Documente Academic
Documente Profesional
Documente Cultură
Edited by
Johannes Bronkhorst
In co-operation with
VOLUME 50
Edited by
Bart Dessein
Weijen Teng
LEIDEN | BOSTON
iv
Cover illustration: Mogao Grottoes, cave 341, northern wall. Illustration of Maireya’s Pure Land, Early Tang
dynasty (courtesy of Dharma Drum).
Want or need Open Access? Brill Open offers you the choice to make your research freely accessible
online in exchange for a publication charge. Review your various options on brill.com/brill-open.
Typeface for the Latin, Greek, and Cyrillic scripts: “Brill”. See and download: brill.com/brill-typeface.
issn 0925-2916
isbn 978-90-04-31666-9 (hardback)
isbn 978-90-04-31882-3 (e-book)
⸪
vi
Contents
Contents vii
Contents
Preface ix
List of Figure and Tables xi
Notes on Contributors xii
Introduction 1
Part 1
Mātṛkā and Abhidharma Terminologies
Part 2
Intellectual History
Part 3
Philosophical Studies
Index 413
PrefacePreface ix
Preface
The papers included in this volume are the outcome of the conference From
Abhidhamma To Abhidharma. Early Buddhist Scholasticism in India, Central
Asia, and China held on 8 and 9 July 2013 at Ghent University, Belgium. The
conference was initiated to commemorate the 120th anniversary – two 60-year
cycles in Chinese chronology – of the appointment of Louis de La Vallée Pous-
sin (1869–1938) as professor at Ghent University. During the almost 30 years he
was affiliated with Ghent University (he resigned in 1929), Louis de La Vallée
Poussin, son of a French father and a Belgian mother who chose to become a
Belgian national, laid the foundations of Belgian Buddhology. Louis de La Val-
lée Poussin is generally recognized as one of the greatest Abhidharma scholars
in the history of Buddhology. The contribution “Rétrospective: L’oeuvre de
Louis de La Vallée Poussin” in the Bibliographie bouddhique of 1955 enumerates
no less than 323 works of his hand, published between 1891 and his death in
1938.
The conference From Abhidhamma To Abhidharma. Early Buddhist Scholas-
ticism in India, Central Asia, and China was a joint initiative of Bart Dessein
and Weijen Teng, and was organized by the Ghent Centre for Buddhist Studies,
Belgium, and the Department of Buddhist Studies of Dharma Drum Institute
of Liberal Arts (Formally Dharma Drum Buddhist College), Taipei, Taiwan. Re-
searchers on Abhidhamma / Abhidharma work at different institutes all over
the globe, and their precise research activities concern different epochs of
Buddhist history, spanning from the life time of the historical Buddha to the
contemporary period; deal with Abhidharmic developments in different geo-
graphical regions, extending from India, over Southeast Asia, Central Asia, Chi-
na, Mongolia, Japan, and Tibet; and concern different types of materials, with
some researchers working on (recent) manuscript founds, and others working
on edited editions of texts in Pāli, Sanskrit, Chinese, Japanese, Mongolian or
Tibetan. While some scholars investigate the Abhidharma of one particular
school, others work in a comparative framework, or work cross-culturally.
Bringing together scientific papers of scholars dealing with such a wide va-
riety of subjects and materials in one coherent conference did not announce
itself to be an easy task. To our surprise, however, the conference papers did
show a surprising coherence, thus proving that despite the long chronological
development and the large geographical dissemination of Abhidhamma /
Abhidharma traditions, old Abhidharmic materials that are at the foundation
of the tradition have continued to determine and impregnate later develop-
x Preface
Bart Dessein
Sint Amandsberg, Belgium
Weijen Teng
Jinshan, Taiwan
List of Figure and Tables xi
Figure
Tables
Notes on Contributors
Johannes Bronkhorst
(PhD Pune 1979, doctorate Leiden 1980) is professor emeritus at the University
of Lausanne. He has published numerous research papers and books, includ-
ing Greater Magadha (Brill, 2007) and Buddhism in the Shadow of Brahmanism
(Brill, 2011). His book How the Brahmins Won (Brill, 2016) has recently come out.
Lance S. Cousins
was active at Oxford University until his death on March 14, 2015. He taught
Pali and Middle Indian at the Faculty of Oriental Studies, and Buddhism in the
Faculty of Theology. He is the author of numerous publications on Pali texts,
early Buddhist schools, Buddhist meditation and Buddhist ethics.
Dhammajoti KL Bhikkhu
is currently the Glorious Sun Professor at the Centre of Buddhist Studies, Uni-
versity of Hong Kong. He has taught Buddhism in several institutes around the
world including the University of Kelaniy, Sri Lanka; the University of Calgary,
Numata Chair in Buddhist Studies; and the International Buddhist College,
Thailand. Prof. Dhammajoti specialized in Theravāda and Sarvāstivāda Abhi
dharma, as well as Mahāyāna doctrines of the Madhyamaka and Yogācāra
schools. He has published many journal articles and is the author of The Chi-
nese Dharmapada, The Sarvāstivāda Abhidharma, and is the Editor of Journal
of Buddhist Studies, Sri Lanka.
Bart Dessein
is Professor at the Department of Languages and Cultures of Ghent University,
Belgium. His research focus is on philosophical developments in early Bud-
dhism and early school formation. He has published mainly on the
Sarvāstivādins and the Mahāsāṃghikas.
Tamara Ditrich
has been researching and lecturing Sanskrit, Pali and a variety of academic
subjects related to Asian religions and languages at several universities in Eu-
rope and Australia. Her research areas include Buddhist studies (mainly in
Pāli), Sanskrit linguistics and Vedic philology.
Notes on Contributors xiii
Dylan Esler
is a scholar and translator of Tibetan Buddhist texts. He holds an MA in Bud-
dhist Studies from the School of Oriental and African Studies (SOAS), London,
and is currently a research scholar at the Institut Orientaliste of the Université
Catholique de Louvain (Belgium), where he is preparing a PhD. He is working
on a translation and study of an important tenth-century Tibetan text on the
subject of meditation. His research interest focuses on early rNying-ma exposi-
tions of rDzogs-chen and Tantra. He is also an associate member of the Ghent
Centre for Buddhist Studies, Ghent University.
Goran Kardaš
is Assistant Professor at the Department of Philosophy and the Department of
Indology and Far Eastern Studies, Faculty of Philosophy and Social Sciences,
University of Zagreb, Croatia. His research interests are Buddhist philosophy
(mainly Abhidharma and Madhyamaka), Nyāya epistemology and Indian the-
ories of meaning.
Eric M. Greene
is Assistant Professor of Religious Studies at Yale. His research focuses on the
early history of Buddhism in China, with a particular emphasis on the trans-
mission of Buddhism to China, early translation practices, and the history of
Buddhist meditation. His articles have appeared in T’oung Pao, Journal of Chi-
nese Religions, Artibus Asiae, and History of Religions.
Jowita Kramer
completed a doctorate (Hamburg, 2004) and habilitation (Munich, 2010) in In-
dology. The main focus of her research lies on Indian and Indo-Tibetan Bud-
dhism. She has held positions at the Universities of Oxford, Heidelberg and
Göttingen in the past and is currently a research fellow at the University of
Munich.
Chen-kuo Lin
is a Distinguished Professor in both the Department of Philosophy and the
Graduate Institute of Religious Studies at National Chengchi University. His
research interest includes Buddhist philosophy (Buddhist logic and epistemol-
ogy, Yogācāra, Mādhyamika), Chinese philosophy (Neo-Confucianism, Dao-
ism), and comparative philosophy. In addition to three books and many
articles, recently he has published (co-edited with Michael Radich) A Distant
Mirror: Articulating Indic Ideas in Sixth and Seventh Century Chinese Buddhism
(Hamburg University Press, 2014).
xiv Notes On Contributors
Andrea Schlosser
studied Indian philology, Indian art history, and culture sciences at the Free
University in Berlin. She received her Master (Magister) degree in 2008 with an
edition of a South Indian copperplate donation record of the Western Gaṅgas.
As a research assistant she focused on Buddhist manuscripts, working with the
Sanskrit fragments of the Berlin Turfan Collection in cooperation with the In-
ternational Dunhuang Project, as well as participating in a project about the
edition of Kharoṣṭhī manuscripts from the Bajaur Collection. For her disserta-
tion, submitted in 2013, she edited two Gāndhārī manuscripts that illustrate
the practices of a bodhisattva. Currently, she is a research associate in the
project “Buddhist Manuscripts from Gandhāra”, where her focus lies on early
Mahāyāna texts.
Ingo Strauch
is Professor for Sanskrit and Buddhist Studies at the University of Lausannne.
He received his PhD (2000) and his habilitation (2010) in Indian Philology at
Freie Universität Berlin. His current research focusses on early Buddhist manu-
scripts from Gandhāra. He is currently preparing an edition of Vinaya and
Sūtra texts from the Bajaur Collection of Kharoṣṭhī manuscripts. Together with
Andrea Schlosser (Munich) he is editing an early Mahāyāna sūtra from
Gandhāra. His research also covers early Buddhist epigraphy in Brāhmī and
Kharoṣṭhī scripts.
Weijen Teng
received his PhD in the Study of Religion from Harvard University. He is cur-
rently teaching at Dharma Drum Institute of Liberal Arts, Taiwan. His interests
of research include Chinese Buddhist intellectual history, the study of Sanskrit
grammar in Chinese Buddhism, and Pāli Abhidhamma theory of meditation.
Yao-ming Tsai
Professor of Philosophy at National Taiwan University and editor-in-chief of
Taiwan Journal of Buddhist Studies, received his Ph.D. in Buddhist Studies in
1997 from the University of California, Berkeley. His teaching and research fo-
cus on Buddhist philosophy as well as philosophy of life. He is the author of
four books: The Teachings of Prajñāpāramitā and the Purification of the Bud-
dha-field (2001), An Open Path for Constructing Buddhology (2006), Research
Methods and Academic Resources for Buddhist Studies (2006), Philosophy of Life
and Worldview from the Perspective of Buddhist Teachings (2012), as well as of
dozens of scholarly articles and book chapters on Buddhist studies.
Introduction
Introduction 1
Introduction
1 For an overview of scholarship on the meaning of the word ‘Abhidhamma / Abhidharma’: see
Malalasekera, 1961: 38–40; Willemen, Dessein & Cox, 1998: 13–15 and the contribution by Dylan
Eslar in this volume. For a recent study of the meaning of the term: see Anālayo, 2014: 70–71,
78–79, 171.
2 For discussions on the development of the Abhidharmic genre and the Abhidharmapiṭaka:
see Gethin, 1992 and Dessein, 2013. For a discussion of the development of the textual format
of the Abhidharma texts: see Cox, 1995: 29–37.
Judging from early Buddhist texts in which especially the Jains are referred
to,10 it appears that in the region of Magadha where the historical Buddha lived,
the Jains (and sometimes also the Ājīvikas) were important religious competi-
tors of the Buddhists.11 This competitive context gives Buddhism a peculiar
position in the Indian religious/philosophical landscape, and is important for
our understanding of the development of Abhidharma as exegetical method
– above, we have claimed that the argumentative skills that religions/philos-
ophies develop in a competitive context are important for them to become
conversant and cross-cultural.12 Indeed, because texts of the Vedic oral/aural
tradition were primarily aimed at delivering a message to the realm of the Vedic
gods, who, by definition, did not need to be convinced of the truth revealed in
the Vedic texts, these texts do not show a development of argumentative tech-
niques. In contradistinction to the Vedic texts, the Brāhmaṇa prose texts – as
do the Upaniṣads – explain the offers that are performed. This suggests that,
with the development of Brahmanism, the religious audience were no longer
the mere passive spectators they had been in the Vedic period, but became
– at least passive – participators in a religious dialogue. It therefore may be
the development of Brahmanism that, in the Indian context, laid the basis for
what eventually would become a tradition of rational inquiry.13 This transfor-
mation should not be overstressed though, as the debates that are included in
the Brāhmaṇas and the Upaniṣads are not ‘rational’ in the sense that they want
to convince someone of the Brahmanic or Upaniṣadic truth. Their explana-
tions are for internal use.
While on the one hand pointing to the Indian oral/aural tradition in which
Buddhism originated,14 the introductory formula to many Buddhist sūtras
“evaṃ me sutaṃ” (“Thus have I heard”) also alludes to it that, for the early
Buddhists too, the word spoken by the Buddha primarily had the value of
‘revealed truth’ and was the guideline to be followed.15 The following passage
included in the Mahāparinirvāṇasūtra testifies of this: When Ānanda lamented
that the doctrine would be without teacher after the Buddha had died – thus
urging the Buddha to appoint a successor – the Buddha is reported to have
replied that the doctrine and discipline he had taught them should also serve
as teacher after his demise.16 Also the following two viewpoints, attributed to
the Mahāsāṃghikas in the Samayabhedoparacanacakra (Yibuzong lun lun 異
部宗輪論) testify of this acceptance of the word of the Buddha as ‘revealed
truth’: “Everything that has been preached by the World-honored One is in
conformity with the truth (yathārtha),”17 and “The sūtras proclaimed by the
Buddha are all perfect in themselves (nītārtha)”.18 In his Yibuzong lun lun shuji
異部宗輪論述記, his commentary to the Chinese version of the Samayabhedo
paracanacakra, Kuiji 窺基 (632–682) explains these claims as that nothing the
Buddha has said is without benefit for living beings, and that, therefore, every-
thing he has said should be considered as the turning of the wheel of the
doctrine.19
Mentioning of the Samayabhedoparacanacakra (Treatise on the Development
of the Different Sects) brings us to one of the major peculiarities connected with
Abhidharma as exegetical method: the formation of Buddhist (Abhidharma)
sects. The territorial expansion of the Buddhist doctrine and the concomitant
different living conditions for Buddhist monks and nuns, must have made it
impossible for all monastics to live according to exactly the same set of rules
(Vinaya), and the Buddha’s choice not to appoint a successor must have invoked
different interpretations of the precise meaning of his sermons. Debates with
non-Buddhists must further have encroached on this tendency and must be
responsible for it that non-Buddhist – particularly Vedic and Jain – concepts
and elements were introduced in the Buddhist doctrine. All this eventually
resulted in the formation of different schools and sects.20 It is especially in the
Abhidhamma / Abhidharma literature that developed from the Sutta / Sūtra
and the Vinaya literature that these developments can be discerned.21
There is evidence that not only the formation of a corpus of sūtra texts, but
that also the Abhidharma practice originated as part of an oral/aural tradi-
tion.22 It has hereby been suggested that it are the mnemotechnical skills of
the Brahmans who joined the Buddhist order that made the memorization of
the Buddhist texts possible.23 Although we may indeed assume that the oral/
aural tradition of the early Buddhist community with synods (saṃgīti) meant
to recite the words of the Buddha, occurred in a context very similar to that of
the Brahmanic tradition – early saṃgha members claimed that they had main-
tained in their memories the words as they had been spoken by the Buddha
himself and had, generation after generation, been publicly recited – it is very
likely that, contrary to general acceptance, Buddhism did not start as a reac-
tionary movement against Vedic Brahmanism. There is evidence that
Brahmanism only started to reach the northeastern domains of the Indian sub-
continent during the Aśokan period. It are the territorial expansion and
enhanced possibilities for adherents of all beliefs and faiths to freely travel the
country that characterizes this period that must have brought Buddhists in
contact with Brahmins.24 This claim is supported by the following: tradition
connects king Aśoka with the so-called Buddhist ‘synod’ (saṃgīti) that resulted
in the first schism in the Buddhist community – the one between the
Mahāsāṃghikas and the Sthaviravādins.25
Contrary to what was the case for Brahmanism, however, Buddhists, as
adherents of a new faith, were quickly confronted with the fact that, for their
religion to survive, non-Buddhists had to be convinced of the Buddhist truth
– one may be born a Brahmin, however, one is not born a Buddhist.26 The
necessity to convert non-Buddhists helps to explain why early Buddhists culled
the most important elements of the Buddhist doctrine from the Buddha’s ser-
mons, and organized them in lists of technical elements, the so-called mātṛkās
that are the fundament of all later Abhidharmic developments.27 These
22 See T.22.1428: 968b15 ff. This oral recitation is connected to Ānanda, one of the first five
disciples of the Buddha, in Sumaṅgala (Rhys Davids and Carpenter, 1968, vol.1: 17),
Aṭṭhasālinī (Bapat and Vadekar, 1942: 3), Samantapāsādikā (Takakusu and Nagai, 1975: 18),
T.1.1: 1a9–10; T.49.2030: 14b8; T.22.1428: 968b25–26; T.24.1463: 818a28–29.
