gzung (gzung] D; bzung P) ba la|| nyan pa po ni gus byed 'gyur|| de phyir thog mar dgos (dgos] D; dgongs P) pa brjod|| Hanner (2020, p. 136) takes up this passage (which is actually separated at the halfway point) and translates it as follows: "If it is asked: What is the purpose of an objection if one has [raised] objections and scrutinized [the matter]? [It is] because the replies are easier to understand after an objection. If one has previously heard of the greatness of the sūtras, the listener will act respectfully towards what he hears and remembers. Therefore, first express the purpose [for the teachings]." The word "previously" in " [i]f one has previously heard" translates the word sngar. However, first, the word sngar stands outside of the four-line stanza, in which each line consists of seven syllables, so one cannot incorporate it inside the verse. In the first place, this verse is the same one that appears in Chapter 1, as introduced in Section "Exegesis based on five aspects" of this article. Thus, it should be clear that here "before" refers to this verse in Chapter 1 located "before" Chapter 5. It is also inappropriate to separate the sentences as Hanner does since they continue below (III): the above is in the middle of a question from the interlocutor, as is evident from the phrase zhe na, *iti cet, which concludes the range of a question. Vasubandhu clarifies this chapter's position after zhe na. Thus, this "before, sngar" does not pertain to thos, hear, but to bstan, taught, which immediately follows that verse. 32 The rest of the passage is as follows: (III) ces bstan pa yin na| de ci'i phyir thog mar gus par mnyan pa dang ldan pa bshad par bya zhe na| [A] dgos pa bstan pa la yang gus par mnyan par bya ba'i phyir ro|| kha cig mdo sde'i don khong du chud par mi nus pa dag 'byung bas de dag la chos tsam la yang gus par mnyan pas bsod nams bskyed par bya ba'i phyir dang| don khong du chud pa la smon (smon] P; smon D) pas shes rab kyi khams bskyed par bya ba'i phyir| gdon mi za bar thog mar gus par (par] D; pa P) mnyan pa dang ldan pa bshad par bya'o|| A natural translation of (II) and (III) based on the above discussion, on the other hand, must be as follows: (II) Question: If [the preacher] has plied with questions (*paryanu-√yuj), why is the objection (*codya) raised further? Answer: To make it easy to understand the answers immediately after the objection. Question: Although you have previously (sngar) taught (bstan): "If one hears of the greatness of the sūtras (*sūtrasya māhātmya), the hearers will strive to hear and grasp with reverence. Therefore, the purpose should be explained first," (III) nevertheless, why should one first explain [the story] on respectful hearing [before the "purpose"]? Answer: It is in order for the explanation of the purpose to also be heard [by hearers] with respect. Some do not understand the meaning or content of the sūtras. Therefore, that they may, at the very least, hear the dharma respectfully, thereby giving rise to merit (*puṇya). Also, they may desire to understand the meaning [of the sūtras], thereby giving rise to the element of wisdom (*prajñā), and so the [story] about respectful hearing should necessarily (gdon mi za bar) be given at the outset (thog mar).
If we were to clarify the meaning of the above passages in light of the overall structure of the VyY, it would be as follows: as a whole, the VyY presents a method of sūtra interpretation that consists of five aspects, beginning with the purpose, prayojana. However, in Chapter 5, Vasubandhu says that prior to it, one should quote the scriptures and ply with questions (which may mean asking the audience what this 32 In connection with the Sanskrit parallel and the context, I would like to note one more point: Hanner's translation "(act respectfully towards) what he hears and remembers" is also inappropriate. This phrase should be translated as "hearing and grasping" since what is being said here is that one will respectfully hear and grasp if one hears the greatness of the sūtra; the Sanskrit is śravaṇa-udgrahaṇa, as we have seen in Section "Exegesis based on five aspects". scripture teaches), 33 thereby creating a thirst for answers in the audience. However, since some people have little thirst for answers, he says that a preacher should present the story about hearing the Buddha's word with respect. Should this story only be relayed to those with little thirst for answers? Judging from the immediately following statement, "in order for the explanation of the purpose to also be heard [by hearers] with respect," it is probably Vasubandhu's intention that preachers provide stories before presenting the purpose, etc., for all hearers. If this is the case, we can illustrate the process of preaching as follows: Citing (presenting) a sūtra [ plying with questions [ creating a thirst for answers (in hearers) [ providing stories about hearing with respect [ preparing the hearers to hear respectfully [ interpreting scripture with the five aspects, beginning with the purpose (as described in Chapter 1).

