Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa1
Robert Kritzer
I. Introduction
The terms Dārṣṭāntika and Sautrāntika have a long history of being used to
designate a group (or groups) of thinkers who were associated with the Sarvāstivādin school but who differed from the mainstream regarding a number of
important doctrinal issues. In the earliest relevant source, the *Vibhāṣā, a
dissident group is referred to as Dārṣṭāntika, but beginning with the Abhidharmakośabhāṣya, the term Sautrāntika appears. In fact, Vasubandhu uses both terms:
Sautrāntika when he approves of a position, and Dārṣṭāntika when he disapproves. Some scholars conclude that Dārṣṭāntika and Sautrāntika refer to the
same group and that the term Dārṣṭāntika is used negatively while Sautrāntika is
used with approbation. Others, notably Harada and myself, believe that Vasubandhu uses the term Sautrāntika to describe his own opinions, which, I believe, are
actually based largely on the Yogācārabhūmi, even though many of them resemble
and may have originated in positions attributed to Dārṣṭāntika by the *Vibhāṣā or
to the Sthavira (Śrīlāta) by the *Nyāyānusāra or found in Harivarmanʾs *Tattvasiddhi. However, in the centuries after Vasubandhu, the terms come to be used
more and more synonymously: for example, Yaśomitra, in his seventh-century2
commentary, the Abhidharmakośavyākhyā, famously glosses Dārṣṭāntika as
Sautrāntika.3
It has generally been assumed that the author of the Abhidharmadīpa,4 the
earliest commentary on the Abhidharmakośabhāṣya extant in Sanskrit, uses the
1
A shorter version of this paper was given at the XIVth Conference of the International Association
of Buddhist Studies, London, 2005. I would like to thank Professors Elizabeth Kenney and Yamabe
Nobuyoshi for reading and commenting upon earlier drafts and Mitomo Kenyō for access to
portions of a draft of his Japanese translation of the Abhidharmadīpa.
2
For this date, see Mejor (38-41).
3
dārṣṭāṃtikāḥ sautrāṃtikāḥ (Abhidharmakośavyākhyā: 392.21). dārṣṭāṃtikāḥ sautrāṃtikaviśeṣā ity arthaḥ
(Abhidharmakośavyākhyā: 400.17).
4
Although there is no translation of the Abhidharmadīpa into a western language, Mitomo Kenyō
will soon publish a Japanese translation. In this paper, I cite an unpublished draft of this translation
as “Mitomo.”
112
Robert Kritzer
terms Dārṣṭāntika and Sautrāntika interchangeably. Jaini says as much in his
introduction to the text (Introduction 70), and his opinion is echoed by Hirose
(126) and Yoshimoto (90-91). More recently, this assertion has been repeated by
Honjō (326), Yamabe (239 n. 47), and myself (“General Introduction” 211).
Until now, only Hirose has dealt with this subject in any detail. However, his
understanding of Sautrāntika is based heavily on later sources, and at the time of
his article, the relationship between Vasubandhuʾs Sautrāntika positions and the
Yogācārabhūmi had not been thoroughly explored. Nor does Hirose look at every
occurrence of the two terms in the Abhidharmadīpa.
In this paper, I examine each use of the terms Dārṣṭāntika and Sautrāntika in
the text. Although the results are not conclusive, I think that the words are not,
in fact, used interchangeably and that the author of the text, the Dīpakāra,
preserves at least some of Vasubandhuʾs original distinction between Dārṣṭāntika
and Sautrāntika.
I have suggested that Vasubandhu uses the term Sautrāntika when he accepts
an idea found in the Yogācārabhūmi that goes against orthodox Sarvāstivāda.
These ideas may or may not have antecedents in the Dārṣṭāntika sources mentioned above (Kritzer, “Sautrāntika”). Vasubandhu uses the term Dārṣṭāntika, on
the other hand, to characterize ideas from these sources with which he
disagrees.5
The Dīpakāra nowhere mentions the Yogācārabhūmi, and, like the other commentators on the Abhidharmakośabhāṣya, he generally does not criticize it for
covertly propounding Yogācāra ideas.6 But if my assumptions about the meanings of Dārṣṭāntika and Sautrāntika for Vasubandhu are correct, we might at least
expect the Dīpakāra to use Sautrāntika only for Vasubandhu’s own opinions. As
for Dārṣṭāntika, we would expect the Dīpakāra to use it to describe positions that
have antecedents in the earlier sources but are rejected by Vasubandhu.
Furthermore, the Dīpakāra could also use it in the case of opinions similar to
those accepted by Vasubandhu in order to criticize Vasubandhuʾs sources or
teachers rather than to attack Vasubandhu personally.
Of the Dārṣṭāntika positions mentioned in the Abhidharmadīpa, two concern
perception, one, action and result, and four, the question of the reality of past
and future. In the remaining case, the Dīpakāra quotes Vasubandhu’s statement
approving the Sautrāntika theory that anuśayas are seeds of kleśas but substitutes Dārṣṭāntika for Sautrāntika.
Of the Sautrāntika positions, one relates the theory of momentariness to the
question of culpability in murder. The other three are about the theory of seeds.
5
In each case, the Yogācārabhūmi contains in different places both Vasubandhu’s position and the
position that he rejects (Kritzer, Vasubandhu xxviii-xxix n. 39).
6
However, Jaini points out that the Dīpakāra alludes to the notion of trisvabhāva, a well-known
Yogācāra doctrine (Introduction 128).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
113
In the case of six of the eight positions attributed by the Dīpakāra to
Dārṣṭāntika, there is no confusion between the two terms: there is evidence that
these positions were actually associated with Dārṣṭāntika. Two of the references
to Dārṣṭāntika, however, may indicate some conflation of Sautrāntika and
Dārṣṭāntika. On the other hand, all four positions associated with Sautrāntika can
actually be traced to Vasubandhuʾs Sautrāntika opinions.
There is one more problematic attribution in the Abhidharmadīpa: in a discussion of mental misconduct, the Dīpakāra states that the Sthitibhāgīyas classify
greed, etc., as karma, not kleśa. Here, the Dīpakāra mistakenly assumes that Vasubandhu agrees with a position that is identified as Dārṣṭāntika in the Abhidharmakośabhāṣya. However, despite Jainiʾs attempt to associate the term Sthitibhāgīya
with the Sautrāntika conception of the saṃtati (see Section IV), we really do not
know what it signifies.
Thus, when the Dīpakāra uses the word Sautrāntika, he is, in fact, referring to
ideas that Vasubandhu himself has identified as Sautrāntika, ideas that are also
found in the Yogācārabhūmi. On the other hand, the Dīpakāra uses the word
Dārṣṭāntika more broadly than Vasubandhu does. However, if we set aside
Dārṣṭāntika position II.D, which may represent a problem with the text, the
Dīpakāra does not use the word Dārṣṭāntika to describe any of the theories that
are exclusively attributed to Sautrāntika and not to Dārṣṭāntika, namely bīja and
saṃtatipariṇāmaviśeṣa (see Kritzer, “Sautrāntika” 204-207).
II. Dārṣṭāntika in the Abhidharmadīpa
A. C ons ci ou snes s see s. 7
The first position attributed to Dārṣṭāntika occurs in a discussion of cognition
in the Vṛtti introducing verse 44, where the Dīpakāra states that, if consciousness
is said to see, then there will be ten dharmas with the nature of dṛṣṭi.8 On the
other hand, if it is denied that consciousness sees, the Dārṣṭāntika position will be
abandoned.
The question of what sees is discussed as early as the *Pañcavastukavibhāṣāśāstra (T. 1555: 991b20-c13), but the treatment most frequently referred to
is found in the *Vibhāṣā, in which five opinions are identified. Of these, four are
attributed to various schools or teachers and are rejected in favor of the position
that it is the two eyes that see rūpa.9 The rejected positions are attributed as
follows: 1) according to Dharmatrāta, it is eye-consciousness that sees rūpa; 2)
according to Ghoṣaka, prajñā related to eye-consciousness; 3) according to
7
atha vijñānaṃ paśyaty atha na paśyati / yadi paśyati daśadharmā dṛṣṭisvabhāvā bhavanti /atha na paśyati
dārṣṭāntikapakṣas tarhy ujjhito bhavati (Abhidharmadīpa: 31.1-3).
8
Presumably there should only be nine such dharmas. According to Mitomo (personal communication), however, no such list of dharmas is to be found in the abhidharma literature.
9
For a very complete discussion of this controversy, see Harada 1997: 21-33).
114
Robert Kritzer
Dārṣṭāntika, the coming together (sāmagrī) of certain factors;10 4) according to
Vātsīputrīya, a single eye. (T. 1545: 61c7-21; partially translated in La Vallée
Poussin, L’Abhidharmakośa 1: 82 n. 1).11
The Abhidharmakośabhāṣya (30.11-31.11; La Vallée Poussin, L’Abhidharmakośa 1:
82-85) contains a long discussion on the question of what sees, but Vasubandhu
does not follow the *Vibhāṣā in mentioning five opinions. Instead, he presents in
detail the argument between the Vaibhāṣikas and those who believe that
consciousness sees. Saṃghabhadra comments at great length on the argument
presented by Vasubandhu but does not attribute the belief that consciousness
sees to a particular person or group (T. 1562: 364a23ff). The Abhidharmakośavyākhyā (80.17) uses the term vijñānavādin here, but simply to designate the
person who holds the view that consciousness sees; the name does not refer to
the proponents of the idealist Mahāyāna school, for which Yaśomitra always uses
the word yogācāra.12 Like Vasubandhu and Saṃghabhadra, Yaśomitra mentions
neither Dharmatrāta or Dārṣṭāntika in connection with this opinion. The two
most frequently cited Chinese commentators, Pʾu-kuang and Fa-pao, use the term
vijñānavādin (Chinese shih chien chia 識見家) in the same way as Yaśomitra, and
they both identify Dharmatrāta as the vijñānavādin. P’u-kuang adds that vijñānavādin also refers to the Mahāsāṃghikas, etc.,13 but Fa-pao does not mention any
schools (T. 1822: 507b12-508a5).
The Dīpakāra after introducing the topic, lists four alternatives: 1) the eye
sees; 2) consciousness sees; 3) prajñā sees; 4) sāmagrī sees. At this point, the
Dīpakāra voices objections to all four positions, but he does not attribute them to
anyone. However, as we have seen, he has already attributed the second position
to Dārṣṭāntika, thus differing from the attribution in the *Vibhāṣā. In verse 44, he
states his own Vaibhāṣika opinion: the eye sees, but consciousness knows
(Abhidharmadīpa: 32.1-2; see Jaini, Introduction 75).
Thus, only the Dīpakāra attributes the position that consciousness sees to
Dārṣṭāntika, while the tradition of the *Vibhāṣā, which P’u-kuang and Fa-pao
10
Katō, relying on the *Pañcavastukavibhāṣāśāstra (T. 1555: 991b20-c13), shows that these factors are
consciousness and the dharmas associated with consciousness and that the idea that the sāmagrī
sees is actually a continuation of the idea that prajñā sees: it is not prajñā alone but the
conglomeration of all the other caittas as well, in addition to cakṣurvijñāna, that sees rūpa (23-24).
11
The discussion in the earlier translation of the *Vibhāṣā is arranged somewhat differently. First,
the positions of Dharmatrāta and the Vātsīputrīyas are refuted (T. 1546: 50c12-51b23). Then four
alternatives are listed (the eye sees, eye-consciousness sees, prajñā associated with eyeconsciousness sees, the coming together of factors sees) together with objections. The position that
the eye sees is accepted, but it is acknowledged that consciousness also must be present (T. 1546:
51b24-29). Here, the four positions are not attributed to any person or group.