23 See von Hinüber, 1989: 68.
24 See Bronkhorst, 2001: 2–4 and 8–11.
25 This event is likely to have occurred 116 years after the demise of the Buddha. See Nattier
and Prebish, 1976–77: 239, 271–272.
26 It can also be recalled here that many of the first converts used to be adherents of Jainism.
27 See Bronkhorst, 2000: 77. The Sarvāstivāda Vinaya, Shi song lü 詩誦律, T.23.1435: 449a19 ff.
identifies the Abhidharma with teachings that are presented in list form. For arguments
that the first mātṛkās predate the ‘age of Abhidhama’ and that the beginnings of the
Abhidharma seem to go back to an early period before the formation of schools: see
Anālayo, 2014: 21–25; 167–168.
6 Introduction
28 See Norman, 1995: 309; von Simson, 1965: 142; Stache-Rosen, 1968: 8.
29 See Watanabe, 1983: 76. For the early appearance of the questions-and-answers format in
Buddhist literature: see Anālayo, 2014: 28; 167–168.
30 See Norman, 1992: 248 and Norman, 1993: 280.
31 The final version of this record was rendered into Pāli in the fifth century CE. See Finot,
1992: 7–9.
32 See Willemen, Dessein & Cox, 1998: 104–105.
Introduction 7
33 Beckwith, 2012: 24 remarks that “it is highly improbable that the Greeks and the early Bud-
dhists had no influence at all on each other”. See also Beckwith, 2012: 55–56.
34 See Salomon, 1999: 2.
35 MN I 55–63, DN II 290–315.
8 Introduction
36 This explains the occurrence of the term ‘Mātṛkāpiṭaka’. See T.50.2042: 113c2–9; T.50.2043:
152a15. That a separate authoritative collection of mātṛkās – a Mātṛkāpiṭaka – must have
existed prior to the moment the originally orally transmitted texts were submitted to writ-
ing, is also suggested in the Mahāsāṃghikavinaya. See T.22.1425: 334c20–22.
37 See Frauwallner, 1964. In his contribution to this volume, Bhikkhu KL Dhammojoti sug-
gests that it could even be said that, to some extent, the Theravāda Abhidhamma develop-
ment cannot be satisfactorily understood without at least a basic understanding of the
possible influence from the broad Sarvāstivāda lineage. Also the contribution by Lance S.
Cousins hints in this direction.
Introduction 9
in which this was done was Pāli. The use of Sanskrit in southern India is inter-
esting because it to all probability is Brahmanism that introduced the use of
Sanskrit in virtually the whole of South Asia and larger parts of Southeast Asia,
a development that is likely to be related to the fact that Brahmanism had
become a socio-political ideology, with Brahmins offering their services as
advisor to the political elites.41 With respect to the question of philosophical
interaction, the evidence provided in the sanne to the Visuddhimagga thus
opens a new perspective on the activities of the Brahmins, as well as of the
Theriya in South Asia in the later part of the first millennium CE.42
It is an important trait of Vasubandhu’s Abhidharmakośa that it criticizes
Vaibhāṣika viewpoints as they are expounded in the so-called vibhāṣā compen-
dia, the commentarial literature that was produced by the so-called Vaibhāṣika
Sarvāstivādins of Gandhāra. According to the Chinese monk and translator
Xuanzang 玄奘 (602–664), Saṃghabhadra (ca. fifth century CE) determined
to refute Vasubandhu’s critique on the Vaibhāṣikas.43 Bhikkhu KL Dhammajoti
discusses the importance of Saṃghabhadra in “Saṃghabhadra’s Contribution
to our Understanding of Abhidharma”. Kuiji, a disciple of Xuanzang, even
records and identifies Vaibhāṣika theories as “neo-Sarvāstivāda”.44 This makes
Saṃghabhadra, who is by Xuanzang’s pupils taken to represent the correct
Vaibhāṣika viewpoint, an important exponent of Abhidharma thinking. In this
way, Saṃghabhadra’s doctrinal expositions allow us insight in at least that sec-
tion of the Dārṣṭāntika Sarvāstivādins who developed into the Sautrāntikas.
The work of Saṃghabhadra is also important for the study of the doctrinal
evolution of the Mahāyāna, as his importance was acknowledged by the
Yogācāra masters (as also discussed in Jowita Kramer’s contribution) who took
him as representative of the true Vaibhāṣika viewpoint in their Sarvāstivāda
studies.
It is through the intellectual context of Greater Gandhāra that both
Śrāvakayāna and Mahāyāna Buddhism reached China around the beginning of
the common era. While the Chinese Buddhist monastic code developed to be
essentially the Dharmaguptaka code, the doctrinal school that had the largest
impact in China was without any doubt the highly rationalistic Sarvāstivāda
45 See Shiji 史記, 63.123: 345a2–11 and Han Shu 漢書, 61: 250b35-c1.
46 See Gernet, 1990: 188.
47 See Lloyd and Sivin, 2002: 27.
48 See Lloyd and Sivin, 2002: 73. Cheng, 1997: 318, note 4 remarks: “Concernant la notion
d’écoles dans la Chine ancienne, Nathan Sivin (cf. Philosophy East and West, 42, 1, 1992, 27)
remarque que, contrairement à la conception grecque de l’école formée d’orateurs et de
polémistes sur la place publique, elle correspond bien plus à des classifications bibli-
ographiques qu’à des groupements de personnes. En Chine, les écoles se distinguaient
entre elles en ce qu’elles préservaient et transmettaient des corpus différents de textes
12 Introduction
écrits, dans une lignée de transmission qui ressemblait fort à une filiation (d’où le mot jia
qui désigne le clan)”.
49 Bauer, 2006: 17.
50 See Seiwert, 1994: 532.
51 See Bronkhorst, 2011: 236–237. For some concrete examples: see Forte, 2000: 9–10, 51; Des-
sein, 2003: 329–332.
Introduction 13
52 On the history of the Chinese Abhidharma and Kośa Schools: see Dessein, 2010: 57–62,
66–67.
14 Introduction
positions and yet also refute, point by point, positions thought to represent
rival groups”.53 In this process, the original mātṛkās were not only adapted
to suit doctrinal interpretations and developments, but they also became
the force of doctrinal development themselves.54 New interpretations of the
transmitted material were also included in commentarial texts on earlier
Abhidharma texts, and sectarian self-identification also culminated in the cre-
ation of well-structured pedagogical manuals of Buddhist doctrine in which,
through sophisticated methods of argument, the own position is established
and the view of others is refuted.55 This development of peculiar philosophical
positions is the subject of the third part of this volume: “Philosophical Studies”.
Harivarman lived in the third-fourth century and is likely to have belonged to
the Bahuśrutīyas, a sub-group of the Mahāsāṃghikas.56 He is credibly attested
as being a pupil of Kumāralāta who, himself, was connected to the Sautrāntika
development of the Sarvāstivādins.57 By the time of Harivarman, the early
Buddhist dharma theory – the Sarvāstivādin atomic world view that all that
exists is composed of phenomena (dharma) that are themselves subject to
constant change, and that therefore all that exists is devoid of intrinsic nature
(svabhāva) – had experienced a major reinterpretation. Rational inquiry could
not but lead to the advanced opinion that when all that exists is composed
of phenomena, it should necessarily be so that these phenomena (themselves
dharmas too) should equally consist of further dharmas. The only logical out-
come was that all that exists was taken to be empty (śūnya).58 It especially
were the Madhyamaka philosophers, the most important of whom are the sec-
ond-third century Nāgārjuna and his disciple Āryadeva, who took this thinking
to its logical and methodological extreme. As masters of logic argumenta-
tion, they even dared to use logical argument to discuss the proper dogmas of
Buddhism. “Madhyamaka in Abhidharma Śāstras: The Case of Harivarman’s
*Tattvasiddhi” is Goran Kardaš’s investigation of the Madhyamaka view-
points in Harivarman’s *Tattvasiddhiśāstra. In this Abhidharma treatise that
is one of the very few to refer to the Mādhyamikas, Harivarman is seen to put
Madhyamaka argumentation – that most likely stems from Āryadeva – in favor
of the non-existence of things in the mouth of his opponent who, through this,
appears as a Madhyamaka adherent. This contribution thus witnesses of the
Thus far, we have signified the very nature of the Abhidharma tradition as an
enterprise of rational thinking, sketched its historical transmission and trans-
formation across South and East Asia, and outlined the arrangement and the
main points of the chapters. The following part of the introduction highlights
the methodological significance of this edited volume as a whole, and teases
out some of the methodological features reflected in the individual chapters.
The compilation of this volume, which covers cross-regional and longue
durée studies of Abhidhamma / Abhidharma traditions, appears to encapsu-
late a methodological intuition – a demonstration of how the Abhidharma
study, which is often envisioned as a lofty, hypertechnical, and relatively mar-
ginal subject within Buddhist Studies, could be approached in a way that could
contribute to a broader inquiry into the Humanities. Given the intention of
offering some new lights to the study of the Abhidharma tradition method-
ologically, this volume does not claim to offer a methodological ‘paradigm
shift’. As the three thematic divisions of this volume suggests, the authors have
adopted the methods that are fundamental to the study of Abhidharma study,
namely a meticulous textual criticism and historical philology, contextualized
intellectual history, and doctrinal philosophy. Nonetheless, in their studies,
these authors take up issues that go beyond the Abhidharma tradition itself
and therefore reveal verily some methodological significances and potentials.
We will point out these methodological features shortly, but before that, we
62 For example, he proposes that “conflict is the energy source of intellectual life, and con-
flict is limited by itself”. Collins, 1998: 1.
63 For a general discussion of the term, see Pym, 2010, Chapter 8 “Cultural Translation”.
Introduction 19
64 Kramer, 2014.
20 Introduction
Abbreviations
65 For a previous discussion on this issue of “normative discourse”, see Cabezón, 1995: 231–
268.
Introduction 21
Bibliography
Cousins, Lance S. (1982). “Pali Oral Literature”, In Buddhist Studies – Ancient and Modern
(Collected Papers on South Asia No. 4). Edited by Philip Denwood and Alexander
Piatigorsky. London: Curzon Press, pp. 1–11.
Cox, Collett (1995). Disputed Dharmas. Early Buddhist Theories on Existence. An Annotated
Translation of the Section on Factors Dissociated from Thought from Saṅghabhadra’s
Nyāyānusāra. (Studia Philologica Buddhica. Monograph Series XI). Tokyo: The
International Institute for Buddhist Studies.
Dessein, Bart (2001). “Climbing a tree to catch fish: Some reflections on Plato, Aristotle,
and China”. La rationalité en Asie / Rationality in Asia. Études de Lettres 2001/3,
pp. 97–125.
Dessein, Bart (2003). “The Glow of the Vow of the Teacher Samantabhadra ‘Puxian pusa
xing yuanzan’ (T.297) *Samantabhadrācāryapraṇidhānarāja”, Acta Orientalia 56/2–4,
pp. 317–338.
Dessein, Bart (2007). “The First Turning of the Wheel of the Doctrine – Sarvāstivāda and
Mahāsāṃghika Controversy”, In The Spread of Buddhism. (Handbuch der Orientalistik.
8. 16). Edited by Ann Heirman and Stephan Peter Bumbacher, pp. 15–48.
Dessein, Bart (2009). “The Mahāsāṃghikas and the Origin of Mahayana Buddhism:
Evidence Provided in the *Abhidharmamahāvibhāṣāśāstra”, The Eastern Buddhist
New Series 40/1–2, pp. 25–61.
Dessein, Bart (2010). “The Abhidharma School in China and the Chinese Version of
Upaśānta’s Abhidharmahṛdayasūtra”, The Eastern Buddhist, New Series 41/2,
pp. 49- 69.
Dessein, Bart (2012). “‘Thus Have I Heard’ and Other Claims to Authenticity: Development
of Rhetorical Devices in the Sarvāstivāda Ṣaṭpādābhidharma Texts”, In Zen Buddhist
Rhetoric in China, Korea, and Japan. Edited by Christoph Anderl. Leiden: Brill,
pp. 121–162.
Dessein, Bart (2013) “Lists as Dynamic Devices in Early Buddhist Doctrine and Textual
Tradition”, Antiquorum Philosophia. An International Journal 7, pp. 29–48. (Special
Issue: Forme del catalogo).
Finot, Louis (1992). Les Questions de Milinda. Paris: Gaillimard.
Forte, Antonino (2000). A Jewel in Indra’s Net. (Italian School of East Asian Studies.
Occasional Papers 8). Kyoto: Istituto Italiano di Cultura. Scuola di Studi sull’Asia
Orientale.
Frankenhauser, Uwe (1996). Die Einführung der buddhistischen Logic in China. Opera
Sinologica 1. Wiesbaden: Harrassowitz.
Frauwallner, Erich (1956). The Earliest Vinaya and the Beginnings of Buddhist Literature.
(Serie Orientale Roma 8). Roma: Istituto Italiano per il Medio ed Estremo Oriente.
Frauwallner, Erich (1964). “Abhidharma Studien”, Wiener Zeitschrift für die Kunde Süd-
und Ost-Asiens 8, pp. 59–99.
Garbe, Richard (1987). The Philosophy of Ancient India. Chicago IL: The Open Court.
Introduction 23
Norman, Kenneth R. (1992) “The Development of Writing in India and its Effect upon
the Pāli Canon”, Wiener Zeitschrift für die Kunde Südasiens und Archiv für indische
Philosophie 36, pp. 239–249.
Norman, Kenneth R. (1993). “Der Beginn der Schrift und frühe Schriftlichkeit in Indien.
By Oskar v. Hinüber. (Abhandlungen der Geistes- und Sozialwissenschaftlichen
Klasse; Jahrgang 1989, Nr.11) pp. 75. Mainz, Akademie der Wissenschaften und der
Literatur; Stuttgart, Franz Steiner Verlag GmbH, 1990)”, Journal of the Royal Asiatic
Society (3rd series), vol.3, part 2, pp. 277–281.
Norman, Kenneth R. (1995). “Oskar von Hinüber: Untersuchungen zur Mündlichkeit
früher mittelindischer Texte der Buddhisten. Untersuchungen zur Sprachgeschichte
und Handschriften-Kunde des Pali III. (Abhandlungen der Geistes- und Sozial
wissenschaftlichen Klasse, Jahrgang 1994, No. 5). Mainz/Stuttgart (Akademie Der
Wissenschaften und der Literatur, Mainz. Franz Steiner Verlag), 1994, 45 pp.”, Acta
Orientalia LVI, pp. 302–311.
Potter, Karl H. (1999). Encyclopedia of Indian Philosophies. Volume VIII: Buddhist
Philosophy from 100 to 350 ad. Delhi: Motilal Banarsidass.
Przyluski, Jean (1932). “L’influence iranienne en Grèce et dans l’Inde”, Revue de l’Univer-
sité de Bruxelles XXXVII, pp. 283–294.
Pym, Anthony (2010). Exploring Translation Theories. London and New York NY:
Routledge.
Rhys Davids, Thomas W. and J. Estlin Carpenter (eds.) (1968). The Sumaṅgala-Vilāsinī,
Buddhaghosa’s Commentary on the Dīgha Nikāya. 3 volumes. Pali Text Society.
London: Luzac & Co.
Rhys Davids, Thomas W. and Carpenter, Joseph E. (ed.) ([1890–1911] 1995–2007).
Dīghanikāya. 3 vols. London: The Pāli Text Society.
Salomon, Richard (1999). Ancient Buddhist Scrolls from Gandhāra. The British Library
Kharoṣṭhī Fragments. London: The British Library.
Scharfstein, Ben-Ami (1989). The Dilemma of Context. New York NY and London: New
York University Press.