An Example of a Story in VyY Chapter 5
Chapter 5 of the VyY is thus full of such stories to arouse faith in the Buddha in the audience. For a sampling of such stories, including humorous ones, see Ueno (2021a). To illustrate the characteristics of this chapter, I will present the simile of a vessel, which first occurs in the Aṅguttaranikāya 3.3. (I. 130.5ff.) and is explained in many treatises (cf. Skilling, 2000, p. 301). Although Skilling properly understood the gist of this simile, Nance's (2012, p. 121) understanding differs: VyY, D116a4-5, P135a3-5: (de bzhin du skye bo nyan pa po'i yid kyi) snod gsum dag tu chos smra bas chos kyi char phab kyang| chos kyi chu'i bya ba mi byed de| (1) rnam par g-yeng ba dang| rmugs pa dang| gnyid dag gis mi nyan pa'i phyir| gang du mi 'bab pa dang| (2) tshul bzhin yid la mi byed pa'i phyir| gang du bab kyang skyon can du 'gyur ba dang| (3) dran pa brjed ngas pa'i phyir gang du mi gnas pa'o|| Nance: "in the three worlds, though the dharma preacher may cause the rain of the dharma to fall, there may be no water of dharma. It may fall on deaf ears, due to distraction, stupidity, or sleep. Though falling, it may become polluted, since one may fail to attend to it in the right way. And [though it may fall and be pure], it may not remain, due to forgetfulness." 34 Although Nance separates the sentence at the halfway point and cites it from snod gsum dag tu, this is inappropriate since a description in parentheses precedes this. Moreover, we should understand this paragraph in association with the preceding paragraph, which describes the three kinds of deficient vessels. As Skilling properly understood, the point here is that the hearer is likened to a vessel (snod, *bhājana, not the "world," as Nance understands it). If a vessel is left outside when it rains, then the rain water can be used as water. However, if the vessel has three kinds of faults-(1) it is overturned, (2) dirty, or (3) leaky-even if rain falls on the vessel, it cannot be later used as water. In the same way, even if a preacher preaches, namely, a preacher sends down the rain of the dharma, if the hearers are experiencing these faults, then the preaching will be in vain. Of course, the point being made here is that one should hear the Buddha's words respectfully. Thus, a translation of this passage could read as follows: (Similarly,) even if the preacher makes the dharma rain on the three vessels (the minds of the hearers), it will not do the work of the dharma water. (The following compares minds similar to faulty vessels:) (1) [The hearers ' minds] in which the rain [of the dharma] does not fall because of non-hearing due to distraction, depression, or dozing off; (2) [the hearers' minds] in which the rain [of the dharma] falls but [the water] is polluted because of wrong comprehension; and (3) [the hearers' minds] in which the rain [of the dharma] is not stored because of forgetfulness.
In this way, the main point of this fifth chapter is to prepare hearers to hear the dharma through such didactic stories. In other words, this chapter acts as a manual for vivid preaching.
Thus, this section reaffirms the position of Chapter 5 within the framework of sūtra interpretation in the VyY, especially in relation to the "purpose," and introduces an example of this story with regard to hearing the Buddha's words respectfully.

Conclusions
In this paper, we have reviewed the overall structure of the VyY and have mainly discussed the "purpose, prayojana" through the use of some new material. I will not reiterate the conclusions obtained through the investigations in each section, but the following summarizes what has been clarified in this paper.
In Chapter 1 of the VyY, Vasubandhu presents a method of sūtra interpretation that consists of five aspects, with the purpose being placed at the beginning. This paper identified for the first time two parallel passages from the AAA regarding the purpose of a sūtra and the purpose of synonyms and scrutinized the passages in question by comparing various sources. In addition, the paper also reread related materials, using this as a starting point. In particular, this paper first clarified the systematic structure of the argument about the Buddha's preaching through the use of synonyms. According to Vasubandhu, the Buddha envisaged all possible hearers, and even future preachers, as beneficiaries and preached in synonymous terms so that everyone would benefit.
On the other hand, Chapter 5 indicates that stories about listening to the Buddha's teachings respectfully must be taught first rather than the purpose. Thus, Chapter 5 contains many examples of such stories for preachers of the dharma. In Section "The position of VyY Chapter 5 in relation to the "purpose"", I clarified the preaching procedures taught in the first part of Chapter 5. In sum, the VyY is a guide to methods of sūtra interpretation, and it is also a manual for lively preaching.
As mentioned in the introduction, only a Tibetan translation of the VyY is extant. Because of the difficulties involved in reading the text, at times we have engaged in rather detailed philological discussions. However, it is necessary to thoroughly examine each passage of this text so as to accurately read it in the original context of the Indian author. Only then will the arguments of the VyY become clear. This paper has examined only a few passages as a start, but I believe that it will provide an essential foundation for future work.