12
Stcherbatsky refers to “an idealist opponent,” probably on the basis of Yaśomitra’s use of the
term vijñānavādin, but he gives us no reason to take this characterization seriously (61).
13
識見家難。即是法救。大衆部等 (T. 1821: 49c1-2).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
115
follow, attributes it to Dharmatrāta.14 Perhaps an explanation lies in the fact that
this Dharmatrāta is closely associated with Dārṣṭāntika (see Lin 316-320, Miyamoto 125-180, Yamada 416-438). Among the various personages of the same
name, there is a Dharmatrāta who frequently espouses Dārṣṭāntika-like positions.
Miyamoto thinks that the discussion regarding what sees is one of the unusual
cases in which Dharmatrāta in the *Vibhāṣā does not maintain a Dārṣṭāntika
position, and he points to the fact that both the *Vibhāṣā and P’u-kuang (T. 1821:
48c3-8) associate Dharmatrāta with a different position from that of the
Dārṣṭāntikas (171-172). On the other hand, Yamada, who distinguishes five
different types of Dārṣṭāntika positions in the *Vibhāṣā, thinks that the position
attributed to Dharmatrāta is an example of a Dārṣṭāntika position that resembles
Mahāsāṃghika (431). If this is the case (and P’u-kuang’s mention of Mahāsāṃghika may support Yamada), then it is possible that the positions attributed
respectively to Dharmatrāta and to Dārṣṭāntika are examples of two different
types of Dārṣṭāntika positions. Furthermore, Harivarman, who very frequently
maintains positions attributed to Dārṣṭāntika in the *Vibhāṣā, agrees with the
notion that consciousness sees. This suggests that this opinion is, in fact, somehow associated with Dārṣṭāntika,15 even though the *Vibhāṣā here attributes a
different opinion to Dārṣṭāntika.
Vasubandhu, after presenting the disagreement between the Vaibhāṣikas and
their opponents, states in the name of Sautrāntika that the whole argument
regarding what sees is nonsensical. In fact there is nothing that sees or is seen;
consciousness simply arises in dependence on the sense-organ and the object
(Kritzer, “Sautrāntika” 333-334). Saṃghabhadra attacks Vasubandhuʾs statement,
but he identifies it with Dārṣṭāntika rather than Sautrāntika (T. 1562: 367b24368a11; see Cox, “On the Possibility” 38-39).16 There are several possible explanations for this attribution. One is that Śrīlāta (the “Sthavira” of the *Nyāyānusāra
and a famous Dārṣṭāntika teacher) held the same opinion, and Saṃghabhadra
was aware of this fact. Another possibility is that Saṃghabhadra knew that the
*Vibhāṣā attributed the opinion that sāmagrī sees to Dārṣṭāntika, and assumed
that sāmagrī referred there, as in the Abhidharmakośabhāṣya, to the coming
together of sense-organ, object, and consciousness, not to the conglomeration of
consciousness, prajñā, and the other caittas (see note 9).
14
或有執眼識見色。如尊者法救 (T.1545: 61c8-9).
15
問曰。諸根爲到塵故知。爲不到能知。答曰。非根能知。所以者何。善根能知塵。則可一時遍知諸塵 。
而實不能。是故以識能知 (T. 1646: 267a7-9). Although Harivarman uses the word “know” rather than
“see,” it is clear from the context that he is discussing the same question. See especially T. 1646:
267a22-b1. The Kokuyaku Issaikyō translator states that Harivarman takes the position of the
“Vijñānavādin” here (Ronjū-bu 3: 132 n. 27).
16
Surprisingly, Saṃghabhadra does not attribute this statement to the “sūtra master,” the term
that he frequently uses to describe Vasubandhu when Vasubandhu is deviating from Sarvāstivāda.
116
Robert Kritzer
As for the Dīpakāra, he devotes the most attention to the idea that the sāmagrī
sees. Like Saṃghabhadra, he understands sāmagrī here as referring to the coming
together of sense-organ, object, and consciousness and refutes this position at
length (Abhidharmadīpa: 31.11-28). At no point in this refutation does the
Dīpakāra mention Dārṣṭāntika, Sautrāntika, or the Kośakāra (an expression he
uses often when criticizing Vasubandhu’s unorthodox positions), but it seems as
though his criticisms are directed against Vasubandhuʾs Sautrāntika ideas, which
the Dīpakāra equates with the position that sāmagrī sees. In the course of his
refutation, the Dīpakāra attacks an explanation of causation in terms of bīja,
which is a further indication that his target is Vasubandhu.
Later, at the end of his comment on Abhidharmadīpa verse I.44, the Dīpakāra
quotes Vasubandhuʾs statement that the disagreement regarding what sees is
meaningless. Here, the Dīpakāra refers to Vasubandhu as the Kośakāra. Then the
Dīpakāra, implying that Vasubandhu has defected to Mahāyāna, reproaches him
for “not only showing his ignorance of the Abhidharma but also of heading for
the precipice of ayoga-śūnyatā ”(Jaini, Introduction 77).
In summary, the Dīpakāra explicitly identifies Dharmatrātaʾs opinion in the
*Vibhāṣā as a Dārṣṭāntika position. As for the position attributed to Dārṣṭāntika in
the *Vibhāṣā (but not in the Abhidharmadīpa), he understands sāmagrī as the
coming together of all causes and conditions, and he suggests that this position is
the one favored by Vasubandhu. In this case, there is no evidence that the
Dīpakāra uses the terms Dārṣṭāntika and Sautrāntika interchangeably. In fact, he
uses Dārṣṭāntika to describe an opinion that is not accepted by Vasubandhu and
that is not identified as Sautrāntika in the Abhidharmakośabhāṣya.
B. There i s n o prat yak ṣa be cau se the ob je ct s of con s ci ous nes s ( an d
the sen se org an s) are alread y in the p as t at the m omen t of consciousne ss . 17
This statement attributed to Dārṣṭāntika is found in the Dīpakāraʾs discussion
of Abhidharmakośa II 2, where Vasubandhu says that the ādhipatya of the indriyas
is with regard to the perception of the specific object of each indriya. The
discussion in the Abhidharmakośabhāṣya is very complicated. In verse II 1,
Vasubandhu presents the orthodox Vaibhāṣika explanation of the power
17
dārṣṭāntikasya hi sarvam apratyakṣam / pañcānāṃ vijñānakāyānām atītaviṣayatvād yadā khalu
cakṣūrūpe vidyete tadā vijñānam asat / yadā vijñānaṃ sat cakṣūrūpe tadāsatī vijñānakṣaṇasthityabhāve
svārthopalabdhyanupapatteś ca (Abhidharmadīpa: 47.13-482; Jaini, Introduction 79; Katō 313). This is
translated by Cox: “For the Dārṣṭāntikas, nothing is directly perceived. This is due to the fact that
the five groups of perceptual consciousness have past object-fields; indeed, when the eye and visual
material form are found, perceptual consciousness does not exist, and when perceptual
consciousness exists, the eye and visual material form do not exist. Further, this is due to the fact
that the apprehension of their own object is impossible given the absence of the continuation [of
the object] in the moment of perceptual consciousness” (“On the Possibility” 78 n. 56).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
117
(ādhipatya) of the various indriyas. According to the Vaibhāṣikas, the five material
sense faculties have a variety of powers: the beautification of the body, the
protection of the body, the production of the corresponding consciousness and
the mental dharmas associated with it, and their own special activities, namely
seeing rūpa, etc. (Abhidharmakośabhāṣya: 38.4-10; La Vallée Poussin, L’Abhidharmakośa 1: 102-103). However, Vasubandhu mentions the opinion of “others,” according to whom the sense faculties do not have such a wide range of powers (Abhidharmakośabhāṣya: 39.3-5; La Vallée Poussin, L’Abhidharmakośa 1: 107). In the
following three verses (Abhidharmakośa II 2-4),18 Vasubandhu departs from his
usual procedure of giving the Vaibhāṣika position in the verse, and in II 2, he
states that each of the five material sense faculties only has power with regard to
its corresponding consciousness (Abhidharmakośabhāṣya: 39.6-11; La Vallée Poussin, L’Abhidharmakośa 1: 107-8).19
The Dīpakāra characterizes this as the position of the “Kośakāra, etc.,” and
stresses that Vasubandhu, who is supposedly giving his own opinion in the group
of verses II 2-4, is actually appropriating a Vaibhāṣika position here. The Dīpakāra
then mentions the Dārṣṭāntika position that pratyakṣa is impossible. There is no
implication that Vasubandhu himself is sympathetic to Dārṣṭāntika here; in fact,
the juxtaposition of the Kośakāraʾs view and the Dārṣṭāntikaʾs suggest that they
are incompatible.
As is clear from the passage in the Abhidharmadīpa, the reason that pratyakṣa is
impossible for the Dārṣṭāntikas is that they deny the simultaneous existence of
cause and effect. In other words, the material sense faculty and object must
precede the consciousness that results (Cox, “On the Possibility” 40-41). Katō
shows the connection among the passage in the Abhidharmadīpa, Harivarmanʾs
denial of the possibility of the simultaneous existence of cause and effect (T.
1646: 258c5-6; 276b6-7), and the Sthavira’s denial of sahajātahetu in the context of
cognition (T. 1562: 421b18-22, c16-17). Katō also states that the argument
concerning simultaneous cause and effect is not found in abhidharma texts before
Harivarman (313).
The question remains as to whether Vasubandhu agrees with this Dārṣṭāntika
reasoning. Vasubandhu discusses the question of whether sahabhūhetu exists on
at least two occasions, once in an exposition of the hetus (Abhidharmakośabhāṣya:
18
Saṃghabhadra omits these three verses in the *Abhidharmapiṭakaprakaraṇaśāsanaśāstra (T. 1563),
but he includes them in the *Nyāyānusāra (T, 1562).
19
The discussion here is confusing. Vasubandhu does not say whose opinion verses II 2–4 represent,
but it seems to be the opinion of the “others” whom Vasubandhu mentions as criticizing the
Vaibhāṣika position stated in verse II 1. Yaśomitra suggests these others are Sautrāntikas (Abhidharmakośavyākhyā: 96.16) while P’u-kuang suggests “Vijñānavādins” (T. 1821: 57b6; see item II.A in
this paper). Nishi, the Kokuyaku Issaikyō translator, follows P’u-kuang (Kokuyaku Issaikyō Bidon-bu
25: 118 n. 50).
118
Robert Kritzer
84.18-85.6; La Vallée Poussin, L’Abhidharmakośa 1: 253-255) and once in a
definition of vedanā in the context of pratītyasamutpāda (Abhidharmakośabhāṣya:
145.3-146.21; La Vallée Poussin, L’Abhidharmakośa 2: 101-107). In both cases, P’ukuang (T. 1821: 117b14, 176b10) and Fa-pao (T. 1822: 560c3, 607c12) attribute the
position that sahabhūhetu does not exist to the Sautrāntikas, but they do not state
that Vasubandhu agrees. In neither case does Yaśomitra attribute the position
that sahabhūhetu does not exist to a particular school. 20 Finally, in both
discussions, Vasubandhu lets the Sarvāstivādins have the last word. 21 This
suggests, although it certainly does not prove, that Vasubandhu accepts the
existence of sahabhūhetu.22
Nowhere else have I been able to find the position, asserted, if we can believe
the Dīpakāra, by Dārṣṭāntika, that there is no pratyakṣa. This is not surprising:
according to Potter, “all schools of Indian philosophy take perception as a
legitimate pramāṇa” (58). When the Dīpakāra attributes this to Dārṣṭāntika, what
he is really doing is criticizing the idea that cause and effect can never be
simultaneous. This idea is associated with Dārṣṭāntika, through Harivarman and
Śrīlāta, while with the exception of P’u-kuang and Fa-pao, none of the commentators characterizes it as Sautrāntika. It seems likely that the Dīpakāra intentionally uses the term Dārṣṭāntika here to refer to figures other than Vasubandhu.