Seiwert, Hubert (1994). “Orthodoxie, Orthographie und Zivilreligion im vorneuzeitli-
chen China”, In Gnosisforschung und Religionsgeschichte. Festschrift für Kurt Rudolph
zum 65. Geburtstag. Edited by Holger Preißler, Hubert Seiwert and Heinz Mürmel.
Marburg: Diagonal-Verlag, pp. 529–541.
Shiji 史記, Ershiwu shi 二十五史 (1988). Shanghai: Shanghai guji chubanshe.
Simson, Georg von (1965). Zur Diktion einiger Lehrtexte des buddhistischen Sanskritkanons.
München: J. Kitzinger.
Staal, Frits (1989). “The Independence of Rationality from Literacy”, European Journal
of Sociology 30, pp. 301–310.
Staal, Frits (1993). Concepts of Science in Europe and Asia. Leiden: International Institute
for Asian Studies.
Introduction 25
Watanabe, Fumimaro (1983). Philosophy and its Development in the Nikāyas and
Abhidharmma. Delhi: Motilal Banarsidass.
Willemen, Charles (2006). The Essence of Scholasticism: Abhidharmahṛdaya. T1550.
Delhi: Motilal Banarsidass.
Willemen, Charles; Dessein, Bart & Cox, Collett (1998). Sarvāstivāda Buddhist
Scholasticism. (Handbook of Oriental Studies 2/11). Leiden etc.: Brill.
Zürcher, Erik ([1959] 1972). The Buddhist Conquest of China. The Spread and Adaptation
of Buddhism in Early Medieval China. 2 vols. Leiden: E.J. Brill.
27
Part 1
Mātṛkā and Abhidharma Terminologies
∵
28
Abhidharma and Indian thinking 29
Chapter 1
1 Introduction
Much could be said, and has been said, about the influence of Buddhism on
other currents of thought in India. This chapter will deal with this same topic,
but limit its attention to one specific form of Buddhism: Abhidharma. And
even here, it will only take into consideration the form of Abhidharma that was
created in the northwestern corner of the Indian subcontinent, in an area that
nowadays belongs to three different political entities: Pakistan, Afghanistan
and Kashmir. This form of Abhidharma is primarily, but not exclusively, associ-
ated with the Sarvāstivādins, and the region in which it arose is often referred
to as “Greater Gandhāra,” a somewhat vague term, to be sure, but appropriately
so, for it is hard to determine with certainty which regions can legitimately be
included in it, and which cannot.1 Kashmir and Gandhāra may be taken to
belong to it, and if we add Bactria, we can call it, with David Gordon White
(2012), KGB.
The Abhidharma of Greater Gandhāra (I will call it Gandhāran Abhidharma)
exerted a major influence on Buddhist thought in India, to be sure. Most, if not
all, of the philosophical schools of Indian Buddhism – both Main Stream and
Mahāyāna – are based on the foundations laid here. I have argued elsewhere
that already the very earliest Mahāyāna texts we possess are based on
Abhidharma thought; these texts may date from the first century BCE.2
* This paper brings together a number of observations made and conclusions drawn in other
publications, duly referred to in the footnotes, along with new observations. In its present
form it has profited from various critical remarks, most notably by Collett Cox and Shoryū
Katsura. Bronkhorst, 1996 overlaps to some extent with parts of this paper, but concentrates
on Buddhist notions of language and their influence on Brahmanical philosophy.
1 A Buddhist presence in this region from at least the 2nd century BCE seems certain.
Cp. Behrendt, 2004: 256: “[Phase I] began with the founding of the earliest Buddhist centers
in Greater Gandhāra: Butkara I and the Dharmarājikā stūpa in Taxila […]. An early 2nd century
bce date seems a conservative benchmark for the beginning of this period.” On the extent of
Greater Gandhāra, see also note 6, below.
2 Bronkhorst, 2013.
3 It is true that this question is rarely, if ever, explicitly formulated in the technical texts of early
Abhidharma. However, this aspect is emphasized, presumably at least from the time of King
Menander onward, in texts that draw the consequences of this form of Abhidharma, includ-
ing the Milindapañha and Prajñāpāramitā texts right from the beginning. See Bronkhorst
2013.
Abhidharma and Indian thinking 31
This, then, was the first step taken by the Gandhāran Ābhidharmikas. They
introduced an atomic vision of reality, in the sense that they maintained that
there are ultimate constituent elements of all that exists, and that these ulti-
mate elements are the dharmas which they had collected in lists. But they did
not stop there. The next step was the claim that dharmas are the only things
that exist. The objects that they constitute – this covers all there is, including
the macroscopic objects of our everyday experience – do not exist. And the list
of dharmas is an exhaustive enumeration of all that exists.
This second step can easily be understood in the light of the persistent
Buddhist tradition that the self, or the person, does not exist. If the person is
conceived of as the aggregate of all its constituent dharmas (and this became
the generally accepted way in Buddhism to think about it), this can easily be
interpreted to mean that the person does not exist because it is an accumula-
tion of dharmas. Other accumulations of dharmas therefore do not exist either.
The Gandhāran Buddhist scholiasts found themselves in this way in the pos-
session of an exhaustive enumeration of all there is, viz., the inherited and
slightly adjusted list of dharmas. These dharmas were now thought of as
momentary, i.e. as each lasting no more than one single moment. Once again,
a traditional Buddhist doctrine could be invoked in defense of this new view:
the Buddha had taught that all conditioned factors are non-eternal. This could
easily be interpreted to mean that all dharmas are momentary, and this is what
happened.4
Gandhāran Abhidharma reduced in this manner the whole world, both ani-
mate and inanimate, to an uninterrupted sequence of some seventy-five
momentary dharmas; more precisely, some seventy-two momentary dharmas,
plus three eternal ones, the so-called asaṃskṛta ‘non-conditioned’ dharmas.
The world of our experience thus turned out to be ultimately unreal, with the
real world of momentary dharmas hidden below it.
How could we possibly be misled into believing that we live in a world of
persons and other macroscopic objects? The Buddhist texts frequently respond
by pointing out that this or that macroscopic object is a mere name and does
not really exist. This is what, in the Milindapañha, Nāgasena pointed out with
regard to King Milinda’s chariot, and numerous other Buddhist texts state the
same. The objects we are familiar with in our everyday world owe their relative
existence – or rather our mistaken conviction that they exist – to the words of
language.
There is one more point I wish to add to this brief characterization of
Gandhāran Abhidharma. Words, as we have seen, are responsible for our
4 On the beginning of Buddhist momentariness, see Rospatt, 1995, along with Bronkhorst, 1995.
32 Bronkhorst
mistaken belief in the reality of a world of persons, chariots and much else. But
words, one might object, are themselves ultimately non-existing entities. Does
this not undermine the system in some vital manner?
It is possible that the Gandhāran scholiasts were aware of this objection. All
we know for certain is that they introduced some dharmas in their list that we
might call linguistic dharmas. There are three of them, originally perhaps only
two. They correspond to individual speech sounds, words and sentences. If we
concentrate on words, this means that, beside the sequence of speech sounds
that make up a word, there is a momentary dharma (or perhaps better: a series
of identical momentary dharmas) that are the word. The word exists in this
way beside, and independently of, the speech sounds. This is a highly remark-
able conception, and the only justification for its existence I can think of is to
save the reality, the real existence, of words and with it the fundamental coher-
ence of the Abhidharma system that was being developed.
2 Grammar
It is time to leave the details of the Abhidharma system and turn to the schools
of non-Buddhist thought on which it exerted an influence. The first to be con-
sidered is the tradition of Pāṇinian Sanskrit grammar, and more in particular
one specific text that belongs to it, Patañjali’s Mahābhāṣya “Great Commentary”.
It is known that Patañjali wrote this work during the decades following the
middle of the second century BCE, and that he was at some time in the service
of King Puṣyamitra, one of the successor kings of the Mauryan empire. He was
already alive during the inroads of Indo-Greeks into northern India, around
150 BCE. There is some evidence to believe that he settled in Kashmir, and this
may explain his interest for us at present.5 For Kashmir was on the one hand a
part of Greater Gandhāra,6 with a substantial Buddhist presence. On the other
hand, Kashmir was conquered by Puṣyamitra, a Brahmanical ruler who, as we
saw, had scholars like Patañjali in his service. Kashmir may therefore have been
the place par excellence for Buddhist philosophy, i.e. Gandhāran Abhidharma,
and Brahmanical scholarship to meet. Judging by certain features of Patañjali’s
Mahābhāṣya, this is what actually happened. Let me explain.
Gandhāran Abhidharma was preoccupied with ontology; this we have seen.
There are reasons to think that the Buddhists who created it were the first to
ceding dharma, not by dharmas that occurred earlier and that have gone long
since, nor indeed by dharmas that are still to appear.8
A grammatical derivation, too, is a succession of stages. What determines
each next stage? For Pāṇini this could be any of the preceding stages, or occa-
sionally even a stage that was still to come. Patañjali changed all this. For him,
each next stage was completely determined by the immediately preceding
one. He imposes in this way a linear scheme that is in all essential respects
parallel to the linearity of Gandharan Buddhism. This imposition confronted
him with major difficulties, for Pāṇini had never intended anything of the kind.
Patañjali was therefore obliged to introduce new procedures, add metarules
and use various other tricks to make derivations conform to his vision.
But this vision was extraneous to Pāṇini’s grammar. Patañjali imposed it,
without telling us why. However, it seems safe to assume that the explanation
of this strange change imposed by Patañjali lies in his acquaintance with the
Abhidharma vision of the world.9
3 Jainism
One of the old texts of the Śvetāmbara Jaina canon, the Sūyagaḍa (in Sanskrit:
Sūtrakṛtāṅga), shows awareness of a number of characteristics of Buddhism. It
mentions the five skandhas, but also – and this is more interesting in the pres-
ent context – the notion of momentariness. Momentariness, as we have seen,
is one of the innovations introduced by Gandhāran Abhidharma. We may
assume that the author of this text (or of this portion of the text) was acquainted
with the developments that had taken place in north-western Buddhism.
Other texts of the Śvetāmbara canon – all of them no doubt younger than
the Sūyagaḍa – do not only know the notion of momentariness: they have
adopted it themselves. The moment (samaya) as the smallest unit of time
appears to occur for the first time in the Uttarajjhayaṇa, and this same text
further knows the notion of santati, the sequence of moments that is also com-
mon in Buddhism. Beside moments, other significant notions are found in the
Uttarajjhayaṇa, among them pradeśa (the smallest unit of space) and
paramāṇu (atom). In other words, we find here an atomic vision of time, space
and matter. It is true that Buddhism does not appear to have accepted the
8 The justification for this way of viewing the succession of dharmas lies in the Buddhist doc-
trine of dependent origination (pratītyasamutpāda).
9 For details, see Bronkhorst, 2004. See further Bronkhorst, 1994; 2002b.
Abhidharma and Indian thinking 35
atomic nature of space, it did accept and elaborate the atomic nature of mat-
ter, and of course of time.
The Sūyagaḍa is also acquainted with the Buddhist notion of the person as
a collection of skandhas. It tells us that the Buddhist person is neither different
nor not different from the skandhas; this is a position that was held by the
Buddhist Pudgalavādins. The Sūyagaḍa, we learn from this passage, was
already aware of the issue regarding wholes and their parts that occupied the
Buddhists. Most Buddhists, with the exception of the Pudgalavādins, rejected
the existence of wholes; subsequent texts from the Śvetāmbara canon accept
it.
Perhaps the most surprising feature that Jainism took over from Abhidharma
Buddhism, if my reflections are correct, is the use of the word pudgala. In
Buddhism the pudgala is the person, conceived of as the combination of the
skandhas, i.e. of the dharmas, that make up a person. Many Buddhists thought
that this conception does not correspond to any reality; the Pudgalavādins
were of a different opinion, and specified that the pudgala is neither different
nor not different from the skandhas. All this we know.
Interestingly, Jainism too adopted this word pudgala. For them, it does not
refer to the person, but rather to material objects. A closer study of the relevant
passages brings to light that one of the earliest occurrences of this word in the
Jaina canon does refer to a person. In subsequent Jaina developments the
emphasis is more and more on the aggregate. We know that in Buddhism pud-
gala refers to an aggregate, but only to one special kind of aggregate: the
person. In Jainism it comes to include other aggregates as well: early texts use
it in the sense of “portion, quantity”. It appears indeed that within Jainism the
meaning of this word developed from “person” (the meaning also used in
Buddhism) to “material object”. This development only makes sense if we start
from the Buddhist notion of the person as an aggregate.
It is not possible to enter more deeply into this discussion. Further reflec-
tions and textual references can be found in another publication, to which I
must refer for details.10 Here a few words must be said about the time and
place of the interaction between Gandhāran Abhidharma and Jainism. Jainism
was not present in Gandhāra.11 Gandhāra is beyond the lands where a Jaina
monk is allowed to travel, which extend westward until Thaneshwar.12 However,
10 Bronkhorst, 2000.
11 See however Pal, 2007, which shows that Jaina merchants may have ventured into
Gandhāra and Afghanistan.
12 See Jain, 1984: 23–24 (with note 2), 337 ff.; further PPN II s.v. Saṃpai (Samprati), with refer-
ences. Kalpasūtra 1.50 reads: “monks and nuns may wander eastward as far as
36 Bronkhorst
Jainism was strongly present in Mathurā under the Kuṣāṇas, i.e. at a time when
Buddhism, too, had a strong presence in that region. There are also reasons to
think that Buddhism and Jainism interacted in other respects in this region
during this period. It seems, for example, that Jainism at that time abandoned
the worship of relics of the Jina and of stūpas that contained them. This would
not be an example of Jainism borrowing from Buddhism but rather of Jainism
consciously differentiating itself from Buddhism. Once again, there is no space
to pursue this issue further at present.13
4 Brahmanical Philosophy
5 Influence Elsewhere
It cannot be the purpose of this chapter to give a complete survey of the ways
in which Gandhāran Abhidharma has influenced non-Buddhist traditions in
India. Some of its notions, such as that of momentariness and that of the per-
son as an aggregate, appear unexpectedly in texts such as the medical Caraka
Saṃhitā.16 Instances could no doubt be multiplied.
Here I will mention one more feature of Gandhāran Abhidharma that was
enthusiastically taken over by certain Brahmanical thinkers. Recall that in this
philosophy the world of our experience is ultimately unreal. This became a
central element of Buddhist systematic philosophy, one that distinguished it
for a long time from the Brahmanical systematic philosophies with which it
coexisted.17 This changed around the middle of the first millennium CE, when
philosophies like Advaita Vedānta joined the inter-philosophical debate. One
of its most important early thinkers was Śaṅkara, and it is not surprising that
some of his contemporaries accused him of being a crypto-Buddhist:18
Śaṅkara’s philosophy had adopted the essentially Buddhist notion that the
world of our experience is not ultimately real. There is no need to recall that
Advaita Vedānta became in due time India’s most popular philosophy, and
Śaṅkara its most famous thinker.
There is one more case of Abhidharma influence that deserves to be men-
tioned. However, for reasons that will become clear, it makes most sense to
deal with this case after a brief discussion of the origin of Gandhāran
Abhidharma. That is therefore the topic to which I will turn now.
It should be clear from what has been mentioned so far that Gandhāran
Abhidharma has been extremely influential in the subsequent development of
Indian thought. Not only did it to a large extent determine the shape of subse-
quent Buddhist philosophy in all of its forms. Non-Buddhist philosophies and
other forms of Indian thought, too, were profoundly influenced by it. I think
therefore that we can state without hesitation that Indian philosophy – or at
any rate systematic, rational philosophy in South Asia – began in Greater
Gandhāra.
How did this happen? Hard-headed philologists rarely ask this question.
They often feel that it is difficult enough to extract from the texts what their
authors thought, and that those same texts rarely, if ever, tell us why these
authors thought the way they did. There are some exceptions. T.R.V. Murti –
whose book The Central Philosophy of Buddhism (1960) was once widely
admired, but whose influence seems to be beyond its peak – emphasized the
importance of intellectual strife and the resolution of contradictions in the
development of new ideas. According to him, “the Ābhidharmika schools […]
grew as the rejection of the ātmavāda of the Brāhmanical systems”.19 This can-
not be right. To the best of our knowledge there were no, or few, Brahmins in
Greater Gandhāra at that time.20 Gandhāran Abhidharma cannot therefore
have been based on a rejection of the Brahmanical ātmavāda.