C. There i s n o vip āka be si des sukh ā v ed an ā; th ere is n o kar m a bes ide s
volit i on. 23
20
However, the opinion of “others” regarding the temporal relationship between sparśa and vedanā,
mentioned by Vasubandhu in his discussion of vedanā, is attributed by Yaśomitra
(Abhidharmakośavyākhyā: 307.16) and Sthiramati (Tattvārthat: tho 74b3) to Śrīlābha or Śrīlāta and by
P’u-kuang to the Sthavira, i.e., Śrīlāta (T. 1821: 176c10-11).
21
eṣām api nāmānye ‘pi sabhāgahetutvādayo bhavantīti siddhaḥ sahabhūhetuḥ (Abhidharmakośabhāṣya:
85.6; La Vallée Poussin, L’Abhidharmakośa 1: 255). āyuruṣmaṇoḥ sāhabhāvye saṃspṛṣṭavacanāt siddhaḥ
kṣaṇaniyamaḥ / yac cāpi sūtre uktaṃ trayāṇāṃ saṃnipātaḥ sparśaḥ iti / tat kathaṃ vijñānaṃ cāsti na ca
trayāṇāṃ saṃnipāto na ca sparśa iti / tasmād avaśyaṃ sarvatra vijñāne sparśaḥ sparśasahajā ca
vedanaiṣṭavyā / alam ati prasaṅginyā kathayā / prakṛtam evānukramyatām (Abhidharmakośabhāṣya:
146.18-21; La Vallée Poussin, L’Abhidharmakośa 2: 107; see also Abhidharmakośavyākhyā, in which
Yaśomitra indicates that this passage represents Vasubandhu’s opinion [309.31]).
22
In his discussion of the controversy on the reality of sahabhūhetu in Chapter Two of the
Abhidharmakośabhāṣya, Dhammajoti assumes that the argument against it is that of Sautrāntika,
probably on the basis of the Chinese commentaries (9-10). However, he notes that Yogācāra “finds
it an indispensable doctrinal category in the their theory of bīja.” He refers to a discussion in the
Mahāyānasaṃgraha, as well as to a passage in the Yogācārabhūmi, which includes sahabhūhetu as one
of five hetus (Dhammajoti 22 n. 5). The Yogācārabhūmi equates sahabhūhetu with parigrahahetu
(Yogācārabhūmi: 111.9-11; Yogācārabhūmit: dzi 67a5-7; T. 1579: 302b1-2), one of the ten hetus of the
Yogācara list. This passage is quoted verbatim in the Hsien-yang sheng chiao lun (T. 1602: 572a7-9),
which elsewhere states that all of the six hetus (i.e., the six on the Sarvāstivāda list), including
sahabhūhetu, are “inappropriate,” except for vipākahetu (T. 1602: 570a12-23).
23
dārṣṭāntikānāṃ tu sukhaiva vedanā vipākaḥ cetanaiva ca karma (Abhidharmadīpa: 139.9-10).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
119
Before mentioning this Dārṣṭāntika opinion, the Dīpakāra asks rhetorically if
vipāka consists only of vedanā and answers that, although vedanā is predominant,
it is accompanied by the other four skandhas. After presenting the Dārṣṭāntika
opinion, the Dīpakāra goes on to say that, for the Ābhidhārmikas, the five
skandhas are both vipākahetu and vipāka.
This argument is found in the *Vibhāṣā, where the Dārṣṭāntika and Vaibhāṣika
opinions are essentially the same as they are portrayed in the Abhidharmadīpa. In
the Vibhāṣā, however, the order of the statement attributed to Dārṣṭāntika is
reversed: there is no vipākahetu except for cetanā; there is no vipākaphala except
for vedanā.24
Concerning vedanā, Vasubandhu here does not attribute to any particular
group the idea that vipāka consists only of vedanā, and he agrees with
Sarvāstivāda, saying that it is not only vedanā that is vipāka, but also its
accompaniments. 25 Yaśomitra explains that “accompaniments” refers to the
sense faculties, objects, and the āśraya, i.e., nāmarūpa, which is tantamount to the
five skandhas.26
Concerning cetanā, the thrust of Vasubandhu’s Sautrāntika argument
regarding vijñapti in the Abhidharmakośabhāṣya (195.15-19; La Vallée Poussin,
L’Abhidharmakośa 3: 12) is that karma is essentially cetanā (Katō 77). In the passage
mentioned above, an identical position is attributed to Dārṣṭāntika in the
*Vibhāṣā: vipākahetu is nothing other than karma. On another occasion in the
*Vibhāṣā, the statement that physical, vocal, and mental karma are all cetanā is
attributed to Dārṣṭāntika.27 However, this idea is also implied in the Yogācārabhūmi (Kritzer, Vasubandhu 184-185), and Vasubandhu agrees with it. Therefore, I
believe, he calls it Sautrāntika rather than Dārṣṭāntika.
Here, the Dīpakāra is primarily concerned with vedanā: it is the Dārṣṭāntika
position on vedanā with which he is arguing. He is not criticizing Vasubandhu’s
understanding of karma as volition, and so his characterization of the two-part
24
云何異熟因。乃至廣説。問何故作此論。答為止他宗顯正理故。謂或有執離思無異熟因。離受無異熟果。
如譬喩者。為止彼執顯異熟因及異熟果倶通五蘊 (T. 1545: 96a24-27; see also T. 1546: 79c22-26). Mizuno
(“Hiyushi” 292-293 ) refers to two passages in the *Tattvasiddhiśāstra that he thinks express ideas
similar to those of the Dārṣṭāntikas in the *Vibhāṣā: 問曰。是諸業亦得色報。何故但説受耶。答曰。於
諸報中受為最勝。受是實報色等為具 (T. 1646: 298a24-26); 答曰。是三種行皆但是心。所以者何。離心無
思無身口業 (T. 1646: 296b1-3). The second of these passages actually goes even further and reduces
cetanā to citta. Katō points out that the idea that there are no caittas but only citta is also attributed
by the *Vibhāṣā to the Dārṣṭāntikas and Buddhadeva. Katō stresses the contrast with Śrīlāta, who
accepts an abbreviated list of caittas (199; 222 n. 4).
25
na caiṣāṃ vedanaiva phalam / kiṃ tarhi / sasaṃbhārā (Abhidharmakośabhāṣya: 228.11-12; La Vallée
Poussin, L’Abhidharmakośa 3: 109-110).
26
saṃbhriyate utpādyate’neneti saṃbhāraḥ. indriyaviṣayāśrayalakṣaṇaḥ (Abhidharmakośavyākhyā: 390.15).
Funahashi (260 n. 7) points out that Sthiramati omits āśraya in his comment (Tattvārthat: tho 173b5).
27
又譬喩者説。身語意業皆是一思 (T. 1545: 587a7-8; see Katō 71).
120
Robert Kritzer
statement as Dārṣṭāntika does not reflect a confusion between Sautrāntika and
Dārṣṭāntika. He is merely quoting, albeit in reversed order, a statement from the
Vibhāṣa that attributes both positions, on vedanā and cetanā, to Dārṣṭāntika.
D. The K oś akār a s ays th at the Dār ṣ ṭānti k a t heory th at an u śa ya s ar e
the see ds of k leśas i s g ood . 28
The original passage from which the Dīpakāra quotes mentions Sautrāntika,
not Dārṣṭāntika.29 Honjō adduces this passage in the Abhidharmadīpa as evidence
that the Dīpakāra, like other Indian authors besides Vasubandhu, makes no
distinction between the terms Sautrāntika and Dārṣṭāntika (325-326). Before I
comment on Honjō’s conclusion, a summary of the available related materials
may be useful.
The *Vibhāṣā attributes to the Vibhajyavādins the theory that anuśayas are the
seeds of the active kleśas (paryavasthāna). Unlike Vasubandhu, who considers
them to be neither associated (cittasaṃprayukta) nor disassociated (cittaviprayukta) with consciousness, the Vibhajyavādins say that anuśayas are disassociated, while active kleśas are associated.30
The Vibhāṣā does not attribute to the Dārṣṭāntikas a statement defining
anuśaya as bīja. However, Mizuno cites two separate passages that he says
mention Dārṣṭāntika with reference to this subject. According to him, one of
these shows that Dārṣṭāntika believed that anuśayas are latent,31 while the other
uses the phrase, “seeds of kleśas” to refer to kleśas in latent form (“Shinshiki ron”
325, 327). The second passage, in fact, does not mention Dārṣṭāntika or any other
group: it is simply an explanation of sūtra.32
28
evaṃ tu sādhu yathā dārṣṭāntikānām iti kośakāraḥ / kathaṃ ca dārṣṭāntikānām / kāmarāgasyānuśayaḥ
kāmarāgānuśayaḥ (Abhidharmadīpa: 222.3-4).
29
evaṃ tu sādhu yathā sautrāntikānām / kathaṃ ca sautrāntikānām / kāmarāgasyānuśayaḥ
kāmarāgānuśaya iti (Abhidharmakośabhāṣya: 278.17-18; La Vallée Poussin, L’Abhidharmakośa 4: 6).
30
分別論者又説隨眠是纒種子. 隨眠自性心不相應. 諸纒自性與心相應. 從隨眠生 (T. 1545: 313a1-3; see
also T. 1546: 236a3-5). Interestingly, according to the commentary on the Kathāvatthu, the
Mahāsaṃghikas and the Sammitīyas say that anuśayas are cittavippayutta (Jaini, “The Sautrāntika
Theory” 241).
31
或復有執。隨眠不於所縁隨増。亦不於相應法有隨増義。如譬喩者彼作是説。若隨眠於所縁隨増者於他
界地及無漏法亦應隨増。是所縁故如自界地。若於相應法有隨増義者。則應未斷已斷一切時隨増。相應畢
竟不相離故猶如自性 (T. 1545: 110a22-27). The meaning of this passage is unclear to me. Ichimura
summarizes it: "(1) If contaminants related (sic) to supporting objects, then pure factors could be
objects of the growth of contaminants, and (2) if contaminants grow with regard to their connected
factors, these factors will always grow, since connected factors never become dissociated" (528). See
also T. 1546, where the same opinion is attributed to Dharmatrāta: 尊者曇摩 多羅作如是. 諸使 不相應
使. 亦不縁使. 彼作是. 若作縁使相應使無漏法. 若作相應使. 一切時恒使無不與心心數相應時. 以是事故 (T.
1546: 89a22-25).
32
如契經. 二因二縁留諸有情久住生死. 謂煩惱業。由煩惱業為種子故生死難斷難破難滅 (T. 1545:
244b22-25; see also T. 1546: 189c27-29, T. 1547: 425b26-27).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
121
Harivarman also says that anuśayas are latent kleśas but states that they are
cittasaṃprayukta and does not use the term bīja.33 In an extensive discussion,
Fukuda traces the development of the theory of anuśaya from Harivarman to
Śrīlāta to Vasubandhu. Yamabe, on the other hand, thinks that Fukuda may overstate the similarity between Harivarmanʾs concept of anuśaya and Vasubandhu’s
identification of anuśaya and bīja. Yamabe concludes: “Perhaps Dārṣṭāntikas were
not completely ignorant of the theory of latent anuśaya, but this point is not at all
certain at this stage” (234).