I have elsewhere dealt with Murti’s ideas in general, which I do not accept.21
However, it does seem right to assume that revolutionary new ideas tend to
arise in appropriate surroundings, especially in challenging intellectual sur-
roundings. It does therefore make sense to ask, if there were no Brahmins in
19 Murti, 1960: 8.
20 Bronkhorst, 2011: 202–205; 2016: § I.1.3.
21 Bronkhorst, 2006.
Abhidharma and Indian thinking 39
Greater Gandhāra at that time, who else may have been there to challenge the
Buddhists into creating a vision of the world that was to inspire Indian think-
ers, directly or indirectly, for the next two thousand years.
The answer to this question seems simple and straightforward. Gandhāra
and most of what we call Greater Gandhāra (with the exception of Kashmir,
see above) fell again in the hands of the Indo-Greeks around 185 BCE, at the
time of the collapse of the Mauryan empire. At that time, and perhaps already
before that time, the Buddhists of that part of the subcontinent had to deal
with Hellenistic rulers. And we know that Hellenistic rulers had a tendency to
surround themselves with philosophically cultivated sages, and that discus-
sions of a philosophical nature had become part of Hellenistic tradition. It is
hard to imagine that representatives of the clearly numerous Buddhists who
inhabited those regions were not sometimes challenged to take part in such
discussions. However, to take part in a sophisticated discussion, you better
present, and represent, a coherent position yourself. Presumably the Buddhist
scholiasts of Gandhāra realized this, and the result is that they created, for the
first time in Buddhist history, a coherent philosophical ontology: Gandhāran
Abhidharma was born.
I realize that these few words about the possible origin of Gandhāran
Abhidharma do not but scratch the surface of an important and complex his-
torical question. I have tried to do it more justice in some other publications,22
and I remain aware that the question has not been fully explored even there.
Others may take it up and, who knows, they may come to different conclusions.
However, I do wish to emphasize that, if it is true that Indian systematic philos-
ophy began in the north-western region we call Greater Gandhāra, reflections
as to how this happened, and why, cannot be ignored.
After these summary reflections about the origin of Gandhāran Abhidharma,
we are ready to address a theory recently launched by Christopher Beckwith in
his book Warriors of the Cloisters: The Central Asian Origins of Science in the
Medieval World (2012). Beckwith’s theory concerns what he calls the recursive
argument method that, he claims, came to be used in Indian philosophical
texts, but not only there. The recursive argument method, according to
Beckwith, ended up being used in medieval Europe which borrowed it from
the Islamic world. The Islamic world itself, still according to Beckwith, took it
from Buddhism in Central Asia. This recursive argument method supposedly
played a crucial role in the development of European science. This is why the
The earliest text so far identified that uses a primitive version of the
method, and indeed uses it throughout the text, is the Central Asian
Aṣṭagrantha [or Aṣṭaskandha; Taishō 1543]. In this work, each topic argu-
ment is followed by a list of arguments about it – usually a rather long list
– and then they are repeated and disputed, one by one, in order.24 By
contrast, the later Jñānaprasthāna [Taishō 1544] does not use the recur-
sive argument method at all. It strictly follows the two-part Question:
Answer format.
The Vibhāṣā, a scholastic work of the Bactrian-Gandhāran branch of
the Sarvāstivāda school dated possibly to the first century AD, during the
Kushan Empire, contains the earliest known example of what eventually
became the fully developed recursive argument method. The method
apparently thus developed specifically within the Bactrian-Gandhāran
branch of the Sarvāstivāda school, and was only later partially adopted by
the Kashmiri Vaibhāṣika sect of Sarvāstivāda.
Examples of the recursive argument do not occur in earlier Buddhist
texts, earlier non-Buddhist Indian texts, or earlier texts connected to
other branches of the Sarvāstivāda school.
23 Beckwith’s theory does not only concern the recursive argument method, but also the
development of medieval European colleges under the influence of Islamic madrasas,
which themselves presumably arose under the influence of Buddhist vihāras. Beckwith
claims, with a reference to Dutt, 1962: 62ff. and 211 ff., that “the plan of the vihāra is strik-
ingly different from that of the saṅghārāma, the typical earlier, strictly Indian, Buddhist
monastic design” and he adds that “[t]he vihāra design is […] a specifically Central Asian
innovation developed under the Kushans and spread by them” (p. 41). These claims, and
especially the second one, are not substantiated but cannot here be further examined; see
however Schopen, 2004: 73–80; 2006.
24 Beckwith refers here to Willemen, Dessein & Cox, 1998: 223, where we read: “the two Chi-
nese translations of the *Aṣṭaskandhaśāstra (30 fascicles) and Jñānaprasthāna (20 fasci-
cles) do differ in length, at least in part, as a result of a difference in format: the
*Aṣṭaskandhaśāstra lists the questions that will be addressed at the beginning of each
section and then repeats the questions with each answer; the Jñānaprasthāna gives the
questions only once prior to the answer.”
Abhidharma and Indian thinking 41
This passage raises a number of questions, among them the following: Does
Beckwith not overemphasize the role of Central Asia? Is the Aṣṭagrantha (or
Aṣṭaskandha) really a Central Asian text?25 And is the Vibhāṣā really a scholas-
tic work of the Bactrian-Gandhāran branch of the Sarvāstivāda school?26 At
present we can leave these questions aside, and reformulate Beckwith’s theory
slightly, so that it now states that the recursive argument method originated in
Abhidharma Buddhism. Put this way, Beckwith’s claim is that the origins of
science in the Medieval European world have to be looked for in Gandhāran
Abhidharma.
What is this recursive argument method? Beckwith dedicates a chapter
(Chapter Two) of his book to explaining and illustrating what he means by this.
Here I will merely repeat his statement of its essence, which occurs on page 89
and reads:27
Suppose now, for argument’s sake, that Beckwith is right in maintaining that
examples of the recursive argument do not occur in earlier texts, whether
25 Beckwith refers to Willemen, but all Willemen says is, 2006: 6: “Kātyāyana’s Aṣṭagrantha
[…], the main text of the Gandhārans, [was] probably written in Gāndhārī and Kharoṣṭhī,
in the late first century BCE. In the second century CE this text was rewritten in Sanskrit,
and called Jñānaprasthāna.” (my emphasis, JB). See also Willemen, 2012: 163–164: “The
Sarvāstivāda ‘orthodoxy’ rewrote the old Gandharan Aṣta-grantha in Sanskrit, now called
Jñāna-prasthāna. Because the old text had many Vibhāṣās, commentaries, the new text
needed a new commentary. This is the Mahāvibhāṣā.”
26 On p. 59, Beckwith specifies that “[t]he earliest vibhāṣā preserved (in Chinese translation)
[…] is the Central Asian work known as the Vibhāṣā [Taishō 1547]”. And p. 62: “the Bac-
trian Aṣṭagrantha and Vibhāṣā are the models for the later Jñānaprasthāna and
Mahāvibhāṣā.” Contrast this with Cox’s remark: “It is with the composition of the vibhāṣā
compendia that the Sarvāstivāda school within Kaśmīra comes to be defined both doc-
trinally and textually.” (Willemen, Dessein & Cox, 1998: 229). See also Willemen, 2006:
6–7: “The western [i.e. non-Kashmirian, JB] Sarvāstivādins, a very heterogeneous group,
seem to have had more than one Vibhāṣā on the Aṣṭagrantha.” Beckwith’s concern with
Central Asia finds expression in his earlier book Empires of the Silk Road: A History of
Central Eurasia from the Bronze Age to the Present (2009).
27 See also pp. 25–26: “The recursive argument is, minimally, an argument that is disputed
by an argument that is disputed by an argument, or more simply (but in reverse order), an
argument about an argument about an argument.”
42 Bronkhorst
28 Beckwith emphasizes the difference between structure and content, but believes that the
former can influence the latter (2012: 35–36): “The overt, explicit, formal structure of the
recursive argument is its most crucial factor. It is not quite true that ‘the medium is the
message’ in recursive method books, but because they typically consist exclusively of lists
of recursive arguments, each of which contains many contrasting views on the same
problem, they clearly did encourage scepticism and speculation by the authors. In that
respect, therefore, it is true that the form of the recursive argument did have a significant
indirect impact on the content of works written according to it. Nevertheless, it must be
stressed that the specific overt structure, per se, of a recursive argument is not directly or
even implicitly connected, structurally or semantically, to its specific overt content or to
the implicit logical structure of the internal content. In other words, in a recursive argu-
ment method, the way it is said has essentially nothing to do with what is said. It does,
however, have a great deal to do with the general way the content is approached and
understood […].”
Abhidharma and Indian thinking 43
Abbreviation
PPN Āgamic Index, vol. I: Prakrit Proper Names (ed. Malvania, 1970–1972)
Bibliography
1 Introduction
The older manuscripts preceding this group were written most likely in the late
first, early second century CE. Again this date could be confirmed by radio
carbon dating (for the Prajñāpāramitā cf. Falk, 2011: 20). Although two of these
early manuscripts contain texts which can also directly be linked to extant ver-
sions of Mahāyāna works, two of them seem to represent texts which are
hitherto unknown and have not been transmitted in any of the known Buddhist
literary traditions. These texts promise new insights into the formative phase
1 For more details cf. Strauch, forthcoming. See also Allon and Salomon, 2010.
At least two of these four early texts – the “Bajaur Mahāyāna sūtra” and the
Prajñāpāramitā – hail, according to reliable records, from the region along the
Pakistan-Afghanistan border, in the modern-day districts of Bajaur and Dir, i.e.
east of the Hindukush range.2
Among these early texts, the Bajaur Mahāyāna sūtra is of special interest.
First, it is by far the largest of these early Mahāyāna texts, and the longest text
in Gāndhārī known so far, comprising around six hundred lines on a large com-
posite birch bark scroll of about 2 meters length. Second, it belongs to those
texts, for which no parallel in another language is known.
The Bajaur Mahāyāna sūtra is part of the Bajaur Collection, which was alleg-
edly discovered in the ruins of a Buddhist monastery near the village Mian Kili
at the Dir-Bajaur border. The collection comprises texts of various Buddhist
literary genres, such as āgama, vinaya, rakṣā and stotra texts. A considerable
number of texts belong to the genre of scholastic literature, some of which
have a distinctive Mahāyāna tendency.3 Even non-Buddhist texts such as a
rājanīti verse anthology and a loan contract could be identified among the
birch barks of the Bajaur Collection.4
The study of the large Bajaur Mahāyāna sūtra is still ongoing. Although the
edition still needs some further research regarding certain passages, a more
general discussion and summary of its contents will be published soon.5
2 For the origin of the Bajaur Collection cf. Strauch, 2008; for the Split Collection see Falk, 2011.
3 The best preserved texts of this group were edited by Andrea Schlosser in her dissertation “On
the Bodhisattva Path in Gandhāra – Edition of Fragment 4 and 11 from the Bajaur Collection
of Kharoṣṭhī Manuscripts” (2013, revised version 2016).
4 For a general survey of the collection see Strauch, 2008. Separate texts are dealt with in
Strauch, 2011; Strauch, 2014a and 2014b.
5 The editing of the text is carried out by Ingo Strauch and Andrea Schlosser within a coopera-
tion between the Chair of Buddhist Studies at Lausanne University and the project “Early
Buddhist Manuscripts from Gandhāra” of the Bavarian Academy of Sciences, Munich.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 49
The entire sūtra can be divided into two different narratives. The first, frame
narrative represents a dialogue between the Buddha and Śāriputra. This dia-
logue occurs at the Vulture Peak in Rājagṛha. The Bajaur sūtra shares this
location with other early Mahāyāna sūtras, as e.g. the Aṣṭasāhasrikā
Prajñāpāramitā and the Gāndhārī Prajñāpāramitā (cf. Falk and Karashima,
2012: 28).
The second narrative, which is mainly devoted to the description of the
bodhisattva path, reports the dialogue between the Buddha and 84,000
gods (devaputra). This part contains many elements which are known from
other early Mahāyāna texts, as e.g. the prediction of a future buddha land
(vyākaraṇa). It is a distinctive feature of the text that this prediction refers to
the buddha land Abhirati of the Buddha Akṣobhya. Moreover, the text cele-
brates dharmakṣānti, “endurance towards the [non-arising of the] factors of
existence,” as the major characteristic of a bodhisattva. As a characteristic pas-
sage of the sūtra’s approach we cite the following passage:6
Thus addressed, the Blessed One said to the Venerable Śāriputra: (*Of all
dharmas), Śāriputra, a beginning (ādi) is not conceived, a middle (mad-
hya) is not conceived, an end (paryavasāna) is not conceived. And
because, Śāriputra, of all dharmas a beginning is not conceived, a middle
is not conceived, an end is not conceived, of this [single] dharma,
Śāriputra, a decrease (hāni) is not conceived, a stability (sthiti) is not con-
ceived, an increase (vaipulyatā) is not conceived. And because, Śāriputra,
6 The quotations and translations in this article are based on the ongoing edition by Ingo
Strauch and Andrea Schlosser. The conventions are those of the series Gandhāran Buddhist
Texts, i.e. [ ] uncertain reading, (*) editorial restoration of lost text, ⟨* ⟩ editorial addition of
omitted text, { } editorial deletion of redundant text, ? illegible akṣara, + lost akṣara, /// textual
loss at left or right edge of support (cf. <http://gandhari.org/a_dpreface.php>). ◊ signifies an
intentional space.
50 Schlosser And Strauch
Passages like this are not rare in early Mahāyāna sūtras, as is, for example,
shown by an almost literal parallel from the Aṣṭasāhasrikā (see Aṣṭa §2, ed.
Vaidya, 1960: 32).
The focus of early Mahāyāna sūtras towards the character of dharmas was
interpreted by Johannes Bronkhorst as clear evidence for the influence of
Gandhāran scholasticism on early Mahāyāna.7 As Bronkhorst (forthcoming)
argues:
7 On Sarvāstivāda scholastisicm in the northwest (Gandhāra and Bactria) see Willemen, Dessein
and Cox, 1998: 255–285.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 51
185 BCE. The foundations for the new Abhidharma may therefore have
been laid toward the middle of the second century BCE.
2 Groups of Four
8 Thanks go to Paolo Visigalli and Lin Qian for reading and commenting upon an earlier draft
of this part.
9 The meaning of this word is yet uncertain. The vi is inserted based on the spelling in the
preceding passage, where the word occurs twice as vidavidi.
52 Schlosser And Strauch
Since the end of the line is not preserved any more (1B.14, see fig. 1), the exact
wording cannot be established and also the verbal connection to the following
passage is unclear, but it should still be spoken by the bhagavant and the miss-
ing portion should only comprise the end of the rhetorical question addressed
to Śāriputra.
10 Possibly written with a stroke above the j (G. j̄aṇa) as etymologically expected, but the
manuscript is folded here, concealing the upper part.
11 Perhaps a term corresponding to the four ārūpyasamāpattis or the four samādhibhāvanās.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 53
12 Most probably this is G. viñaṇaṭ́hiḏio, based on the remaining traces of ink and the
sequence in two versions of the Saṃgītisūtra/-paryāya, namely G Cm and T.1.1 (cf.
table 3).
13 Cf. also Gethin, 2001, especially pp. 264–283 about the ‘seven sets expanded’.
14 Cf. Bronkhorst, 1985: 306, note 8: “It is remarkable that the Dīrghāgama preserved in Chi-
nese seems to have only list II, not I”. Also a list in another Gāndhārī manuscript (frag-
ment 5 of the Senior Collection) contains this extended list (II) corresponding to the
Chinese Dīrghāgama (see Glass, 2007: 35).
15 This combination is not extant in any text but a “hypothetical construction” (Bronkhorst,
1985: 308).
54 Schlosser And Strauch
I II III *IV
The same set of items (III, excluding the ārūpyas) also occurs in the Dhātukathā16
or the Vibhaṅga. In the latter, the next (and last) group comprising four items
are the four paṭisambhidās, quite similar to the BajC2 list: (4) ariyasaccāni,
(7) satipaṭṭhānā, (8) sammappadhānā, (9) iddhipādā, (12) jhānāni, (13)
appamaññāyo, (15) paṭisambhidā (the numbering reflects the chapters, cf.
also Frauwallner, 1995: 17f.).17 Thus, all seven fourfold categories discussed in
16 Bronkhorst, 1985: 306 (page 1 of the PTS edition, cf. also Narada, 1962: xlviii).
17 The ārūpyas have been included under the heading jhāna, but this is thought to be a
later addition. According to Bronkhorst, 1985: 308, a part of the Vibhaṅga (pp. 193–305)
is based on the following list: (1) 4 satipaṭṭhāna, (2) 4 sammappadhāna, (3) 4 iddhipāda,
(4) 7 bojjhaṅga, (5) 8-aṅgika magga, (6) 4 jhāna, (7) 4 appamaññā, (8) 5 sikkhāpada,
(9) 4 paṭisambhidā, thus excluding the 4 sacca. The Vibhaṅga is believed to “have
developed out of an earlier work [before 200 BCE] which also underlay the Dharma
skandha of the Sarvāstivādins” (Bronkhorst, 1985: 308). There, the ārūpyas as well as
the satyas are contained (T.26.1537: 453b24–514a10, 阿毘達磨法蘊足論, Apidamo
fayun zu lun, tr. by Xuanzang 玄奘). However, the sequence of the chapters is differ-
ent (Frauwallner, 1995: 15f. (= 1964: 73–74)): (2) srotāpattyaṅgāni, (3) avetyaprasādāḥ,
(4) śrāmaṇyaphalāni, (5) pradipadaḥ, (6) āryavaṃśāḥ, (7) samyakpradhānāni, (8)
ṛddhipādāḥ, (9) smṛtyupasthānāni, (10) āryasatyāni, (11) dhyānāni, (12) apramāṇāni, (13)
ārūpyāṇi, (14) samādhibhāvanāḥ (only chapter (1) [the 5 śikṣāpadāni] and chapter (15)
[the 7 bodhyaṅgāni] of the first part (–494b29) are not groups of four).
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 55
the Vibhaṅga are contained within the BajC2 list under the first eight items,
leaving no indication for the uncertain item no. 7 though. Nevertheless, one
tentative conclusion might be that the listing in BajC2 is based on a list also
occurring in the Vibhaṅga and Dhātukathā, but with the inclusion of the
ārūpyas or another additional group of four. The satyas, however, are posi-
tioned not at the beginning but in the middle of two categories pertaining to
meditation. Interestingly, they are placed likewise in the Sanskrit version of
the Saṃgītisūtra (‘Skt’) and its Chinese commentary (T.26.1536), additionally
followed by saṃjñā / xiang 想, see table 2. The Pāli version (‘P’) lists samādhi-
bhāvanā instead, the satyas (P. sacca) or saṃjñās (P. saññā) are not contained
at all; in the Gāndhārī commentary and the other Chinese versions they are
inserted later (satya: G Cm 23., T.1.1: 23., T.1.12: 9.; saṃjñā: G Cm 34., T.1.1: 34.,
T.1.12: –).18
Table 2 Sequence of dhyāna, apramāṇa and ārūpya in the different versions of the Saṃgītisūtra/-
paryāya.
Thus, G Cm, T.1.2, and T.1.12 seem to represent an older version of this particu-
lar sequence, where the first three fourfold items of the bodhipakṣya-dharmas
plus the dhyānas are immediately followed by the ‘unlimited’. The list in BajC2
should be more recent than the list preserved in these versions, but older than
the one in Skt/T.26.1536, however being part of the same strand of develop-
ment that included the satyas at this position (so far only attested in
18 For references and more information about the different versions of the Saṃgītisūtra/-
paryāya see table 3.
56 Schlosser And Strauch
- va[ś]ida
- [veharaja]
- as̱aharia dharma
19 In the Sanskrit Dīrghāgama manuscript from Gilgit, the relevant passages of the
Saṃgītisūtra are too fragmentary to be taken into consideration here. Apparently, only
IV.12 and IV.20 are preserved partly (thanks to Jens-Uwe Hartmann for sharing unpub-
lished information). In the table, ~ indicates that the equivalence is uncertain.
Table 3 Comparison of the list in BajC2 with the groups of four occurring in the Saṃgītisūtra/-paryāya.
01. (*catvari ś̱paḏovaṭ́haṇa ·) 01. smṛtyupasthāna 01. 念住 01. satipaṭṭhānā 01. 念處觀 11. 念處 11. [ś̱paḏ]ova[ṭ́ha]ṇa
02. (*catvari) samapra[s̱aṇa] · 02. samyakprahāṇa 02. 正斷 02. sammappadhānā 02. 正斷 12. 意斷 12. saṃmepras̱aṇa
03. [catvari] irdh[ipada] · 03. ṛddhipāda 03. 神足 03. iddhipādā 03. 神足 13. 神足 14. idhivaḏa
04. catvari jaṇa · 04. dhyāna 04. 靜慮 04. jhānāni 04. 禪定 14. 禪 13. j̱aṇa
05. catvari saca · 05. āryasatya 05. 聖諦 – 09. 聖諦 23. 聖諦 23. arias̱aca
06. catvari apramaña · 07. apramāṇa 07. 無量 06. appamaññāyo 05. 無量 15. 梵堂 15. apravaṃña
07. catvare ? + + + + +
08. [catvare] (*pa)ḍ̱isabhida · – – – 26. 無礙解 27. 辯才 27. paḍisaṃbiḏa
09. catvare va[ś]ida · – – – – – –
10. ca[tvare veharaja] · – – – – – –
11. catvare ñaṇamulea dharma · ~ 13. jñāna ~ 13. 智 ~ 11. ñāṇāni ~ 07. 智 ~ 26. 智 ~ 26. ñaṇa
12. catvare so[ḏavati](*aga ·) 12. aṅgaiḥ samanvāgataḥ 12. 證淨 14. sotāpannassa aṅgāni 18. 預流身 20. 須陀洹支 20a. soḏavatiaga
srotāpannaḥ
13. + + + + + + ? ? +
14. catvare taṣ̄amulea dharma (*·) ~ 24. tṛṣṇotpāda ~ 34. 愛 ~ 20. taṇhuppādā ~ 30. 愛生 – –
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism
15. catvare paḍ̱i[va](*da ·) 31. pratipad 21. 行 21. paṭipadā 17. 神通道 22. 道 22. paḍivaḏa
16. + + + + + + + + + +
17. + + + + + + + ? ? ?
18. [catvare] ? ? ? ? + + 23. vijñānasthiti 33. 識住 18. viññāṇaṭṭhitiyo 13. 識住 28. 識住處 28. viñaṇaṭ́hiḏio
(= viñaṇaṭ́hiḏio?)
57
58
Table 3 Comparison of the list in BajC2 with the groups of four occurring in the Saṃgītisūtra/-paryāya (cont.).
19. + + + + + + ? ? ?
20. [ca]tvare sakṣig̱araṇia dharma 20. sākṣīkaraṇīya dharma 20. 應證法 30. sacchikaraṇīyā – 21. 受證 21. sakṣiḵatava
dhammā
21. catvari as̱aharia dharma – – – – –
G Cm: BL15, ca. 0–100 CE, not published yet (groups of four to be edited by Stefan Baums).
P: Saṃgītisuttanta, DN III 221–233.
Skt: Saṃgītisūtra, Central Asian manuscripts, ca. 7th c. CE, ed. Stache-Rosen, 1968.
T.1.1: ~ Saṃgītisuttanta (眾集經, Zhongji jing, DĀ, sūtra no. 5), T.1.1.50b23–51b4,
tr. Buddhayaśas, ca. 5th c. CE, tr. in Behrsing, 1930.
T.1.12: ~ Saṃgītiparyāyasūtra (大集法門經, Daji famen jing = *Mahā-saṃgīti-sūtra), T.12.1.228b16–230a5, tr. by Dānapālaa, ca. 1000 CE.
T.26.1536: ~ Saṃgītiparyāya (阿毘達磨集異門足論, Apidamo jiyimen zulun = *Abhidharma-saṃgīti-paryāya-pāda-śāstra), T.26.1536.26.391b11–411c11), tr. by
Xuanzang, 660–663 CEb, tr. in Stache-Rosen, 1968.
Skt is closely connected to T.26.1536 (Sarvāstivāda), which is a commentary on it. P (Theravāda) and T.1.12 (affiliation unknown) seem to stand for themselves,
although being connected to Skt/T.26.1536 due to the same beginning. G Cm is closely connected to T.1 (Dharmaguptaka).
a Also BajC1 (*Gautamīsūtra) has a closely related, though not identical Chinese version that has been translated by Dānapāla (T.1.84, ca. 980–1000 CE, cf.
Strauch, 2007/2008: 19–21, and Strauch, 2014a: 33). He was from Uḍḍiyāna (Swat), his school affiliation is unknown (cf. Strauch, 2014 a: 26 and 35).
b Willemen, Dessein and Cox, 1998: 177.
Schlosser And Strauch
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 59
20 Dhsgr 74: āyur-, citta-, pariṣkāra-, dharma-, ṛddhi-, janma-, adhimukti-, praṇidhāna-,
karma-, jñāna-◦. A partly different explanation is given in the Abhisamayālaṃkāravṛttiḥ
sphuṭārthā (AAV, ed. Tripathi, 1977: 3–44) on Abhisamayālaṃkāra (AA) 8.4: āyuś-citta-
pariṣkāra-karmopapatty-adhimukti-praṇidhāna-rddhi-jñāna-dharma-vaśitā iti daśa vaśi
tāḥ. See Brunnhölzl, 2011: 114 and also Brunnhölzl, 2010: 659 (chart 12) for a translation.
The same list is given e.g. in the PvsP (fol. 532b; ed. Kimura, 2006 [VI–VIII]: 59) and the
Sāratamā (ed. Jaini, 1979: 176), and – slightly varied – the Catuḥstavasamāsārtha (ed.
Tucci, 1956: 239).
21 Cf. e.g. Dhsgr 75: bodhisattvānāṃ daśa balāni / tadyathā // adhimuktibalaṃ pratisaṃ
khyānabalaṃ bhāvabalaṃ kṣāntibalaṃ jñānabalaṃ prahāṇabalaṃ samādhibalaṃ
pratibhānabalaṃ puṇyabalaṃ pratipattibalaṃ ceti // 76. tathāgatasya daśa balāni /
tadyathā // sthānāsthānajñānabalaṃ karmavipākajñānabalaṃ nānādhātujñānabalaṃ
nānādhimuktijñānabalaṃ sattveṃdriyaparāparajñānabalaṃ sarvatragāminīpratipatti
jñānabalaṃ dhyānavimokṣasamādhisamāpattisaṃkleśavyavadānavyutthānajñānabalaṃ
pūrvanivāsānusmṛtijñānabalaṃ cyutyutpattijñānabalam āsravakṣayajñānabalaṃ ceti.
22 T.31.1599: 455a7–8 (MAVBh, Paramārtha), T.31.1600: 468b5–6 (MAVBh, Xuanzang); root
text: T.31.1601: 478b25, MAV, Xuanzang). Another explanation of the ‘four sovereign
powers’ is: jie 戒 the moral law; shentong 神通 supernormal powers; zhi 智 knowledge;
and hui 慧 wisdom (Soothill, according to the Digital Dictionary of Buddhism, www.
buddhism-dict.net).
23 A worthwile start would be the passages in T.1.13 (Chang ahan shi bao fa jing 長阿含十報
法經), T.2.125 (Zengyi ahan jing 増一阿含經 ~ Ekottarāgama), T.9.272 (Da sazhe niganzi
suoshuo jing 大薩遮尼乾子所説經 ~ Mahāsatyanirgrantha-sūtra), T.10.279 (Dafang-
guang fo huayan jing 大方廣佛華嚴經).
60 Schlosser And Strauch
G [veharaja]. In the Pāli canon, the four vesārajjas are explained as the
self-confidences or fearlessnesses of a buddha, because of which he cannot
be reproved by an “ascetic or brahmin or deva or Māra or Brahmā or anyone
in the world” of (1) not having reached full enlightenment, of (2) not having
destroyed all taints, of (3) not having understood the obstructions, of (4) not
having taught the correct way to the destruction of suffering.24 Also, in the
Mahāvastu, they are enumerated as one of the characteristics of a buddha.25
They are more often mentioned in Sanskrit texts, most of which are Mahāyāna-
related, where they likewise determine characteristics of an awakened being.
The four vaiśāradyas are also included in similar lists in Prajñāpāramitā texts,
but here they are always preceded by the (ten tathāgata-) balas and not by
vaśitās (cf. table 4).26 The apparently only text listing the (ten) vaśitās is the
Abhisamayālaṃkāra, where they occur adjacent to and precede the (ten) balas
(cf. table 4, A).27 This list is similar to the list in BajC2 in regard to the sequence
of the fours, even though some of the groups are not mentioned (jaṇa = Skt.
dhyāna, saca = Skt. satya, and the uncertain one). It is however different from
other lists in Prajñāpāramitā texts (cf. table 4, B).
* *
154b19–22: without placed here or
dhyāna etc. at the end of
153a16–19: with the list
dhyāna etc.
149b08–09: with
maitrī/karuṇā
29 References for the LPG parts edited by Conze: ed. Conze, 1962: 57 (fol. 229b), 142f. (fol.
251a–b), 162 (fol. 255b), 180 (fol. 260a), 185f. (fol. 261b); ed. Conze, 1974: 11 (fol. 268b), 24
(fol. 273a), 29 (fol. 274b), 46 (fol. 279a–b), 80 (fol. 290a), 126 (fol. 305a). Four times the list
does not begin with dhyāna etc.; thrice these terms are missing, once they are inserted
after the 37 bodhipakṣya-dharmas (ed. Conze, 1974: 29, fol. 274b). The mahāmaitrī etc. can
be placed after the āveṇikadharmas or before it, but they are – with one exception –
always included.
30 The text references for the PvsP are too numerous to list here, one example is ed. Kimura,
1990 [IV]: 13. In some instances, near the beginning and the end of the whole text, also the
“old” sequence is given, beginning with dhyāna etc. (e.g. ed. Kimura, 2007 [I-1]: 149, and
ed. Kimura, 1992 [V]: 151). The same with the Śatasāhasrikā (ed. Kimura, 2009–2010, see
e.g. Kimura, 2010 [II-3]: 39). In contrast to the LPG and the PvsP, the list in the Śatasāhasrikā
is stable. Cf. also Advayaśatikā, ed. Shakya, 1988: 82–84.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 65
BajC2 seems to agree more closely with the Sarvāstivāda and Theravāda ver-
sions than the Dharmaguptaka. Between these two, the enumeration in BajC2
shares some features with the Pāli (Theravāda) on account of the sequence. At
the same time, however, there are exceptions, where it fits better to the Skt ver-
sion and T.26.1536 (Sarvāstivāda), and it also contains a category (G. saca) not
extant in the Pāli version. A special connection is given to T.1.1, T.1.12 and G Cm
due to the term (*pa)ḍ̱isabhida (T.1.1: biancai 辯才, T.1.12: wu aijie 無礙解, G
Cm: paḍisaṃbiḏa), which only occurs in those versions, although at later posi-
tions. Among all versions, the list in BajC2 is perhaps most similar to T.1.12 in
representing an intermediate state between the Theravāda / Sarvāstivāda and
the Dharmaguptaka versions. A link to Prajñāpāramitā texts is indicated by the
four veharajas (Skt. vaiśāradya), a term that is not known from the Saṃgītisūtra.
Also the va[ś]idas (Skt. vaśitā) point to an early Mahāyāna affiliated context.
The groups that have been mentioned so far are represented as the first ten
items of the list in BajC2, as far as they are characterized as being fourfold. The
subsequent ten items are a seemingly random selection of fourfold groups, also
known from the Saṃgītisūtra. The last of the altogether twenty-one items are
the asaṃhārya-dharmas, a term peculiar to the Aṣṭasāhasrikā, which will be
discussed later. The modular composition of the list is summarized in table 5.