Saṃghabhadra attributes the explanation of anuśaya as bīja to the sūtramaster, i.e., Vasubandhu (T. 1562: 596c24), but he does not mention either
Sautrāntika or Dārṣṭāntika. On several occasions in his discussion, Saṃghabhadra
mentions the Sthavira, who, it seems, has ideas about anuśayas that are very
similar to Vasubandhu’s.34
Yaśomitra does not comment on or quote the phrase evaṃ tu sādhu yathā
sautrāntikāṇām, but earlier in his discussion he states that the Sautrāntika opinion
is that anuśayas are bījas.35 Sthiramati uses the word Sautrāntika in both his
quotation of and comment on the phrase (Tattvārtha: tho 232a2-3), as does
Pūrṇavardhana (Lakṣaṇānusāriṇī: ñu 111a4-5). Similarly, the Chinese commentators Pʾu-kuang (T. 1821: 292b20-28) and Fa-pao (T. 1822: 688a-29) mention
Sautrāntika in both their quotation and discussion of Vasubandhuʾs statement.
To summarize, the idea that anuśayas are the seeds of kleśas cannot be clearly
linked to Dārṣṭāntika in the Vibhāṣā or the Tattvasiddhi. Nor, except for the
Abhidharmadīpa, do any of the commentaries on the Abhidharmakośabhāṣya
attribute it to Dārṣṭāntika. This would seem to support the assertion mentioned
above that the Dīpakāra uses Dārṣṭāntika and Sautrāntika interchangeably, an
assertion that Honjō supports specifically by referring to the Dīpakāraʾs substitution of Dārṣṭāntika here (326).
However, the Dīpakāra’s use of the term Dārṣṭāntika here is problematic
because it occurs in his quotation of Vasbandhuʾs statement, not in his discussion. As I see it, there are three possible explanations. The first is that the Dīpakārā intentionally altered Vasubandhuʾs text. But if the Dīpakāra in fact made no
distinction between Sautrāntika and Dārṣṭāntika, it is difficult to understand why
he would change Vasubandhuʾs words when quoting them. Jaini has noted the
large number of passages in the Abhidharmadīpa that are directly borrowed from
the Abhidharmakośabhāṣya, and he gives the impression that they are generally
faithfully reproduced. He mentions that the passage under discussion “may
specially be noted as it is identical with its corresponding Bhashya” (Introduction
5). In fact, if we set aside the substititution of Dārṣṭāntika for Sautrāntika,
33
*Tattvasiddhiśastra: 故不應言諸使非心相應 (T. 1646: 258c7-8).
In particular, see T. 1562: 597a29-b2, 597b27-c14.
35
sautrāṃtika-nāyena bījam (Abhidharmakośavyākhyā: 442.29).
34
122
Robert Kritzer
Abhidharmadīpa 222.3-223.5 otherwise corresponds almost word for word with
Abhidharmakośabhāṣya 278.17-279.4, except for the omission of one short
sentence. 36 Thus, it does not seem to be the Dīpakāraʾs practice to change
Vasubandhu’s words, and there seems to be no reason for him to do so here.
The second possibility is that the Dīpakāra was quoting accurately from an
inaccurate manuscript of the Abhidharmakośabhāṣya. But as we have seen, every
other extant version of, or quotation from, the Abhidharmakośabhāṣya reads
Sautrāntika, so this seems unlikely.
The final possibility is that the text of the Abhidharmadīpa itself is corrupt.
With the exception of a few fragments discovered in Turfan (Willemen, Dessein,
and Cox 250), the manuscript on which Jaini based his edition is the only version
available, and so there is no way to confirm its readings. Jaini remarks that the
manuscript is generally correct grammatically (Introduction 135), and it seems as
though most of the quotations are fairly accurate. However, this manuscript is
fairly late, and since a general tendency to conflate Dārṣṭāntika and Sautrāntika
seems to have developed over time, it is possible that the scribe, who may have
been distracted at the time of writing this particular line, absentmindedly
substituted the term here. Of course, this is speculation that cannot be supported
unless a new manuscript comes to light.
E. For Vibh ajy av ād a and Dārṣ ṭānti k a, a p orti on, namel y the pre sen t
time , e xis ts . 37
The Abhidharmadīpa is the only text that attributes precisely this position to
Vibhajyavāda and Dārṣṭāntika. However, several other sources mention the idea
that only a portion of the three times exists.
According to Vasubandhu, the name Vibhajyavāda refers to those who
distinguish between that which exists, namely the present time and past karma
that has not yielded its result, and that which does not exist, namely the future
time and past karma that has already yielded its result.38 Vasubandhu does not
mention Dārṣṭāntika here. Saṃghabhadra likewise attributes essentially the same
36
This omission is noted by Jaini (Abhidharmadīpa: 223 n. 1). However, it seems as though Jaini has
inadvertently omitted another phrase: in the last line of page 222, Jaini reads ... kleśotpādakaśaktiḥ
yathā cāṅkurādīnāṃ ... The Abhidharmakośabhāṣya, reads: ... kleśotpādakaśaktiḥ yathānubhavajñānajā
smṛtyutpādana śaktir yathā cāṅkurādīnāṃ ... (278.21.). Although I have no access to the manuscript, I
believe that an inspection would reveal the inclusion of the phrase yathānubhavajñānajā
smṛtyutpādana śaktir. Jaini also fails to indicate the beginning of line 3 of folio 104a of the
manuscript.
37
vibhajyavādinas tu dārṣṭāntikasya ca pradeśo vartamānādhvasaṃjñakaḥ (Abhidharmadīpa: 257.4-5).
38
ye tu kecid asti yat pratyutpannam adattaphalaṃ cātītaṃ karma kiṃcin nāsti yad dattaphalam atītam
anāgataṃ ceti vibhajya vadanti te vibhajyavādinaḥ (Abhidharmakośabhāṣya: 296.4-6; La Vallée Poussin,
L’Abhidharmakośa 4: 52).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
123
position to Vibhajyavāda without mentioning Dārṣṭāntika.39 Harivarman (258c1011) agrees with the Kathāvatthu commentary (Aung and Rhys Davids 101) in
attributing this position to the Kāśyapīyas. It must be noted that the Abhidharmadīpa, unlike these other sources, does not mention unretributed past karma.
The *Vibhāṣā, which not infrequently attributes identical opinions to
Vibhajyavāda and Dārṣṭāntika, does not ascribe the position that a portion exists
to either school. However, it contains at least one passage that suggests that the
Dārṣṭāntikas deny the reality of past and future. They are quoted as saying that
time can be divided into two sets of two: already arisen and not yet arisen;
already extinguished and not yet extinguished; therefore, there is no moment of
arising and no moment of extinction.40 This implies that the past, to which the
moment of arising belongs, and the future, to which the moment of extinction
belongs, do not exist (Kokuyaku Issaikyō bidon-bu 8: 95 n. 64; Ichimura 531).
Elsewhere in the *Vibhāṣā, the names Vibhajyavāda and Dārṣṭāntika are
associated with a position related to the nature of past, present, and future.41 The
main discussion of the three times42 begins with a presentation of the idea,
attributed to these two groups, that time is permanent while the saṃskāras are
impermanent: the saṃskāras move from one time to another, like fruit being
moved from one container to another or people moving from one hut to
another.43 The *Vibhāṣā objects, saying that time and the saṃskāras are the same,
i.e., the future time is none other than the future saṃskāras, etc. (According to
Miyashita, this statement is intended as a proof of the reality of the three times:
unlike fruit or people, the saṃskāras do not come and go anywhere.) Therefore,
the way to distinguish among the times is in terms of activity in accordance with
Vasumitraʾs explanation of sarvāstivāda (12).
It is fairly clear that the Dārṣṭāntika in the *Vibhāṣā, Harivarman (T. 1646:
255c8-256a17), the Sthavira in the Nyāyānusāra (Katō 288-293), and Vasubandhu
(Abhidharmakośabhāṣya: 295.1-301.16; La Vallée Poussin, L’Abhidharmakośa 4: 49-65;
Stcherbatsky: 76-91) all agree that past and future do not exist. However, as Katō
points out, Vasubandhu and the Sthavira give different reasons for their
39
分別論者唯説有現. 及過去世未與果業 (T. 1562: 630c10-11).
40
或復有執. 無正生時及正滅時如譬喩者. 彼説時分但有二種. 一者已生. 二者未生. 復有二種. 一者已滅.
二者未滅 (T. 1545: 141b2-5).
41
42
43
In the older translation (T. 1546: 293c18-26), however, there is no mention of Vibhajyavāda.
T. 1545: 393a9-396b23; La Vallée Poussin, “Documents” 8-25.
復有三法. 謂過去未來現在法. 問何故作此論. 答為止他宗顯正理故. 謂或有執. 世與行異. 如者分別論師.
彼作是. 世體是常行體無常. 行行世時如器中果. 從此器出轉入彼器. 亦如多人從此舍出轉入彼舍. 諸行亦
爾. 從未來世入現在世. 從現在世入過去世. 為止彼意顯世與行體無差別. 謂世即行行即是世. 故大種蘊作
如是. 世名何法. 謂此 語所顯諸行 (T. 1545: 393a9-17; see 700a25-b2 for a very similar passage). Kimura
(290) notes a resemblance between this position and a position refuted in the Kathāvatthu: addhā
parinipphanno ti (511; Aung and Rhys Davids 295). As Kimura points out, Aung and Rhys Davids
suggest that this might be an Andhaka position (295 n. 6).
124
Robert Kritzer
positions (1989: 295).
As for the Abhidharmadīpa, its characterization of the position that only the
present exists would apply to all of the versions of that position expressed in the
texts referred to above. Since the Abhidharmadīpa does not mention Vasubandhu’s specific arguments here (see Kritzer, Vasubandhu 294-335 for examples),
the Dīpakāra’s use of the term Dārṣṭāntika does not seem to be a direct criticism
of Vasubandhu. In this passage, four broad opinions are listed regarding the
reality of the three times, and each is matched with its proponents: Sarvāstivāda
maintains that all three times are real; the Vibhajyavādins and Dārṣṭāntikas only
accept the present; the Ayogaśūnyatāvādins of Vaitulika (i.e., the Mādhyamikas
and the Mahāyāna Yogācāras),44 deny the real existence of anything; the Avyākṛtavastuvādins of Paudgalika, presumably in addition to accepting past and
future, also assert that there is a real pudgala.45 In the context of such a stereotyped scheme, the term Dārṣṭāntika may well indicate the Dārṣṭāntika of the
*Vibhāṣā, together with later teachers such as Harivarman and Śrīlāta who rejected Sarvāstivāda. As for Vasubandhu, the Dīpakāra elsewhere accuses him of approaching ayogaśūnyatā.46 Therefore, the Dīpakāra may be implying here that
Vasubandhu’s Sautrāntika arguments against the reality of past and future place
him among the Ayogaśūnyatāvādins, even further beyond the pale than the
Dārṣṭāntikas.
F. The mi sgui ded view of t he Dār ṣṭ ān ti k a th at on ly the presen t t ime
exi st s is e qu ate d wi th th at of the m ateri ali st L okāyati k a. 47
See item II .E .
G. A ccor ding t o the Dārṣ ṭ ānti k a, the me t aphori cal n oti on of an
[actu al ly] n on-e xi sten t agent i s attr ibute d to the p ower s of t he
cause . 48
44
See Jaini, Introduction 123-124.
tatra sarvāstivādasyādhvatrayam asti sadhruvatrayam [corrected from sa dhruvatrayam; see Nasu 96 n.
17] iti / vibhajyavādinas tu dārṣṭāntikasya ca pradeśo vartamānādhvasaṃjñakaḥ / vaitulikasyāyogaśūnyatāvādinaḥ [corrected from vaitulikasya ayogaśūnyatāvādinaḥ; see Nasu 96 n. 18] sarvam nāstīti /
paudgalikasyāpi avyākṛtavastuvādinaḥ pudgalo ’pi dravyato ’stīti (Abhidharmadīpa: 257.4-258.2).