33 The sequence of terms under each point is: »G: Skt; P; E«. Unless otherwise stated, the
citations are taken from the reconstructed Sanskrit version given in the edition of Stache-
Rosen, 1968. If there are significant differences to the other versions in Pāli or Chinese this
is noted.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 67
11. catvare ñaṇamulea dharma · fourfold groups, also occurring in the dharmas to be
12. catvare so[ḏavati](*aga ·) Saṃgītisūtra, apparently random known
13. + + + + + + ? ? + selection
14. catvare taṣ̄amulea dharma
(*·)
15. catvare paḍ̱i[va](*da ·)
16. + + + + + + + + + +
17. + + + + + + + ? ? ?
18. [catvare] ? ? ? ? + +
(= viñaṇaṭ́hiḏio?)
19. + + + + + + ? ? ?
20. [ca]tvare sakṣig̱araṇia
dharma
7. catvare ? + + + + +
Uncertain. Most probably either the ārūpyasamāpattis36 or the samādhibhā
vanās.37
40 Cf. PTSD s.v. paṭisambhidā: “BSk. pratisaṃvid is a new formation resting on confusion
between bhid & vid”.
41 Skt IV.15 śraddhā, vīrya, samādhi, prajñā; P IV.26 sati, viriya, … There are also four other
balas “leading to a Bodhisattva’s cittotpāda, Bbh 13.22, listed 17.8–9 as adhyātma-, para-,
hetu-, prayoga-bala” (BHSD s.v. bala).
42 Source of explanation: Abhidharmasamuccaya (Abhidh-s 98).
43 Due to the different terminology it is not entirely certain that the Gāndhārī refers to the
four jñānas as given in the Saṃgītisūtra. Another secondary explanation of the four
knowledges relates to the four truths (Skt IV.14: duḥkha, samudaya, nirodha, mārga; also P
IV.12). G Cm seems to mix the two alternative explanations.
44 Cf. the explanation in T.26.1536 (Stache-Rosen, 1968: 100).
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 71
13. + + + + + + ? ? +
Uncertain.
45 There are two different explanations. The above-mentioned one seems more likely here,
since the four avetyaprasādas / abhedyaprasādas (G. abhejapras̱ada) are referred to later
in the text, although this is no conclusive evidence. Most versions list both explanations,
the first is named Skt. srotāpattyaṅgaḥ / T.26.1536 yuliuzhi 預流支 / P sotāpattiyaṅgāni /
T.1.12 – / T.1.1 –; the second one is named Skt. caturbhir aṅgaiḥ samanvāgataḥ srotāpannaḥ
/ T.26.1536 zhengjing 證淨 / P sotāpannassa aṅgāni / T.1.12 yuliushen 預流身 / T.1.1 xutuo-
huanzhi 須陀洹支. G Cm calls both soḏavatiaga, and says that the second explanation
is favored “here” (G. iśa). The alternative (and probably older) explanation would be:
1. associating with good people (satpuruṣasaṃseva); 2. listening to the good doctrine
(saddharmaśravaṇa); 3. investigating it thoroughly (yoniśo manasikāra); correct behav-
iour according to the doctrine (dharmānudharmapratipatti).
46 Skt avetya◦ / T.26.1536 zhengjing 證淨 / P avecca◦ / G Cm aveca◦ / T.1.12 buhuai 不壞 /
T.1.1 buhuaixin 無壞信. In the subsequent text of BajC2, the term is spelled abhejopras̱ada
(= abhedyaprasāda). In the Saṃgītisūtra preserved in the Gilgit Dīrghāgama manuscript,
the form avetya◦ is used (thanks to Jens-Uwe Hartmann for the information on this
unpublished manuscript portion). For a more comprehensive discussion of the term
abhedyaprasāda cf. § 3 of this article.
47 Similar to the four jñānamūlaka-dharmas (11.), it is uncertain if the tṛṣṇāmūlaka-dharmas
equate to the tṛṣṇotpādas in the Saṃgītisūtra at all, since they denote things that have
tṛṣṇā as a cause and not as a result. There are however no four tṛṣṇāmūlaka-dharmas, but
only nine (taṇhāmūlaka-dhamma) in the Aṅguttaranikāya (AN IV 400–401, cf. DN II
58–61), the Paṭisambhidāmagga (Ps 130), and the Vibhaṅga (Vibh 390).
48 Mss. (Hoernle, Hs. 47/48): bhavatibhava◦ (cf. Stache-Rosen, 1968: 79, note 133); bhavā
bhava, for which Thomas W. Rhys Davids gives “dainty foods” like “oil, honey, ghee, etc.”
72 Schlosser And Strauch
16. + + + + + + + + + +
Uncertain.
17. + + + + + + + ? ? ?
Uncertain.
19. + + + + + + ? ? ?
Uncertain.
according to the commentary of Buddhaghosa. T.26.1536 explains bhava as the five skan-
dhas, thus ‘existence’.
49 The translation is taken from Walshe, 1995: 492 (P). The Sanskrit and Pāli versions also
contain another explanation, that is however not contained in the Chinese versions,
which is why the given explanation has been preferred. The alternative would be (Skt
IV.32, P IV.22): 1. inability to endure (akṣamā), 2. taming / self-control (damā), 3. ability to
endure (kṣamā), 4. appeasement (chamā).
50 P has ◦upāya instead of ◦upaga; G Cm has ruovao / veḏaṇ[o]vao / saṃñ[o]ao /
saṃkharovao (preliminary unpublished transliteration), which can be both.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 73
May that thought of enlightenment which they have wished for, thought
over and taken hold of, bring to fulfillment in them the dharmas of a
Buddha, and dharmas associated with all-knowledge, the dharmas of the
Self-Existent, the insuperable dharmas [asaṃhārya-dharmāṇāṃ]! (tr.
Conze, 1973; xxvi 434, ed. Vaidya: 215).
An Arhat, a monk whose outflows are dried up, does not go by someone
else whom he puts his trust in, but he has placed the nature of dharma
directly before his own eyes, and Mara has no access to him [asaṃhāryo
bhavati māreṇa]. Just so an irreversible [avinivartanīyo] Bodhisattva
cannot be crushed by persons who belong to the vehicle of the Disciples
and Pratyekabuddhas, he cannot, by his very nature, backslide into the
level of Disciples or Pratyekabuddhas, he is fixed on allknowledge, and
51 The sequence varies: Skt 1–2–3–4; P 2–3–1–4; T.1.1 3–1–2–4 (cp. Behrsing, 1930: 75–76 note
169); G Cm 3–2–1–4.
52 T.1.1 and G Cm have here “cessation [of perception and feeling]” (G. ṇiros̱o; T.1.1 shenshou
mie zheng 身受滅證 ~ P vedayita-nirodha…).
53 T.1.1 and G Cm have here “forms” (~ rūpa).
74 Schlosser And Strauch
He can no longer be led astray by others, and on the stage which is his by
right he cannot be crushed. For, as he has stood firm on it, his mind
becomes insuperable, his cognition becomes insuperable (aparapraṇeyo
bhavati, anavamardanīyaś ca bhavati svasyāṃ bhūmau / tatkasya hetoḥ?
tathā hi sa sthito ’saṃhāryeṇa cittena asaṃhāryeṇa jñānena samanvāgato
bhavati, tr. Conze, 1973; xvii 337, ed. Vaidya: 168).
Also, a passage in the Larger Prajñāpāramitā from Gilgit (fol. 253b) circum-
scribes the practice of the perfection of wisdom as a state, in which
55 Reconstruct vedaṇehi ?
56 Clearly written ku, but perhaps kaṃ was intended, like in other Gāndhārī manuscripts
(next to kadha or ḱadha).
57 Maybe pratītyasamutpāda is to be inserted here, as it follows after skandhadhātvāyatana
(and precedes the bodhipakṣya-dharma) in lists in the Larger Prajñāpāramitā (LPG). Also,
those lists are concluded and analysed by terms like saṃkleśa and vyavadāna, just as in
BajC2 a few words later.
58 Maybe tasagadadharma (Skt. tathāgatadharma) is to be reconstructed.
59 G. b[o]s[o]: Skt. bodho (?). Skt. buddho is excluded, since it should be written budho or
bodho. Nevertheless, since a translation as “awakening” seems rather unlikely due to the
associated verbal forms (G. samuṇag̱ada = samanvāgata and sahariadi = saṃharīyate), it
appears to be an unusual bahuvrīhi (“possessing awakening”). Alternatively, the transla-
tion would be “A state of awakening, which is endowed with these four unconquerable
things is not conquered”.
76 Schlosser And Strauch
[He] is not conquered by adherents of other sects like the caraka mendi-
cants60 or the nirgrantha mendicants61,
[He] is not conquered by the fourfold army of Māra62,
[He] is not conquered by …,
[He] is not conquered by feelings of suffering,
[He] is not conquered by (feelings ?) [such as] non-suffering [or]
non-happiness,
[He] is not conquered by states of [wrong] views,
[He] is not conquered by …,
[He] is not conquered by states of propensity [or] by states of aversion
(?)63,
[He] is not conquered by defiled (sāsrava) aggregates, elements, [or]
sensory bases (skandhadhātvāyatana) leading to rebirth (punarbhavika)
[…]
[etc.] up to all characteristics (dharma) associated with awakening
(bodhipākṣika),
all characteristics associated with defilement (saṃkleśapākṣika),
or all characteristics associated with purification (vyavadānapākṣika),
all …
Why has it been shown? [Now] the characteristics (dharma) (of a tathā
gata ?) are known. Having been asked, I have answered adequately
(samartha).
device to introduce the four abhedyaprasādas, while at the same time illustrat-
ing the author’s knowledge of certain lists and categories, of which he
enumerated all those fourfold ones that came to his mind in order to represent
the Dharma.
The discourse about the groups of four culminates in a long exposition about
the abhedyaprasāda “unbreakable confidence/trust”. Despite its obviously dif-
ferent etymology this term has to be related to its Pāli equivalent aveccapasāda
or its Sanskrit representative avetyaprasāda which are usually translated as
“perfect confidence/trust/faith” or “confidence/trust/faith based on under-
standing,” respectively. As in our text, in canonical literature, these terms
describe one of the characteristic features of an āryaśrāvaka or srotāpanna
“stream-enterer”. But not only the altered etymology distinguishes our text
from these parallels, the inclusion of the abhedyaprasādas/avetyaprasādas
among the group of four is similarly remarkable.
Based on the amount of text devoted by the Bajaur Mahāyāna sūtra to this
issue, the abhedyaprasādas played a key role in the concept of an āryaśrāvaka.
In order to determine the specific role of the abhedyaprasādas in the Bajaur
Mahāyāna sūtra, our exposition will focus on three major points:
3.1 From aveccapasāda to abhedyaprasāda: shifting etymologies
3.2 The abhedyaprasādas as a group of four
3.3 The reinterpretation of this category in the context of the Bajaur
Mahāyāna sūtra
In Pāli texts, the same term regularly occurs as aveccapasāda. The etymology of
the first member of this compound is doubtful. Modern Pāli dictionaries,
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 79
including the Critical Pāli Dictionary, derive avecca from the root ava-(ā)-√i “to
understand, to know”. This explanation appears to be based on two aspects.66
First, some of the Pāli commentaries paraphrase the initial avecca with
words meaning “to understand, to know”. Thus, Buddhaghosa’s commentaries
explain avecca by paññāya ajjhogahetvā, paṭivijjhitvā, ñatvā or jānitvā (cf. CPD
s.v. avecca). Secondly, an etymology based on Skt. ava-(ā)-√i is also indicated
by the Sanskrit variant of this term, which is usually given as avetyaprasāda (cf.
BHSD s.v.). Consequently, modern studies on the Buddhist concepts of faith
and belief characterize this scholastic category as “confidence/trust/faith
based on understanding”. Thus, Rupert Gethin writes:
66 Cf. also the detailed note by Samtani in his edition of the Arthaviniścaya (1971: 241).
67 During the discussion at the 1st Lausanne Gāndhārī Workshop in June 2013, Harry Falk
suggested an alternative etymology of the Pāli word avecca based on the root √vic “to sift,
separate”. In this case one would certainly have to distinguish the absolutive avecca used
in isolated position in a phrase and the gerundive avecca used as first member of a
80 Schlosser And Strauch
There can be little doubt that the text has to be corrected into buddhābhe
dyaprasād°68 and can be translated as:
The same form abhedya also occurs in the summarizing verses (upasaṃhāra
gāthā) devoted to the fourth bhūmi (ed. Vaidya, 1967: 79–80):
compound word. According to this explanation, the meaning of the Gāndhārī variant
abheja would nearly correspond to the original meaning of the term. In any case, the com-
mentaries of Buddhaghosa as well as the Sanskritized term avetya leave no doubt that the
etymological origin of this term was no longer understood.
68 This is in fact the reading given in the older edition by Rahder, 1926: 97.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 81
Unbreakable with regard to the Buddha jewel and the Dharma and the
Saṃgha, he sees that things are inactive with regard to their production,
cessation, and stability.
It is obvious that both the Tibetan and the Chinese translation by Xuanzang 玄
奘 (T.14.476) confirm the reading of the Skt. version and use the already men-
tioned translations for abhedya (Tib. mi phyed pa, Chin. buhuai 不壞). The two
earlier Chinese translations are more difficult to evaluate. Whereas T.14.474 at
69 The glossary’s reference “(10M)” for one of the occurrences seems to refer to the Parīndanā
section, which corresponds in Rahder’s edition to ch. C.
70 The reference to the Vimalakīrtinirdeśa I owe to Dan Stuart.
71 The Tibetan translation from the Derge edition of the Kangyur and the Chinese transla-
tions T.14.474 by Zhiqian 支謙 (223–228 CE), T.14.475 by Kumārajīva 鳩摩羅什 (406 CE)
and T.14.476 by Xuanzang 玄奘 (650 CE) can be easily accessed in Jens Braarvig’s excel-
lent Thesaurus Literaturae Buddhicae (URL: <http://www2.hf.uio.no/polyglotta/index.
php?page=volume&vid=37>, accessed 2/11/2105).
82 Schlosser And Strauch
least in the second case seems to render abhedya by buli 不離, the other trans-
lations leave the element abhedya untranslated or altered its meaning. Thus,
abhedyaprasāda is represented in T.14.475 either as changxin 常信 “eternal
faith” or as shenxin 深信 “profound faith.”72
Another text which uses this variant is the Sarvatathāgatādhiṣṭhāna
vyūhasūtra, the Sanskrit version of which is only partially preserved in two
manuscripts from Gilgit.73 The text describes the effects of a meditation practice
called sarvatathāgatādhiṣṭhāna-sattvāvalokana-buddhakṣetra-sandarśana-
vyūho nāma samādhiḥ. One of these effects is described as follows:
72 For this last variant cf. the Digital Dictionary of Buddhism, s.v. (<http://www.buddhism-
dict.net/>).
73 The facsimiles of both manuscripts are reprinted in Raghu Vira and Lokesh Chandra,
1995: plates 1746–1815 and plates 1816–1837. The passage, which mentions abhedyaprasāda
is found only in the former.
74 The text given by Dutt is not entirely correct: smṛtimantaḥ prajñāvantaḥ buddhe dhama
saṃghe abhedyaprasādena samanvāgatā (ed. Dutt, 1984: 53).
75 The late Chinese translations T.19.1022 by Amoghavajra (eighth century CE) and T.19.1023
by Dānapāla (tenth/eleventh century CE) cannot confirm this reading. They use dingxin
定信 “firm faith” (T.19.1022A: 710 a23) and shanxin 善信 “good faith” (T.19.1023: 715 a19)
instead.
76 For a detailed description of the Gilgit version of the Larger Prajñāpāramitā see Zacchetti,
2005: 19–28.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 83
The Blessed One said: What do you think, Kauśika, how many people of
Jambūdvīpa are endowed with unbreakable confidence in the Buddha,
[are endowed with unbreakable confidence in the Dharma,] are endowed
with unbreakable confidence in the Saṃgha. Śakra said: Few people of
Jambūdvīpa, Blessed One, are endowed with unbreakable confidence in
the Buddha, [are endowed with unbreakable confidence in the Dharma,]
are endowed with unbreakable confidence in the Saṃgha.
However, the microfilm copy of this manuscript at the Berlin State Library con-
firms the variant avetyaprasāda. This reading is clearly the preferable one since
it corresponds to the following explanation of avetya as avagamya.