46
abhidharmasaṃmohāṅkasthānenātmāpy aṅkito bhavaty ayogaśūnyatāprapātābhimukhyatvaṃ pradarśitam iti (Abhidharmadīpa: 33.9-10). The context is the Dīpakāra’s refutation of Vasubandhu’s own
opinion regarding the question of what sees (item II.A, above).
47
tad anye vādino dārṣṭāntikavaitulikapaudgalikāḥ na yuktyāgamābhidhāyinaḥ tarkābhimāninas te /
mithyāvāditvād ete lokāyatikavaināśikanagnāṭapakṣe prakṣeptavyāḥ (Abhidharmadīpa: 258.6-259.1).
48
dārṣṭāntikaḥ khalu brūte kāraṇaśaktiṣu nirātmakajanikartrupacāraḥ pravartate (Abhidharmadīpa:
274.26-27).
45
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
125
This passage, like items I I. E and I I. F, is found in the discussion of the three
times. According to the Dīpakāra, an agent, i.e., a dharma, enters into the various
states of being.49 Thus, the same dharma exists in past, present, and future (Jaini,
Introduction 122). For the Dārṣṭāntika, there is no enduring agent since dharmas
have only a momentary existence, but due to causes, dharmas will arise and
perish.
This statement is not attributed to the Dārṣṭāntikas by the *Vibhāṣā. However,
we can infer from the *Vibhāṣā’s argument that the opponent of sarvāstivāda has a
similar opinion: those who reject the reality of past and future do not deny that
causes have their results, even though either the cause or the result necessarily
belongs to a time other than the present.50 In other words, a cause that is
nonexistent because it is in the past can have a real result in the present.
Similar reasoning can be found in the denial of past and future in the
*Tattvasiddhi, which confirms the efficacy of the past, although no longer existent, in producing results in the present. In the third of his ten arguments, Harivarman states that things without a cause cannot exist, and therefore past
dharmas cannot exist. 51 The proponent of sarvāstivāda answers that if past
dharmas do not exist, then there can be no results; for example, rice could not
arise from rice.52 Furthermore, we know that past karma produces result in the
future.53 To this, Harivarman responds that a cause only yields its result after it
has perished, as in the case of a seed, which perishes before the sprout appears. 54
Similarly, past karma, even though it no longer exists, produces real results.55
The same idea is also found in Harivarmanʾs refutation of the Kāśyapīya
position that a part of the past, namely karma that has not been retributed (see
item I I.E ), exists. According to Harivarman, anything past is, by definition, gone.
But even though past karma is no longer existent, it acts as a cause for a result
49
tadyathā asti vipariṇamate vardhate kṣīyate vinaśyatīti sati mukhyasattāviṣṭe kartari ete pañca
bhāvavikārā bhavanti (Abhidharmadīpa: 273.16-17).
50
又應詰彼撥無過去未來體者。若有異熟因在現在世時。彼所得果當言在何世。過去耶未來耶現在耶。若
言在過去應説有過去。若言在未來應説有未來。若言在現在應説異熟因果同時 (T. 1545: 393a25-29; La
Vallée Poussin, “Documents” 9).
51
復次若無因而有。是事不然。過去法無因可有。是故不然 (T. 1646: 255a21-22; for the entire discussion of the three times in the *Tattvasiddhi, see Katsura 40-41).
52
又現見從智生智。以修習故。如從稻生稻。是故應有過去。若無過去果則無因 (T. 1646: 255b17-19).
53
又知過去業有未來果。是名正見。又佛十力知去來諸業。又佛自説。若無過去所作罪業。是人終不墮諸
惡道 (T. 1646: 255b2-25).
54
又汝言智生智者。因與果作因縁已滅。如種與牙作因已滅。佛亦説是事生故是事生 (T. 1646: 255c12-
14).
55
業力亦爾。佛知是業雖滅而能與果作因。不言定知如字在紙。罪業亦爾。以此身造業是業雖滅果報不失
(T. 1646: 255c24-26).
126
Robert Kritzer
that arises later, namely in the present.56
The Sthavira, too, explains that the past does not really exist, despite the
efficacy of past karma in producing real results. However, unlike the opponent of
Sarvāstivāda in the *Vibhāṣā and Harivarman, he mentions a mechanism,
anudhātu, by which causation can be maintained.57 Katō notes the similarity
between the Sthavira’s theory of anudhātu and Vasubandhu’s theory of saṃtatipariṇāmaviśeṣa and bīja (251-260).
As for Vasubandhu, what he says about the arising of real results from past
karma is not much different from any of the other non-Sarvāstivādin sources:
past karma is nothing more than the potential to yield result that is deposited in
the saṃtati; if it were something that actually existed, it would be present, not
past.58 If we compare this passage with the passage in the Abhidharmadīpa, the
words sāmarthya in the Abhidharmakośabhāṣya and śakti in the Abhidharmadīpa,
both meaning “power,” might seem strikingly similar. In fact, in a different
context, the Abhidharmakośavyākhyā gives both words as synonyms of bīja. 59
However, both Harivarman and the Sthavira also refer to the power of karma.
Thus, there does not really seem to be an especially close connection between the
statement attributed by the Dīpakāra to Dārṣṭāntika and Vasubandhu’s characteristically Sautrāntika ideas. Rather, this statement should be equally acceptable
to both Dārṣṭāntika and Vasubandhu, and there seems to be no reason to insist
that the Dīpakāra is specifically attacking Vasubandhu here.
H. The p ast i s n ot comple tely n one xi s tent : it exi st s pr ovi si on all y. 60
This statement, which the Dīpakāra attributes to the Dārṣṭāntikas, closely
resembles what Vasubandhu says in the Abhidharmakośabhāṣya: “We too say that
the past and future exist. The past is what has existed previously, while the
future is that which will exist when the cause (for its production) is present. It is
said that it (i.e., the past and future) exists in this way, not in a substantial way.” 61
56
論者言。迦葉鞞道人説。未受報業過去世有。餘過去無。答曰。此業若失則過去。過去若不失是則為常。
失者過去異名。則為失已。復失是業與報作因已滅。報在後生 (T. 1646: 258c10-13). Katsura points out
that Harivarman does not mention saṃtānapariṇāmaviśeṣa (41).
57
彼過去業亦可説有。有因縁故有隨界故。未有能遮彼相續故。彼異熟果未成熟故。最後方能牽異熟故。
然去來世非實有體 (T. 1562: 627b20-22; La Vallée Poussin, “Documents” 66).
58
tatra punas tadāhitaṃ tasyāṃ saṃtatau phaladānasāmarthyaṃ saṃdhāya uktaṃ / anyathā hi svena
bhāvena vidyamānam atītaṃ na sidhyet (Abhidharmakośabhāṣya: 299.10-11; La Vallée Poussin,
L’Abhidharmakośa 4: 59).
59
bījataḥ śaktitaḥ sāmarthyata ity arthaḥ (Abhidharmakośavyākhyā: 125.5-6). I am indebted to Fukuda
Takumi for both the observation and this reference.
60
atra pratyavatiṣṭhante dārṣṭāntikāḥ na brūṃaḥ sarvathā ’tītaṃ na vidyate / kiṃ tarhi / dravyātmanā na
vidyate prajñaptyātmanā tu sad iti (Abhidharmadīpa: 278.6-7).
61
vayam api brūmo ʾsty atītānāgatam iti / atītaṃ tu yad bhūtapūrvam / anāgataṃ yat sati hetau bhaviṣyati
/ evaṃ ca kṛtvā ʾstīty ucyate na tu punar dravyataḥ (Abhidharmakośabhāṣya: 299.1-2; La Vallée Poussin,
LʾAbhidharmakośa 4: 58).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
127
A case can be made here for the Dīpakāra’s freely substituting “Dārṣṭāntika”
for “Sautrāntika.” The wording of the statement attributed by the Dīpakāra to
Dārṣṭāntika and that of Vasubandhu’s statement are very close. Nor can we find a
corresponding statement made by any of the usual sources for Dārṣṭāntika
opinions, namely the Dārṣṭāntika in the *Vibhāṣā, Harivarman in the *Tattvasiddhi,
or the Sthavira in the Nyāyānusāra. However, the statements are not identical:
Vasubandhuʾs seems more doctrinally developed in that it is related to the
formulation that a dharma exists after previously not existing and ceases to exist
after having existed. This is a concept that is found in the Yogācārabhūmi but is
nowhere attributed to Dārṣṭāntika (Kritzer, Sautrāntika 336-338).
The *Vibhāṣā does not record a Dārṣṭāntika statement that past and future are
provisional. However, it indicates that Dārṣṭāntika frequently considers as
provisional dharmas accepted as ultimately real by the Sarvāstivādins.
Harivarman seems to go one step further than Vasubandhu. After he
concludes that dharmas in general exist conventionally but not ultimately, the
question is asked whether the same thing can be said of past and future.
Harivarman then denies the existence of past and future even according to
conventional truth,62 and he says that when the Buddha refers to past and future,
it is upāya and not to be taken as ultimate truth.63 Here, Harivarman sounds more
like a Mādhyamika than like a Dārṣṭāntika.
Saṃghabhadra, in the introduction to his discussion of Vasubandhuʾs
treatment of the three times, attributes the belief in a provisional past and future
to “some” (T. 1562: 624c6-625a9; La Vallée Poussin, “Documents” 47-49), mentioning neither Dārṣṭāntika nor the Sthavira here. Later, he identifies Vasubandhu’s
statement, quoted above, as the opinion of the sūtra-master (ching chu 經主) and
attacks Vasubandhu personally and at considerable length (T. 1562: 626b20627a10; La Vallée Poussin, “Documents” 60-63).
However, there remains the possibility that the Dīpakāra thinks that
Vasubandhu, when he says that past and future exist provisionally, is relying on
an older tradition, one that was indeed associated with Dārṣṭāntika. The whole
point of the argument for sarvāstivāda is that dharmas of the three times are all
real and not provisional (see Cox, Disputed Dharmas 137-141). For example, one of
the arguments given by the *Vibhāṣā in support of the reality of past and future is
as follows: if the past and future were not real, then the present also would not
62
問曰。若以世諦有者。今還以世諦故。説過去未來為有為無。答曰。無也 (T. 1646: 255a18-20; see
Katsura 44).
63
又汝言是正見者。以此身起業。此業與果作因已滅。復後還自受故説有果。於佛法中若有若無皆方便説。
為示罪福業因縁故。非第一義。如以因縁説有衆生。去來亦爾。依過去意者。是方便依。不如人依壁等
(T. 1646: 255c19-23).
128
Robert Kritzer
exist because it is provisionally established with relation to past and future.64
This passage suggests that the opponents of Sarvāstivāda at the time of the
*Vibhāṣā argued that past and future are not real because they are provisionally
established with relation to the present. 65 Although this passage does not
mention Dārṣṭāntika by name, it is quite possible that the position to which it is
reacting is a Dārṣṭāntika one.
In any case, the structure of the Dīpakāra’s argument suggests that he is not
necessarily associating the statement he identifies as Dārṣṭāntika directly with
Vasubandhu. In verse 319, the Dīpakāra refutes the Dārṣṭāntika statement. In
verse 320 and its commentary, he reproduces verse V 27a from the
Abhidharmakośabhāṣya, mentioning the Kośakāra by name, and criticizes Vasubandhu’s questioning of the Sarvāstivādin idea of kāritra. The “Dārṣṭāntikas”
mentioned by the Dīpakāra in verse 320 may refer to Vasubandhu’s predecessors,
and this may not be a veiled attack on Vasubandhu, himself.