Vimuktisena’s commentary was taken up by the later commentator
Haribhadra,80 who comments on the same passage in his ālokā as follows (ed.
Wogihara, 1932: 213):
80 Haribhadra wrote his commentary on the Abhisamayālaṃkāra “about the year 800 dur-
ing the reign of Dharmapāla (rg. C. 770–810), the greatest of the Pāla kings” (Sparham,
2006: xv).
81 Once more, Sparham’s translation suggests that the commentary used the term
abhedyaprasāda: One, [i.e. Ārya-Vimuktisena] says, “they have ‘unbroken faith’ [(abhedya-
prasāda), i.e. ‘knowledgeable faith’ (avetya-prasāda)]” (Sparham, 2008: 160). However,
neither the text edited by Tucci, 1932: 182, nor Wogihara’s edition of the AAA (cf. above)
refer to this reading.
82 In the following the term a°prasāda is used to designate both variants of this term:
avetyaprasāda and abhedyaprasāda.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 85
first member with Skt. abhedya. There seems to be some evidence that this
variant was far more widely used than our small survey suggests. Thus, the
Mahāvyutpatti (ed. Sakaki, 1916: 440) lists not only the more common term
avetyaprasāda, but also adds the variant abhedyaprasāda:
The Tibetan mi phyed pa (< ’byed ba “to split, to break”) clearly points to Skt.
abhedya. It seems that the majority of the early Chinese translators up to
Paramārtha (563 CE) even preferred this variant. The first element of the com-
pound a°prasāda is here usually represented by Chinese buhuai 不壞
“indestructible, unbreakable, incorruptible.” Thus, we find the form buhuaixin
不壞信 in the Chinese Dīrghāgama (T.1.1), in the separately translated
Saṃyuktāgama (T.2.100), in both translations of the Larger Buddhāvataṁsaka
(T.9.278 and T.10.279), and in several Prajñāpāramitā texts translated by
Kumārajīva (e.g. T.8.227, T.25.1509). Only from the late sixth century onwards
(and in particular in the translations by Xuanzang) is the first element of the
compound usually represented by zheng 證 “to realize,” which may be related
to Skt. avetya.83
It is hardly probable that in all these aforementioned cases the underlying
Indic text really contained the reading abhedyaprasāda. But at least in cases
where the older variant buhuai 不壞 is used by post-6th c. translators (such as
some of Xuanzang’s translations), there is good reason to argue that the origi-
nal text contained abhedya rather than avetya. A systematic investigation of
this question is beyond the scale of the present study.
There is some evidence that certain Abhidharma texts are aware of this
alternative interpretation of the term. Discussing this issue with Lin Qian, he
drew my attention to an important passage from the Mahāvibhāṣā (T.27.1545,
534c14–29) and provided the following translation:
83 I am most grateful to Lin Qian and Jan Nattier, who kindly provided this evidence for me.
86 Schlosser And Strauch
It seems that at least two of the five explanations given here, numbers 2
and 3, point to abhedya as the underlying form rather than to avetya.84 A
slightly different explanation based on the same etymology is given by the
Saṃyuktābhidharmahṛdaya:
These two kinds of increase both acquire the first path. Māra cannot
destroy or break [it]. Each is named by depending on the specific expla-
nation (Dessein, 1999,1: 681, my emphasis).
It is not easy to split or break the space, (*just like a hole) that was pierced
by an even hundredfold split tip of a hair.
Is it therefore possible to argue that the new term was introduced in a north-
western environment before it was introduced into other contexts including
the translational practice of early Chinese translators?
There is no easy way to explain the sound change from Middle Indic avecca
to Gāndhārī abheja. Such a transformation can only be justified by an inten-
tional reinterpretation of an inherited, but obscure term. This transformation
might be at least partially due to a hypersanskritisation based on the language
of the Indian northwest. Only in the orthography of the northwest is Old Indian
bh- regularly represented by the sign for the labio-dental v or its aspirated vari-
ant vh. Besides that, the “historical” spelling bh is frequently attested (e.g. Skt.
85 The variant buhuaijing 不壞淨 is also used in the other Hṛdaya works by Dharmaśreṣṭhin
and Upatrāta. See footnote 96 below.
88 Schlosser And Strauch
prabhā > prava, pravha, prabha). Inherited intervocalic bh was obviously pro-
nounced as a fricative with or without aspiration, in clear departure from other
Middle Indic languages where we observe the change bh > h (von Hinüber,
2001: §§ 190f.). In a Gāndhārī environment a term avecca could easily be mis-
understood as a word containing an aspirated labial. The change from cca to
j( j)a is more complicated. It could be explained as an intentional shift from a
no longer comprehensible form aveca / abheca to a hypersanskritized form
abheja (Skt. abhedya). However, it cannot be completely excluded that this
shift also had a phonetical background. That the pronounciation of c and j was
sometimes confused, is demonstrated by some Prakrit grammarians (von
Hinüber, 2001: 155, § 177). Moreover, Kenneth R. Norman (1970: 134–135) lists a
number of words where this change obviously occurred. The interchangeabil-
ity of c and j is also occasionally attested in a Gāndhārī environment, as shown
by one of the Senior fragments where OI añjali is written as G acali (GD, Index
s.v. acali). Thus, both changes (v > bh, c > j) are at least hypothetically within
the range of possible phonetical developments of Gāndhārī. Especially the
characteristic shift from v to bh makes a Gāndhārī influence on the emergence
of this variant highly probable.
This fourfold list is not attested in the earliest layers of Buddhist literature but
appears to belong to a specific scholastic tradition. Usually, the early texts refer
only to three such items, namely the three jewels. These three a°prasādas are
arranged together with (ārya)śīla to another fourfold list: that of the
srotāpattyaṅgas, the constituents of stream entry. As locus classicus for the
definition of the a°prasādas as a part of the srotāpattyaṅgas in canonical lit-
erature, I quote a passage from the Pāli Saṃgītisuttanta (DN 33 III 227):
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 89
This fourfold list of srotāpattyaṅgas is also part of the Sanskrit, Gāndhārī and
Chinese versions of the Saṃgītisūtra and its commentaries (see § 2.3.2). As
seen above, the srotāpattyaṅgas (G so[ḏavati](*aga)) are also mentioned
among the groups of four listed in the respective section of the Bajaur
Mahāyāna sūtra.
90 Schlosser And Strauch
What are the four *avetyaprasādas (si zhengjing 四證淨)? They are: bud-
dha-avetyaprasāda, dharma-avetyaprasāda, saṅgha-avetyaprasāda, and
the virtue favored by the nobles. Why? The four great elements, namely,
the elements of earth, water, fire, and air, are capable of change; those
noble disciples who have achieved the four avetyaprasādas definitely will
not change (translation: Lin Qian).88
86 But cf. Willemen, Dessein and Cox, 1999: 69. Based on the quotations in the Saṃgītiparyāya,
they consider the Dharmaskandha as “the oldest of the seven Abhidharma works”.
87 I am once more indebted to Lin Qian, who guided me through the Chinese texts of the
Dharmaskandha and the SĀ. For more parallels to this sūtra see Chung, 2008: 185.
88 Although the Chinese translation of the SĀ (Guṇabhadra, 443 CE) uses the term buhuai-
jing 不壞淨 (Skt. abhedyaprasāda), the Dharmaskandha (Xuanzang, 659 CE) refers to
zhengjing 證淨 (Skt. avetyaprasāda). This again shows the difficulties in making any con-
clusions that are solely based on the terminology of early Chinese translations.
89 For the school affiliation of the Chinese SĀ see the discussion by Chung (2008: 11–20).
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 91
underlying the Chinese SĀ had already replaced the reference to the four
srotāpattyaṅgas by that to the four a°prasādas.
The character of a°prasādas as a group of four had become firmly estab-
lished in the later Sarvāstivāda Abhidharma tradition, although the distinction
between the first three members of this group and śīla as its fourth element
had been a matter of discussion. For example, the Mahāvibhāṣā discusses this
problem in the passage directly preceding the one cited above (T.27.1545:
534c5–10, quoted after Dessein, 1999: 513, note 450):
trisatyadarśane śīladharmāvetyaprasādayoḥ
lābho mārgābhisamaye buddhatatsaṃghayor api (ed. Pradhan, 1975: 386)
Although Vasubandhu clearly admits the difference between these two types
of avetyaprasāda, his commentary leaves no doubt that āryakāntaśīla-
avetyaprasāda has to be regarded as part of a fourfould list of avetyaprasādas.
For his Abhidharmakośa Vasubandhu used “the Abhidharma system as it
had been systematized by Dharmaśreṣṭhin and revised and enlarged by
Upaśānta and Dharmatrāta” (Willemen, Dessein and Cox, 1998: 270). The
*Abhidharmahṛdayaśāstra, written by Dharmaśreṣṭhin/Dharmaśrī from
Bactria probably between 220 BCE and 220 CE91 and translated into Chinese in
391 CE (cf. Willemen, Dessein and Cox, 1998: 255–256) represents the earliest
preserved systematic compilation of Sarvāstivāda dogmatics. Dharmaśreṣṭhin’s
work was the basis of the two Gandhāran Abhidharmahṛdaya works by
Upaśānta and Dharmatrāta who lived in the third and early fourth centuries
(cf. Willemen, Dessein and Cox, 1998: 259 and 261). All these works consistently
refer to a fourfould list of a°prasādas.92
All these references demonstrate that the tradition, which refers to the
a°prasādas as a group of four was well established in Sarvāstivāda Abhidharma,
although the different character of its three original members (Buddha,
Dharma, Saṃgha) and the later incorporated moral (śīla) continued to be
discussed.
Vasubandhu, as well as his direct predecessors Upaśānta and Dharmatrāta,
lived in Gandhāra. It might, therefore, be hardly surprising that the author of
the Bajaur Mahāyāna sūtra referred in his work to this fourfold list of a°prasādas
that had become commonplace at least in the northwestern Abhidharma
traditions.
That Buddhahood is comprised within the fruit of being without attachment. The quali-
ties of one who has no more training to do are the qualities of a Buddha. When one has
pure faith in these qualities, it is called perfect faith in the Buddha. Having taken up the
realization of that which is right, one is a disciple. The qualities of one in training and of
one who has no more training to do are said to be the qualities of a disciple. When has
pure faith in these qualities, it is called perfect faith in the saṃgha. Pure faith in nirvāṇa
and faith in the remaining formed dharmas, such as the truth of suffering and the truth of
origination, faith in the pure qualities of the bodhisattva, and faith in the qualities of the
pratyekabuddha who is in training or who has no more training to do, this is called perfect
faith in the dharma. Noble morality is pure morality. This is called perfect faith in moral-
ity” (Willemen, 1975: 135–136).
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 95
Dharma, Saṃgha and morality. This is obvious from the above quoted passage
from the Saṃgītisuttanta and it is also evident from the passage extracted from
the Abhidharmakośabhāṣya.
The author of the Gāndhārī sūtra chooses a different approach, which is,
however, in accordance with the general message of the text, which is based on
the notion of emptiness (śūnyatā). Usually the notion of emptiness is expressed
by the phrase na samanupaśyati, “does not perceive”. In accordance with this
rhetoric, the abhedyaprasādas are defined. For sake of briefness I quote only
few characteristic extracts for each of the four items:93
Buddha
utamaṭ́haṇaṭ́hido vi tasag̱ado ◊ ṇa samaṇupaśati ◊ paramaṭ́haṇaṭ́hida vi ◊
tasag̱ada ṇa sa(*maṇupaśati ·) (BajC2, 1A3–4) […]
yado ya śariputra ◊ mamo ṣ̱avag̱a · edehi ca ◊ añehi ca karaṇehi ◊ ṇa
samaṇupaśati · tado budho abhejapras̱a(*deṇa samuṇaga)[d]a bho[di]
(BajC2, 1.A7–8 + 1CD.18)
He also does not perceive the Tathāgata as being in the highest place
(uttamasthāna-sthita). He also does not perceive the Tathāgata as being
in the supreme place (paramasthāna-sthita). […]
And because, Śāriputra, my disciple does not perceive [the Tathāgata (?)]
out of these and other reasons, he is endowed with unbreakable confi-
dence in the Buddha.
Dharma
yoda94 [vi] dharmaviharam eva ṇa samaṇupaśati ◊ tado vi dharmo ◊
abh[e]jopras̱{e}deṇa samuṇag̱ado bhodi (BajC2, 1.A8 + 1CD.18)
Saṃgha
ya[s̱a] yeva tu[a] (*śariputra) dharma ṇa samaṇupaśas̱i ◊ yeṇa dharmeṇa
samuṇag̱ado raha di voharias̱i ◊ evam eva śariputra ◊ yeṇa dharmeṇa ◊
mama ṣ̱avag̱a-sagho ṣ̱avag̱a(*sa)[gha] saṃkho gachati ◊ ta dharmo aria ·
ṣ̱avag̱o ◊ yoṇiśo vavarikṣata ◊ · ṇa as̱ig̱achadi ◊ yado ya ṇa as̱ig̱achadi tado
ya (*sagho a)[bhejo]pras̱adeṇa samu{s}ag̱ado bhoti ◊ (BajC2, 1CD.18–21)
Śīla
yado ya śariputra ◊ mamo ṣ̱avag̱o ◊ ṇa aj̄atvo samaṇupaśati ◊ kudo ⟨*bha⟩
hidho · tado ya (*ṇa aj̄atvasamu)[ṭ́hi]da śilo samaṇupaśati ṇa bhahidha
samuṭ́hida śilo samaṇupaśati ◊ ṇa ajatvabhahidhasamuṭ́hido śilo
(*samaṇupaśati ·) […] (*yavado a)[ria]ṣ̱avag̱o ◊ aribhutehi śilehi samuṇa
g̱ado bhoti · etavado śariputra ◊ caduhi abhejapras̱adehi ◊ samuṇagado
bhoti (BajC2, 1CD.21–23; 2B.13)
Just so, (*Śāriputra), out of these and other reasons my disciple does not
perceive a Buddha, does not perceive a Dharma [or] a Saṃgha. He does
not perceive morality (śīla), concentration (samādhi), understanding
(prajñā), release (vimukti), he does not perceive anything at all. [And]
because he does not perceive [anything], he is not broken. [Therefore] it
is said: ‘[he is] endowed with unbreakable confidence’.
It becomes evident that the Bajaur text explicitly links the well-known
śrāvakayāna concept of the āryaśrāvaka and his a°prasādas with a Mahāyāna
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 97
type of notions. The confidence in the Buddha, the Dharma, the Saṃgha and
the śīla does not arise out of reflexion upon their positive qualities, but out of
their complete non-perception. The entire concept of an āryaśrāvaka and his
characteristic confidences are, thus, clearly reinterpretated in terms of the
theory of emptiness. As in other passages, the text uses well-established cate-
gories of Buddhist thinking and re-defines them according to its own
ideological needs.
The same attitude towards the a°prasādas can be observed in Nāgārjuna’s
Mahāprajñāpāramitāśāstra. In its 20th chapter, the sūtra quotation (and its
commentary) regarding the seventh bhūmi lists twenty things a bodhisattva
should avoid (viṃśatidharmā na kartavyāḥ). Among them are the following
four elements (ed. and tr. Lamotte, 1980: 2421–2422):95
95 The text of the quotation roughly corresponds to the following passage from the
Pañcaviṃsatiprajñāpāramitā: punar aparaṃ subhūte bodhisattvasya mahāsattvasya
saptamyāṃ bhūmau vartamānasya viṃśatidharmā na bhavanti. katame viṃśatiḥ? yad
uta ātmagrāho ‘sya na bhavati sattvagrāho jīvagrāhaḥ pudgalagrāha ucchedagrāhaḥ
śāśvatagrāho nimittasaṃjñā hetudṛṣṭiḥ skandhābhiniveśo dhātvabhiniveśaḥ, āyata
nam ṛddhis traidhātuke pratiṣṭhānaṃ traidhātukādhyavasānaṃ traidhātuke ālayo
buddhaniśrayadṛṣṭyabhiniveśo dharmaniśrayadṛṣṭyabhiniveśaḥ saṃghaniśrayadṛṣṭyabhi
niveśaḥ śīlaniśrayadṛṣṭyabhiniveśaḥ śūnyā dharmā iti vivādaḥ śūnyatāvirodhaś cāsya na
bhavati, ime subhūte viṃśatidharmā bodhisattvasya mahāsattvasya saptamyāṃ bhūmau
vartamānasya na bhavanti (ed. Kimura, 2009: 90).