III. Sautrāntika in the Abhidharmadīpa
A. The quest i on is raise d as t o how s ome one can be s ai d t o be a
murderer if al l sa ṃ skār as (in clud ing the vi cti m’ s ska nd ha s) ar e
au tom ati cally des tr oye d at e very mom ent. The Sautr ān ti k a
expl an ati on i s t hat i t i s li ke the case of a l amp’ s being b l own ou t or
a bell ’s be ing silen ce d. (The mur derer s t ops the vi ctim ’s bre ath
from con tinui ng, and thi s imp lies pre vent in g the future repl icati on
of t he vi cti m’s ska nd ha s. ) 66
The exchange regarding the apparent contradiction between the doctrine of
momentariness and the possibility of murder is found as early as the *Vibhāṣā,
where the question is phrased somewhat differently: “The killing of which
skandhas is given the name ‘murderʾ? (The skandhas) of the past? Of the future? Of
the present?” The answer is that when one destroys the conditions for the
arising of the future skandhas of another person, the sin of murder is incurred.67
Similarly, the *Tattvasiddhi explains that the five skandhas, although momentary,
are continually reborn in a series; the destruction of the series is what is called
64
復次若過去未來非實有者。彼現在世應亦是無。觀過去未來施設現在故 (T. 1545: 393b21-22; La Vallée
Poussin, “Documents” 10).
65
Similar reasoning may also be found in passages in the Prasannapadā (Mūlamadhyamakakārikās:
376-.1-7) and the Yogācārabhūmi (T. 1579: 588c10-22; Yogācārabhūmit: zi 29b1-6) regarding the nonidentity and non-difference of cause and effect (I am indebted to Nobuyoshi Yamabe for these
references).
66
atra sautrāntikāḥ parihāram āhuḥ na / pradīpanirvāpanaghaṇṭaśabdanirodhavat tatsiddheḥ (Abhidharmadīpa: 157.6-7).
67
問殺何蘊名殺生。過去耶。未來耶。現在耶。過去已滅。未來未至。現在不住。悉無殺義云何名殺生耶。
答殺未來蘊。非過去現在。問未來未至云何可殺。答彼住現在遮未來世諸蘊和合説名為殺。由遮他蘊和合
生縁故得殺罪 (T. 1545: 617a27-b3).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
129
murder, not the destruction of the present skandhas.68 Thus, both the orthodox
Vaibhāṣikas and the unorthodox Harivarman answer the question in basically the
same way.
In the Abhidharmakośabhāṣya, however, Vasubandhu introduces the similes of
the lamp and bell, similes that he uses elsewhere in support of a Sautrāntika
position. This seems to be the basis for the Dīpakāra’s associating the simile with
Sautrāntika. In Vasubandhu’s discussion of murder, he offers two answers to the
question of how someone can be called a murderer, which Fa-pao says is posed by
the Saṃmatīyas (T. 1822: 666b19). The first answer is that the word prāṇa in the
compound prāṇātipāta (murder) refers to the breath on which the body and mind
depend. The murderer destroys this breath in the same way that one puts out a
lamp or stops the sound of a bell. The second answer is that, by preventing the
jīvitendriya from being reproduced, the murderer destroys it and becomes guilty
of murder. 69 As Jaini points out, Vasubandhu does not attribute the first
explanation to the Sautrāntikas (Abhidharmadīpa: 157 n. 2).70 However, at the
beginning of Chapter Four of the Abhidharmakośabhāṣya (Karmanirdeśa), Vasubandhu uses the same example of the lamp and the bell in establishing that
destruction has no cause and is inherent in the momentary nature of saṃskṛtadharmas.71 Vasubandhu’s use of this simile does not coincide with the Sarvāstivādin understanding of the process of destruction. According to Saṃghabhadra,
Sarvāstivāda says that the destruction of all saṃskṛtadharmas is caused by a
cittaviprayuktasaṃskāra, anityatā, which arises together with them.72 Vasubandhu,
68
問曰。若此五陰念念常滅。以何為殺。答曰。五陰雖念念滅還相續生。斷相續故名為殺生。又是人以有
殺心故得殺罪。問曰。為斷現在五陰故名殺生耶。答曰。五陰相續中有衆生名。壞此相續故名殺生。不以
念念滅中有衆生名 (T. 1646: 304b27-c4).
69
kathaṃ kṣaṇikeṣu skandheṣu prāṇātipāto bhavati / prāṇo nāma vāyuḥ kāyacittasaṃniśrito vartate / tam
atipātayati / yathā pradīpaṃ nirodhayati ghaṇṭāsvanaṃ vā / jīvitendriyaṃ vā prāṇas tan nirodhayati / yady
ekasyāpi jīvitakṣaṇasyotpadyamānasyāntarāyaṃ karoti prāṇātipātāvadyena spṛśyate nānyathā
(Abhidharmakośabhāṣya: 243.16-19; La Vallée Poussin, L’Abhidharmakośa 3: 153-154). Underlined
words represent Rospatt’s corrections of Pradhan’s text (186 n. 407).
70
In fact, there is some confusion among the commentators on this point. Fa-pao refers to two
different explanations and concludes that the second is preferable (T. 1822: 666b22-29), while P’ukuang refers to two teachers and says that, although their explanations are different, they agree
that the murderer obstructs the future skandhas rather than destroying the present ones (T. 1821:
257c29-258a17). Sthiramati, on the other hand, after explaining the simile of the lamp, says that the
sound of the bell and jīvitendriya are to be understood in the same way, thus suggesting that
Vasubandhu’s two explanations are not contradictory (Tattvārtha: tho 189b4-6). Agostini assumes
that Vasubandhu prefers the second explanation, but he does not refer to any of the commentators
(69).
71
Abhidharmakośabhāṣya: 193.5-194.11; La Vallée Poussin, L’Abhidharmakośa 3: 4-8. See Rospatt 183186; Kritzer, Vasubandhu 174-175.
72
滅謂別法。是倶生行滅壞勝因。性是體義 (T. 1562: 405c9-10). For a translation and for references to
the *Vibhāṣā, see Cox, Disputed Dharmas 306, 356 n. 6).
130
Robert Kritzer
on the other hand, explicitly attributes to Sautrāntika a rejection of the reality of
the saṃskṛtalakṣaṇas.73
Therefore, it is probably because this simile implies a rejection of anityatā as a
real dharma that the Dīpakāra labels it Sautrāntika.74 He objects to Vasubandhu’s
explanation, arguing that the simile is inappropriate since, after the flame has
been blown out or the sound stopped, an identical flame or sound can be
produced again, whereas the breath destroyed by the murderer cannot.75 The
Dīpakāra goes on to give his own explanation of how murder is possible even
though the skandhas are momentary: “When the killer obstructs the capability of
the cause (of production of the skandhas), due to his impeding the means of
having the power and operation of future saṃskāras, there is the occurrence of
murder.” 76 This explanation uses the language of the neo-Vaibhāṣikas 77 to
amplify the answer given in the *Vibhāṣā.
B. T he Ṣaṭ ṣaṭka sūtra s ay s th at a pers on e xpe rienci ng a ple asur ab le
feeling h as r āgā n uśay a.
This is explained by Vasubandhu as referring to the anuśaya of rāga before it is
activated (i.e., when it is still a potential), or as referring to the result, i.e., the
activated kleśa, by the name of its cause, i.e., anuśaya. The Dīpakāra says that this
explanation is inserted by the Sautrāntikas due to their sloth at listening to the
wisdom of the Buddha’s words.78
Here, the Dīpakāra identifies as Sautrāntika Vasubandhu’s explanation of the
Ṣaṭṣaṭkasūtra in support of an argument that Vasubandhu clearly labels as
Sautrāntika. This statement of the Dīpakāra comes immediately after the long
quotation from the Abhidharmakośabhāṣya concerning anuśaya that was
mentioned above in the discussion of Sautrāntika item II.D (Abhidharmakośabhāṣya: 278.17-279.4; La Vallée Poussin, L’Abhidharmakośa 4: 6-7). In the
Abhidharmakośabhāṣya, this passage presents the theory of anuśaya that is
73
tad etad ākāśaṃ pāṭyata iti sautrāntikāḥ / na hy ete jātyādayo dharmā dravyataḥ saṃvidyante yathā
vibhajyante (Emended from abhivyajyante by Shastri [v. 1: 257.2]) / kiṃ kāraṇam / pramāṇābhāvāt / na hy
eṣāṃ dravyato ʾstitve kiṃcid api pramāṇam asti pratyakṣam anumānam āptāgamo vā yathā rūpādīnāṃ
dharmāṇām iti (Abhidharmakośabhāṣya: 76.20-23; La Vallée Poussin, L’Abhidharmakośa 1: 226; see Kritzer,
Vasubandhu 70-71).
74
Rospatt, writing about Vasubandhu’s argument that destruction has no cause, says, “This
argument is characteristic for the Sautrāntikas, whereas it is at odds with the doctrine of the
Sarvāstivādins according to whom destruction is caused by the mark of destruction” (187).
75
na samānatvāt (Abhidharmadīpa: 157.8). I follow Mitomo’s explanation here.
76
ayaṃ tv atra parihāraḥ hantur hetusāmarthyopaghātakaraṇe saty anāgatasaṃskāraśaktikriyādhānavidhānavighnakaraṇāt prāṇātipātopapattiḥ (Abhidharmadīpa: 157.8-9).
77
For this term, see Cox (Disputed Dharmas 58).
78
tad etad sautrāntikair antargataṃ buddhavacananītiśravaṇakausīdyam āvirbhāvyate (Abhidharmadīpa:
224.1).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
131
expressly said to be that of Sautrāntika, and it is clear that Vasubandhu approves.
Saṃghabhadra also comments on Vasubandhu’s interpretation of the
Ṣaṭṣaṭkasūtra, which he criticizes at length, saying that it is the sūtra-master’s
explanation (T. 1562: 598b16-c16; see Kritzer, Vasubandhu 276-277).
C. The con cept of bīja as śakti is said t o be a figmen t of th e
im agin ati on of the S au trānti kas. 79
Again, what the Dīpakāra calls Sautrāntika is indeed part of Vasubandhu’s
Sautrāntika argument. In the Abhidharmakośabhāṣya, the question is asked, “What
is this seed-state [of a kleśa]?” Vasubandhu answers, defining it as a person’s
power (śakti), produced by kleśa in the past, to give rise to kleśa in the future. 80
The Dīpakāra’s statement is a continuation of his criticism of Vasubandhu’s idea
of anuśaya, which was discussed in the previous item. Jaini points out the
similarity between the Dīpakāra’s argument here and Saṃghabhadra’s arguments
against the theory of bīja that Vasubandhu proposes in Chapter Two of the
Abhidharmakośabhāṣya, where he denies the real existence of the
cittaviprayuktasaṃskāra, prāpti.81 Saṃghabhadra, at the beginning of his discussion
of prāpti, again uses the epithet sūtra-master to show that he is arguing
specifically against Vasubandhu (T. 1562: 397a12).
D. The S au trānti kas d am age the ir own p os i ti on by re lyin g on the
Para m ārth aśū ny atā sūtr a, whi ch w as inte nde d to refu te the S aṃ kh y ā
the ory of an etern al sub st an ce w ith ch ang ing aspe ct s ( see Jaini ,
Intr odu cti on 119 ), n ot t o deny the reali ty of pas t an d future . 82
Although Vasubandhu does not mention Sautrāntika in referring to the
Paramārthaśūnyatāsūtra, the context is that of his refutation of the reality of past
and future, in the course of which he does use the word Sautrāntika.83 In the
passage commented on by the Dīpakāra, Vasubandhu adduces a statement in the
Paramārthaśūnyatāsūtra in support of his position: when cakṣurindriya appears, it
does not come from anywhere, and when it is destroyed, it does not go anywhere;
79
sautrāntikaparikalpite pratibījakalpe cittaśaktibījabhāvanāpakṣe nivṛttyuttaram anyānanyatvādidoṣāt
(Abhidharmadīpa: 225.3-4). The entire passage is summarized by Jaini (Introduction 108).
80
ko ’yam bījabhāvo nāma / ātmabhāvasya kleśajā kleśotpādanaśaktiḥ (Abhidharmakośabhāṣya: 278.20-21;
La Vallée Poussin, L’Abhidharmakośa 4: 6).
81
Jaini relies on a quotation from Saṃghabhadra in the Abhidharmakośavyākhyā (148.22-149.2),
which, as Wogihara states, is abridged. The original passage from the *Nyāyānusāra (T. 1562: 397b5c10 is translated in Cox, Disputed Dharmas 187-191).
82
tasmād durvihitavetāḍotthānavat sautrāntikaiḥ svapakṣopaghātāya sūtram etad āśrīyate (Abhidharmadīpa: 268.19-20)
83
naiva hi sautrāntikā atītāt karmaṇaḥ phalotpattiṃ varṇayanti / kim tarhi / tatpūrvakāt saṃtānaviśeṣād
ity ātmavādapratiṣedhe saṃpravedayiṣyāmaḥ (Abhidharmakośabhāṣya: 300.19-21; La Vallée Poussin,
L’Abhidharmakośa 4: 63).
132
Robert Kritzer
it exists after not having existed, and after existing, it does not exist.84
Here again Saṃghabhadra’s argument resembles the Dīpakāra’s: the Buddha
preached this sūtra not to illustrate the non-existence of past and future but to
counter un-Buddhist ideas, including the notions that there is a Creator who
makes the eye or that the result is encompassed by the cause85 (T. 1562: 625c27626a22; La Vallée Poussin, “Documents” 56-58). Again Saṃghabhadra identifies
Vasubandhu’s argument as that of the sūtra-master (T. 1562: 627a10-15; La Vallée
Poussin, “Documents” 63). Immediately after summarizing his objections to
Vasubandhu’s misuse of scripture, Saṃghabhadra turns his attention to the
Sthavira (T. 1562: 627b11-c16; La Vallée Poussin, “Documents” 65-67). Thus,
Saṃghabhadra distinguishes here between Vasubandhu’s and the Sthavira’s
ideas.
IV. Sthitibhāgīya
A. The S thi tibh āg īy as s ay th at ab hid hy ā, e t c., have t he n ature of
karm a. 86
Vasubandhu records a disagreement between the Dārṣṭāntikas and the
Vaibhāṣikas regarding the status of the three types of mental misconduct,
namely, greed, ill will, and false views. According to Vaibhāṣika, they are kleśa,
but the Dārṣṭāṇtikas say that they are karma.87 Katō discusses this passage in
detail, and shows that Saṃghabhadra is mistaken in assuming that Vasubandhu
agrees with Dārṣṭāntika here. Katō points out that Saṃghabhadra elsewhere
attributes to the Sthavira the statement that the three types of mental
misconduct are nothing other than mental karma. He theorizes that
Saṃghabhadra, who considers the Sthavira to be Vasubandhu’s teacher, assumes
that Vasubandhu agrees with the Sthavira, even though he really does not (Katō
81-84).
Yaśomitra, who elsewhere flatly equates Dārṣṭāntika and Sautrāntika, here
glosses Dārṣṭāntika as “a type of Sautrāntika” (see footnote 1). This perhaps
shows that Yaśomitra recognizes Vasubandhu’s disagreement with this position:
in general, Vasubandhu agrees with Sautrāntika, so this Dārṣṭāntika, according to
Yaśomitra, may indeed be a Sautrāntika, but a different type of Sautrāntika from
84
itthaṃ caitad evaṃ yat paramārthaśūnyatāyām uktaṃ bhagavatā cakṣur utpadyamānaṃ na kutaścid
āgacchati nirudhyamānaṃ na kvacit saṃnicayaṃ gacchati / iti hi bhikṣavaś cakṣur abhūtvā bhavati bhūtvā
ca prativigacchatīti [emended from pratigacchatīti by Pāsādika {68}] (Abhidharmakośabhāṣya: 299.12-14;
La Vallée Poussin, L’Abhidharmakośa 4: 59; see Kritzer Vasubandhu: 304-307).
85
This is identified by La Vallée Poussin as a Sāṃkhya theory (“Documents” 57 n. 4).
86
abhidhyādaya eva karmasvabhāvāni iti stithibhāgīyāḥ / tac ca na karmakleśaikatvadoṣāt / sthitibhāgīyānāmaśākyāḥ svalāṅgūlikadvitīyanāmānaḥ (Abhidharmadīpa: 148.10-11).
87
abhidhyādaya eva manaskarmeti dārṣṭāntikāḥ / saṃcetanīyasūtre vacanāt / evaṃ tu sati karmakleśayor aikyaṃ
syāt / kiṃ syād yadi kaścit kleśo ’pi karma syāt / naitad asti (Abhidharmakośabhāṣya: 237.17-18; La Vallée
Poussin, L’Abhidharmakośa 3: 136). See Kritzer, Vasubandhu 250-251).
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
133
the ones whom Vasubandhu usually cites approvingly. Neither Sthiramati
(Tattvārtha: tho 183b1-2) nor Pūrṇavardhana (Lakṣaṇānusārinī: ñu 65a6-7)
mentions Sautrāntika; they both attribute the position to Dārṣṭāntika, and they
both explicitly reject it, presumably in support of what they understand to be
Vasubandhu’s rejection.
The Dīpakāra, like Saṃghabhadra, assumes Vasubandhu’s agreement with the
statement that abhidhyā, etc., are karma, which he attributes to neither
Sautrāntika nor Dārṣṭāntika, but rather to the Sthitibhāgīyas. The meaning of the
term sthitibhāgiya is unclear. Jaini speculates as follows: “The word sthiti in this
term most probably means pravāha or santati. The Sautrāntikas do not recognise
sthiti (subsistence) as a separate saṃskṛita-lakṣaṇa, because they consider that it is
a name given to the series of momentary dharmas (santati). By sthitibhāgīya,
therefore, the Dīpakāra means one who belongs to or is heading for the (theory
of) santati, a term which a Vaibhāshika could use as an abuse to the Sautrāntika”
(Introduction 100).88 Here, however, the Dīpakāra does not actually refer to the
saṃtati or to any other idea that is generally attributed to Sautrāntika. Rather he
says that accepting the equation of kleśa and karma would be tantamount to
accepting the system of the Sāṃkhyas, etc.89 Thus, while it is clear that the
Dīpakāra thinks that Vasubandhu believes that abhidhyā, etc., are karma and
attacks him for his opinion, Jaini may be overstating the extent to which the
term Sthitibhāgīya is specifically directed against Sautrāntika.
V. Conclusion
Beginning with Vasubandhu in his Abhidharmakośabhāṣya, Buddhist authors refer
to the opinions of a group they call Sautrāntikas. Most modern scholars simply
repeat this term uncritically and likewise refer to Sautrāntika ideas, thus giving
the impression that Sautrāntika was a significant doctrinal classification even if it
was not a vinaya sect. It is generally assumed that there were, in fact, Sautrāntika
thinkers who preceded Vasubandhu.
Honjō summarizes the reasons “to believe that the name ‘Sautrāntika’ dates
back to the pre-Vibhāṣā period” (326-327). First, he mentions Katō’s claim that
Śrīlāta is the earliest thinker to refer to himself as Sautrāntika. But Honjō
disagrees and argues for an even earlier use of the word. First of all, according to
Honjō, the commentaries do not recognize a difference between Sautrāntika and
Dārṣṭāntika. Second, as Przyluski argues, the name Dārṣṭāntika, which appears so
often in the *Vibhāṣā, is pejorative, and the group to which it refers must have
called themselves something else, i.e., Sautrāntika. (This argument has been challenged by Harada, who maintains that the term Dārṣṭāntika is not, in fact, pejora88
I have not altered Jaini’s transliteration of Sanskrit terms here.
karmakleśānāṃ cātyantasvabhāvaprabhāvakriyāphalabhedabhinnānām ekatvaparikalpaiḥ sāṃkhyīyādidarśanam abhyupagataṃ syāt (Abhidharmadīpa: 149.2-4).
89
134
Robert Kritzer
tive [2006].) Furthermore, the Dārṣṭāntikas in the *Vibhāṣā implicitly reject the
authority of the abhidharma, thus showing themselves to be Sautrāntikas in the
sense of relying only on sūtra. Finally, Honjō points to two passages in the Abhidharmakośavyākhyā that mention Sautrāntika opinions recorded in the *Vibhāṣā.
However, the evidence for the existence of the Sautrāntikas before Vasubandhu is, to my mind, not very convincing. Ultimately, it is all based on three
things: first, Hsüan-tsang’s translation of the Samayabhedoparacanacakra and
K’uei-chi’s commentary on that and several other texts; second, the testimony of
commentators on the Abhidharmakośabhāṣya; and third, the similarity between
certain of the opinions attributed to Sautrāntika by Vasubandhu and those
attributed to Dārṣṭāntika in the *Vibhāṣā. This similarity may in fact be the
reason why Dārṣṭāntika positions not attributed to Sautrāntika by Vasubandhu
were more and more frequently labeled as Sautrāntika by later commentators.
However, I must reiterate the crucial fact that there is no reliable textual
reference to Sautrāntika prior to the Abhidharmakośabhāṣya.
If my analysis of the Dārṣṭāntika and Sautrāntika positions in the Abhidharmadīpa is correct, it appears as though the Dīpakāra, one of the earliest commentators on the Abhidharmakośabhāṣya, maintains more of a distinction between
Dārṣṭāntika and Sautrāntika than do later authors, such as Yaśomitra. The
Dīpakāra never calls Dārṣṭāntika positions with which Vasubandhu disagrees
Sautrāntika,90 and on several occasions he identifies as Dārṣṭāntika positions elsewhere thus identified, with which Vasubandhu disagrees. However, the Dīpakāra,
unlike Vasubandhu, seems to refer to as Dārṣṭāntika some of the ideas that
Vasubandhu (like the Yogācārabhūmi) inherited from the Dārṣṭāntika of the
*Vibhāṣā and appropriated as his own under the designation Sautrāntika.
Without making a detailed analysis of all the occurrences of Dārṣṭāntika and
Sautrāntika in the entire commentatorial literature on the Abhidharmakośabhāṣya, I cannot say anything definite about the evolution in the use of the two
terms. However, what can be observed is that, in the course of time, references to
Sautrāntika increased dramatically in comparison to those of Dārṣṭāntika.91
One can conjecture that in the earlier period there was a greater awareness
that certain of the unorthodox positions mentioned by Vasubandhu originated
90
But see my comments above regarding Sthitibhāgīya (item IV.A).
A simple chart of the number of occurrences of each word might be useful. Except in the case of
the Abhidharmadīpa and the Abhidharmakośavyākhyā, these numbers are based on a computer search
of digital texts and are only to be taken as approximate.
Dārṣṭāntika
Sautrāntika
Nyāyānusāra
52
31
Abhidharmadīpa
8
4
Abhidharmakośavyākhyā
6
43
P’u-kuang
23
564
Fa-p’ao
36
411.
91
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
135
among the group referred to as Dārṣṭāntika in the Vibhāṣā or among teachers
such as Dharmatrāta, Harivarman, and Śrīlāta, who were closely associated with
Dārṣṭāntika, while other unorthodox views did not. Later, the varied origins of
Vasubandhu’s opinions were forgotten. Because of the similarity between many
of the positions he labels Sautrāntika and the earlier Dārṣṭāntika ideas, it came to
be assumed that Vasubandhu accepted all of these Dārṣṭāntika opinions, which
were now designated as Sautrāntika due to their presumed association with
Vasubandhu.
The question may still be asked: if there really were not a group called
Sautrāntika before Vasubandhu, why would none of the commentators, including
ones like Saṃghabhadra and the Dīpakāra, who probably either were
Vasubandhu’s contemporaries or lived soon after him, show an awareness of this
fact? I cannot answer this satisfactorily, and so the real significance of the term
Sautrāntika must remain something of a mystery.
List of works cited
1. Chinese Sources
*Abhidharmapiṭakaprakaraṇaśāsanaśāstra (A-p’i-ta-mo-tsang hsien-tsung lun). By Saṃghabhadra. T. 1563.
Chü-she lun chi 倶舎論記. By P’u-kuang 普光. T. 1821.
Chü-she lun shu 倶舎論疏. By Fa-pao 法寶. T. 1822.
*Nyāyānusāra (A-p’i-ta-mo shun cheng-li lun 阿毘達磨順正理論). By Saṃghabhadra. T. 1562.
*Tattvasiddhiśāstra (Ch’eng-shih lun 成實論). T. 1646.
*Vibhāṣā (A-p’i-ta-mo ta-p’i-p’o-sha lun 阿毘達磨大毘婆沙論). Trans. Hsüan-tsang. T. 1545.
*Vibhāṣā (A-p’i-t’an p’i-p’o-sha lun 阿毘曇毘婆沙論). Trans. Buddhavarman. T. 1546.
*Vibhāṣā (P’i-po-sha lun 鞞婆沙論). Trans. Saṃghabhadra. T. 1547.
*Pañcavastukavibhāṣāśāstra (Wu-shih p’i-p’o-sha lun 五事毘婆沙論). By Dharmatrāta (法救).
T. 1555.
2. Sanskrit and Pāli Sources
Abhidharmadīpa with Vibhāṣāprabhāvṛtti. Edited by P. S. Jaini. Tibetan Sanskrit Works Series
4. Second edition. Patna: Kashi Prasad Jayaswal Research Institute, 1973.
Abhidharmakośa & Bhāṣya of Acharya Vasubandhu with Sphuṭārthā commentary of Ācārya
Yaśomitra. Edited by Swami Dwarkidas Shastri. Bauddha Bharati Series 6. Varanasi:
Bauddha Bharati, 1971.
Abhidharmakośabhāṣya. Edited by P. Pradhan. Tibetan Sanskrit Works Series VIII. Patna: K.
P. Jayaswal Research Institute, 1967.
Abhidharmakośavyākhyā. Edited by U. Wogihara. Tokyo: Sankibo Buddhist Bookstore, 1990.
Reprint. (First edition: Tokyo: The Publishing Association of the Abhidharmakośavyākhyā, 1932-1936.)
Kathāvatthu. Ed. Arnold C. Taylor. Pali Text Society Text Series 48-49. London: The Pali
136
Robert Kritzer
Text Society, 1979. Reprint. (First edition 1894-1897.)
Mūlamadhaymakakārikās (Mādhyamikasūtras) de Nāgārjuna avec la Prasannapadā Commentaire
de Candrakīrti. Edited by Louis de la Vallée Poussin. Bibliotheca Buddhica 4.
Osnabrück: Biblio Verlag, 1970. Reprint. (First edition 1903-1913.)
3. Tibetan Sources (Cited by Sanskrit title if the original work was in Sanskrit)
Lakṣaṇānusāriṇī. Peking Bstan ʾgyur 5594 (mṅon paʾi bstan bcos ñu).
Tattvārtha. Peking Bstan ʾgyur 5875 (sna tshogs tho).
4. Miscellaneous
Kokuyaku Issaikyō 国譯一切經.
Bidon-bu 7-17 (Japanese translation of *Vibhāṣā).
Bidon-bu 25-26 (Japanese translation of Abhidharmakośabhāṣya).
Ronjū-bu 3 (Japanese translation of *Tattvasiddhiśāstra).
5. Modern Works
Agostini, Giulio. “Buddhist Sources on Feticide as Distinct from Homicide.” Journal of the
International Association of Buddhist Studies 27.1 (2004): 63-95.
Aung, Shwe Zan and Mrs. Rhys Davids. Points of Controversy or Subjects of Discourse (English
translation of Kathāvatthu). London: The Pali Text Society, 1969. Reprint. (First
Edition 1915.)
Cox, Collett. Disputed Dharmas: Early Buddhist Theories on Existence. Studia Philologica
Buddhica Monograph Series 11. Tokyo: The International Institute for Buddhist
Studies, 1995.
Cox, Collett. “On the Possibility of a Nonexistent Object of Consciousness: Sarvāstivādin
and Dārṣṭāntika Theories.” Journal of the International Association of Buddhist Studies
11.1 (1988): 31-87.
Dhammajoti, K.L. “The Sarvāstivāda Doctrine of Simultaneous Causality.” The Numata
Yehan Lecture in Buddhism Winter 2000. Calgary: University of Calgary, 2000.
Fukuda Takumi 福田琢 . “Jōjitsuron no zuimen ron” 『成実論』の随眼論. Katō Junshō
Hakushi kanreki kinen ronshū: Abidaruma Bukkyō to Indo shisō 加藤純章博士還暦記念論
集 アビダルマ仏教とインド思想. Ed. Katō Junshō Hakushi Kanreki Kinen Ronshū
Kankōkai. Tokyo: Shunjūsha, 2000. 151-165.
Funahashi Issai 舟橋一哉著. Kusharon no genten kaimei: Gōhon 倶舎論の原典解明 業品. Kyoto:
Hōzōkan, 1987.
Harada, Wasō 原田和宗. “Kyōryōbu o meguro shomondai (2)” 經量部をめぐる諸問題 (2).
Indogaku Bukkyōgaku Kenkyū 54.2 (2006): 51-55.
–
“‘Kyōryōbu no tansō no shiki no nagareʾ to iu gainen e no gimon [II]” <經量 部 の
「単層の」識の流れ>という概念への疑問(II). Indogaku Chibettogaku Kenkyū
2 (1997): 22-59.
Hirose Tomoichi 広瀬智一. “Abidatsuma Tōronshaku ni okeru bukkyō gakuha hihan” 阿毘達
Dārṣṭāntika and Sautrāntika in the Abhidharmadīpa
137
磨灯論釈における仏教学派批判. Sōtōshūkenkyūin Kenkyūsei Kenkyūkiyō 3 (1971): 142-
126.
Honjō, Yoshifumi. “Sautrāntika.” Journal of the International Association of Buddhist Studies
26.2 (2003): 321-330.
Ichimura, Shohei, et al. “Mahāvibhāṣā.” Encyclopedia of Indian Philosophies, Vol. VII:
Abhidharma Buddhism to 150 A.D. Delhi: Motilal Banarsidass, 1996. 511-568.
Jaini, P.S. Introduction to Abhidharmadīpa with Vibhāṣāprabhāvṛtti. Edited by P. S. Jaini.
Tibetan Sanskrit Works Series 4. Second edition. Patna: Kashi Prasad Jayaswal
Research Institute, 1973.
–
“The Sautrāntika Theory of Bīja.” Bulletin of the School of Oriental and African Studies
22.2 (1959): 236-249.
Katō Junshō 加藤純章. Kyōryōbu no kenkyū 経量部の研究 (Étude sur les Sautrāntika). Tokyo:
Shunjūsha, 1989.
Katsura, Shoryu. A Study of Harivarmanʾs Tattvasiddhi. Ph.D. dissertation, Universi-ty of
Toronto, 1974.
Kimura Taiken 木村泰賢. Abidatsumaron no Kenkyū 阿毘達磨論の研究. Kimura Taiken
Zenshū 4 木村泰賢全集第4巻. Tokyo: Daihōrinkaku, 1968.
Kritzer, Robert. Vasubandhu and the Yogācārabhūmi: Yogācāra Elements in the
Abhidharmakośabhāṣya. Studia Philologica Buddhica Monograph Series 18. Tokyo:
The International Institute for Buddhist Studies, 2005.
–
“General Introduction.” Journal of the International Association of Buddhist Studies 26.2
(2003): 201-224.
–
“Sautrāntika in the Abhidharmakośabhāṣya.” Journal of the International Association of
Buddhist Studies 26.2 (2003): 331-384.
La Vallée Poussin, Louis de. LʾAbhidharmakośa de Vasubandhu. New edition. Mélanges chinois
et bouddhiques 16. 6 vols. Bruxelles: Institut Belge des Hautes Études Chinoises,
1971. Reprint. (First edition: Louvain: J.B. Istas, 1923-1931.)
–
“Documents dʾAbhidharma: la controverse du temps.” Mélanges chinois et
bouddhiques 5 (1936-1937): 7-158.
Lin Li-kouang. LʾAide-mémoire de la Vraie Loi (Saddharma-Smṛtyupasthāna-Sūtra). Paris:
Librairie dʾAmérique et dʾOrient, 1949.
Mejor, Marek. Vasubandhuʾs Abhidharmakośa and the Commentaries Preserved in the Tanjur.
Institut für Kultur und Geschichte Indiens und Tibets an der Universität Hamburg,
Alt- und Neu-Indische Studien 42. Stuttgart: Franz Steiner Verlag, 1991.
Miyamoto Shōson 宮本正尊. “Hiyusha, Daitoku Hokku, Dōju, Yumanron no Kenkyū” 譬喩
者、大徳法救、童受、喩鬘論の研究. Nihon Bukkyō Gakkai nenpo 1 (1928): 115-192.
Mizuno Kōgen 水野弘元. “Shinshiki ron to yuishiki setsu no hattatsu” 心識論と唯識説の発
達. Bukkyō Kyōri Kenkyū 仏教教理研究. Mizuno Kōgen Chosaku Senshū 2. Tokyo:
Shunjūsha, 1997. 301-342.
–
“Hiyushi to Jōjitsuron” 譬喩師と成實論. Bukkyō Kyōri Kenkyū 仏教教理研究. Mizuno
Kōgen Chosaku Senshū 2. Tokyo: Shunjūsha, 1997. 279-300.
138
Robert Kritzer
Nasu Enshō 那須円照. “Abhidharmadīpa (Abidarumadeīpa) no jikanron (sansejitsuyuron)
shiyaku” Abhidharmadīpa (『アビダルマデイーパ』)の時間論<三世実有論>試
訳. Indogaku Chibettogaku Kenkyū 7-8 (2003/2004): 49-101.
Pāsādika (Bhikkhu). Kanonische Zitate in Abhidharmakośabhāṣya des Vasubandhu. SanskritWörterbuch der buddhistischen Texte aus den Turfan Funden: Beiheft: 1.
Göttingen: Vandenhoeck and Ruprech 1986.
Potter, Karl H. Presuppositions of Indiaʾs Philosophies. Englewood Cliffs, N.J., Prentice-Hall,
1963.
Rospatt, Alexander von. The Buddhist Doctrine of Momentariness. Institut für Kultur und
Geschichte Indiens und Tibets an der Universität Hamburg, Alt- und Neu-Indische
Studien 47. Stuttgart: Franz Steiner Verlag, 1995.
Stcherbatsky, Th. The Central Conception of Buddhism and the Meaning of the Word “Dharma.”
Delhi: Motilal Banarsidass, 1970. Reprint.
Willemen, Charles, Bart Dessein, and Collett Cox. Sarvāstivāda Buddhist Scholasticism.
Leiden: Brill, 1998.
Yamabe, Nobuyoshi. “On the School Affiliation of Aśvaghoṣa: ‘Sautrāntika’ or ‘Yogācāra.’
”Journal of the International Association of Buddhist Studies 26.2 (2003): 225-254.
Yamada Ryūjō 山 田龍城 . Daijō Bukkyō Seiritsuron Josetsu 大乗仏教成立論序 説. Kyoto:
Heirakuji Shoten, 1959.
Yoshimoto Shingyō 吉元信行. Abidaruma Shisō アビダルマ思想. Kyoto: Hōzōkan, 1982.