98 Schlosser And Strauch
As Étienne Lamotte rightly remarks, “les articles 17 à 20 sont une critique dis-
crète contre le noble disciple animé d’une foi éclairée à l’endroit du Buddha, du
Dharma et du Saṃgha, et doué des moralités chères aux saints” (Lamotte, 1980:
2422, note 1). The sūtra text and Nāgārjuna explicitly justify these twenty avoid-
able things by referring to the notion of emptiness. The same kind of critique
against the traditional view of an āryaśrāvaka based on the doctrine of empti-
ness can certainly be stated for the treatment of the abhedyaprasādas in the
Bajaur Mahāyāna sūtra.
4 Conclusion
Within the introductory passage of the “Bajaur Mahāyāna sūtra” (BajC2, Frag
ment 2 of the Bajaur Collection) which is concerned with the emptiness of all
dharmas, the text stresses two categories: the four asaṃhārya-dharmas and
the four abhedyaprasādas. Both are explained in longer passages.
The four asaṃhārya-dharmas, “unconquerable things,” are the last item in a
list of altogether twenty-one groups of four. They circumscribe the highest per-
fect awakening and omniscience of a tathāgata, which make him invincible in
regard to Māra or other adversaries. While the asaṃhārya-dharmas are spe-
cific to Prajñāpāramitā texts, especially the Aṣṭadaśasāhasrikā, all other
categories of the list occur in both Abhidharma and/or Prajñāpāramitā texts,
where they represent the teaching or the characteristics of a buddha. The first
three items of this list are part of the bodhipakṣya-dharmas, and the next four
are related to meditation and include the four noble truths. Up to here every-
thing belongs to the śrāvakadharmas and is also known from canonical or
Abhidharma texts. The next three items characterize a tathāgata, an awakened
being. Elsewhere these items are also called buddhadharmas or lokottara-
kuśaladharmas and they are only known from Mahāyāna texts. This seems to
indicate that the original Abhidharma list had been expanded in order to fit
into an explict Prajñāpāramitā or rather early Mahāyāna context.
The following ten groups of four (as far as they are preserved on the manu-
script) appear to be random selections of terms, which also occur in the
Saṃgītisūtra/-paryāya. None of the different versions of the Saṃgītisūtra/-
paryāya shows a particular parallel in regard to the sequence, selection or
spelling of the terms. However, the Gāndhārī text (BajC2) is in principle closer
to the Pāli version of the Theravādin or the Sanskrit version of the Sarvāstivādin
rather than to the Chinese and Gāndhārī versions of the Dharmaguptaka.
Despite certain parallels, it was not possible to determine a close connec-
tion to any of the Prajñāpāramitā texts. Furthermore, its restriction to groups
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 99
of four distinguishes the list of the Bajaur sūtra from all extant parallels.
Nevertheless, a common background of the lists occurring in the Prajñāpāramitā
literature and the Bajaur Mahāyāna sūtra is clearly discernible. This is con-
firmed by the general diction of the text that uses terms and concepts typical
to Prajñāpāramitā. However, the term itself, prajñāpāramitā, is not mentioned
even once in the preserved portions of the text.
The list of the groups of four leads to a discussion of the four abhedyaprasādas,
“unbreakable confidences”. The extensive treatment of the abhedyaprasādas
appears to be an original trait of the sūtra that cannot be found in other early
Mahāyāna texts. The peculiar variant abhedya replacing the more common
avecca / avetya of other traditions as well as the arrangement of the traditional
three a°prasādas together with morality (śīla) in a group of four can be traced
back to early Abhidharma traditions attested for the Sarvāstivādins. While the
arrangement as a fourfold group seems to be a pan-Sarvāstivādin feature, the
specific interpretation of the a°prasādas as abhedyaprasādas, “unbreakable
confidences,” could have its origins in the circles of early Gandhāran
Abhidharma specialists. From there, however, it seems to have spread out to
various traditions including the translational Chinese literature.
The Bajaur Mahāyāna sūtra uses this well-established concept of
“Mainstream Buddhism” and reinterprets, or to use an expression by Paul
Harrison, “Mahāyāna-ises” it “in terms of the doctrines of Śūnyatā” (1978: 55). It
thus follows a strategy that can also be observed in other early Mahāyāna texts.
As Johannes Bronkhorst (forthcoming) correctly noticed, the dogmatic dis-
courses of early Mahāyāna literature presupposed the existence of a
well-developed Abhidharma tradition, and there seems to be good evidence
that the rich Sarvāstivāda Abhidharma, which was particularly influential in
the Indian northwest, was one of the main sources for the upcoming Mahāyāna
and its terminology and scholarly debates in early Gandhāra.
Abbreviations
Cm Commentary
CPD Critical Pāli Dictionary (Trenckner et al., 1924–)
DĀ Dīrghāgama
DN Dīghanikāya
Dhsgr Dharmasaṃgraha of Nāgārjuna (ed. Müller and Wenzel, 1885)
G. Gāndhārī
GD A Dictionary of Gāndhārī, Stefan Baums and Andrew Glass, eds., <http://
gandhari.org/dictionary.php>
LPG Larger Prajñāpāramitā from Gilgit (partly ed. Conze, 1962, 1974, Zacchetti,
2005)
MAV(Bh) Madhyāntavibhāga-bhāṣya (ed. Nagao, 1964)
MPPŚ Mahāprajñāpāramitāśāstra (T.25.1509)
MS Martin Schøyen Collection, fragment no.
Msa Mahāyānasūtrālaṃkāra (ed. Lévi, 1907)
MN Majjhimanikāya
P. Pāli
PP Prajñāpāramitā
Ps Paṭisambhidāmagga (ed. Taylor, 1905)
PTSD Dictionary of the Pali Text Society (Rhys Davids and Stede, 1921–1925)
PvsP Pañcaviṃśatisāhasrikā Prajñāpāramitā (Ms. Cambridge, ed. Kimura,
1986–2009)
SĀ Saṃyuktāgama
Skt. Sanskrit
SN Saṃyuttanikāya
SplitC Split Collection, fragment no.
SWTF Sanskrit-Wörterbuch der buddhistischen Texte aus den Turfan-Funden
und der kanonischen Literatur der Sarvāstivāda-Schule. Begonnen von
Ernst Waldschmidt. Im Auftrage der Akademie der Wissenschaften zu
Göttingen hrsg. von Heinz Bechert u.a. Göttingen: Vandenhoeck &
Ruprecht.
T. Taishō shinshū daizōkyō 大正新脩大藏經 (ed. Takakusu 高楠, Watanabe
渡邊, and Ono 小野, 1924–1934)
Tib. Tibetan
Vibh Vibhaṅga (ed. Rhys Davids, 1904)
Bibliography
Allon, Mark, and Salomon, Richard (2010). “New Evidence for Mahayana in Early
Gandhāra”, The Eastern Buddhist 41, pp. 1–22.
Abhidharmic Elements in Gandhāran Mahāyāna Buddhism 101
Glass, Andrew (2007). Four Gāndhārī Saṃyuktāgama Sūtras. Senior Kharoṣṭhī Fragment
5. Seattle WA: University of Washington Press.
Harrison, Paul (1978). “Buddhānusmṛti in the Pratyutpanna-Buddha-Saṃmukhāvasthita-
Samādhi-Sūtra”, Journal of Indian Philosophy 6, pp. 35–57.
Harrison, Paul and Hartmann, Jens-Uwe (2014). “Introduction”, In From Birch Bark to
Digital Data: Recent Advances in Buddhist Manuscript Research. Papers Presented at
the Conference, Indic Buddhist Manuscripts: The State of the Field, Stanford, June 15–19
2009. Edited by Paul Harrison and Jens-Uwe Hartmann. Wien: Verlag der
Österreichischen Akademie der Wissenschaften, pp. vii-xxii.
Hinüber, Oskar von (2001). Das ältere Mittelindisch im Überblick (Österreichische
Akademie der Wissenschaften, philosophisch‐historische Klasse, Sitzungsberichte,
467. Band / Veröffentlichungen der Kommission für Sprachen und Kulturen
Südasiens, Heft 20). Wien: Verlag der Österreichischen Akademie der Wissenschaften.
Jaini, Padmanabh S. (1979). Sāratamā. A Pañjikā on the Aṣṭasāhasrikā Prajñāpāramitā
Sūtra by Ratnākaraśānti (Tibetan Sanskrit Works Series). Patna: Kashi Prasad Jayaswal
Research Institute.
Jayatilleke, Kulatissa Nanada (1963). Early Buddhist Theory of Knowledge. London:
Routledge.
Kimura, Takayasu (ed.) (1986). Pañcaviṃśatisāhasrikā Prajñāpāramitā II–III. Tokyo:
Sankibo Busshorin.
Kimura, Takayasu (ed.) (1990). Pañcaviṃśatisāhasrikā Prajñāpāramitā IV. Tokyo: Sankibo
Busshorin.
Kimura, Takayasu (ed.) (1992). Pañcaviṃśatisāhasrikā Prajñāpāramitā V. Tokyo: Sankibo
Busshorin.
Kimura, Takayasu (ed.) (2006). Pañcaviṃśatisāhasrikā Prajñāpāramitā VI–VIII. Tokyo:
Sankibo Busshorin.
Kimura, Takayasu (ed.) (2007). Pañcaviṃśatisāhasrikā Prajñāpāramitā I-1. Tokyo:
Sankibo Busshorin.
Kimura, Takayasu (ed.) (2009a). Pañcaviṃśatisāhasrikā Prajñāpāramitā I-2. Tokyo:
Sankibo Busshorin.
Kimura, Takayasu (2009b). Ś atasāhasrikā Prajñāpāramitā II-1. Tokyo: Sankibo Busshorin.
Kimura, Takayasu (2010a). Ś atasāhasrikā Prajñāpāramitā II-2. Tokyo: Sankibo Busshorin.
Kimura, Takayasu (2010b). Ś atasāhasrikā Prajñāpāramitā II-3. Tokyo: Sankibo Busshorin.
La Vallée Poussin, Louis de (1925). L’Abhidharmakośa de Vasubandhu. Cinquième et si-
xième chapitres. Société Belge d’Etudes Orientiales. Paris: Paul Geuthner.
La Vallée Poussin, Louis de / Pruden, Leo M. (1988–1990). Abhidharmakośabhāṣyam of
Vasubandhu. Translated into English from the French Translation of Louis de La Vallée
Poussin. English Translation by Leo M. Pruden. 4 vols.: I.1988, II.1988, III.1989, IV.1990.
Berkeley CA: Asian Humanities Press.
Lamotte, Étienne (1980). Le Traité de la Grande Vertu de Sagesse de Nāgārjuna
(Mahāprajñāpāramitāśāstra). Tome V. Chapitres XLIX–LII, et Chapitre XX (2e série)
104 Schlosser And Strauch
Chapter 3
1 Introduction
1 DN II 290–315; MN I 55–63; MN III 111–112; SN V 143, 294–297; AN III 450; Nidd I 1, 28, 63, 72,
78, 99; 340, 354, 370, 387; Nidd II 78, 124, 128.
2 The only exception appears in the Niddesa section of the Khuddakanikāya where the trip-
let is not attested within the framework of contemplation but instead in the context of
explication of the doctrine; e.g. in the Kāmasuttaniddesa, the three terms are listed in the
discussion on kāmā: “…ajjhattā kāmā bahiddhā kāmā ajjhattabahiddhā kāmā …” (Nidd
I 1). These specific occurrences of the triad in the Niddesa seem to reflect the commen-
tarial nature of the text.
3 Here references are given for the shorter version in the Majjhimanikāya (MN I 55–63),
with parallel passages from the Dīghanikāya (DN II 290–315) noted when relevant; the
longer Mahāsatipaṭṭhānasutta in the Dīghanikāya (DN II 290–315) differs only in giving a
longer explanation of the four noble truths and hence, this difference is not consequen-
tial for the study of the three modes of contemplation.
4 The same refrain appears after each of the following sections: breathing, once (MN I 56,
DN II 292); postures of the body, once (MN I 57, DN II 292); bodily activities, once (MN I 57,
DN II 293); parts of the body, once (MN I 57, DN II 294); elements, once (MN I 58, D II 294);
cemetery contemplations, nine times (MN I 58, DN II 295–298).
5 MN I 56–58; DN II 292–298); trans. Ñāṇanamoli and Bodhi, 1995: 146: “In this way he abides
contemplating the body as a body internally, or he abides contemplating the body as a
body externally, or he abides contemplating the body as a body both internally and exter-
nally”. For all Pāli quotations in this chapter, the existing English translations are given in
110 Ditrich
The formulaic refrain occurs also after the three other sections – the contem-
plation of feelings (vedanānupassanā),6 the mind (cittānupassanā),7 and
mental objects (dhammānupassanā);8 in total it is attested in 21 places in the
sutta, suggesting very clearly that these three modes were an important or per-
haps even essential aspect of the practice of the four satipaṭṭhānas.
The three terms appear also connected asyndetically (without vā) in close
or distant collocation; e.g. in the Dhammānupassīsutta, when describing how
six dhammas are to be abandoned in order to practice contemplation of the
body, feelings, the mind and mental objects.
The triad appears in the Nikāyas consistently only in the context of meditation
instructions, as three different modes or ways of contemplation in the practice
of the satipaṭṭhanas. Since the suttas discussing the four satipaṭṭhānas seem to
be prescriptive, it means that the monks were instructed to contemplate in one
or another of the three modes (vā). In the Satipaṭṭhānasutta the terms
ajjhattaṃ, bahiddhā and ajjhattabahiddhā are linked by the conjunction “or”
(vā); there is no clear indication that they refer to three successive stages of
contemplation respectively, to be practiced in progression, starting with the
internal, followed by the external and then both the internal and external
mode, which has been suggested by some modern interpreters.10 Rather the
the footnotes (usually from Pāli Text Society versions). However, when those translations
are problematic in relation to the terms and topics discussed, alternative options are
suggested.
6 MN I 59; DN II 298.
7 MN I 59; DN II 299.
8 MN I 60–62; DN II 301–304, 313–314; this refrain follows each of the sections comprising
dhammānupassanā, i.e. five hindrances, five aggregates, six sense-spheres, seven factors
of enlightenment and four noble truths.
9 AN III 450; trans. by Bodhi, 2012: 988: “Bhikkhus, without having abandoned six things,
one is incapable of contemplating the body in the body internally … externally … both
internally and externally … contemplating feelings in feelings … mind in mind … phe-
nomena in phenomena … internally … externally … both internally and externally”.
10 Anālayo, 2006a: 94; cf. Gethin, 2001: 54.
Interpretations of the Terms ajjhattaṃ and bahiddhā 111
implication seems to be that each mode is a distinctive and valid way of con-
templation in its own right.
Neither the Satipaṭṭhānasutta nor other suttas in the Nikāyas explain the
meaning of the three terms nor do they inform on their actual practice, espe-
cially of the third mode. To my knowledge, there is only one exception in the
Nikāyas, in the Janavasabhasutta:11
11 This sutta is referred to by most scholars discussing the interpretations of the internal and
external; e.g. Gethin, 2001: 54; Schmithausen, 2012: 292.
12 DN II 216; trans. by Walshe, 1995: 298: “As he thus dwells contemplating his own body as
body, he becomes perfectly concentrated and perfectly serene. Being thus calm and
serene, he gains knowledge and vision externally of the bodies of others”. The translation
adds “his own” body which is not in the Pāli original (ajjhattaṃ kāye kāyānupassī); the
translation suggested here is to use the word “internally”.
13 Explained in the same way for the other three satipaṭṭhānas (DN II 216).
14 Dīrghāgama fascicle 5, T.1.1: 34c25–35a4; trans.: “There being internal body contempla-
tion, knowledge of other bodies arises. There being internal feeling contemplation,
knowledge of other feelings arises. There being internal mind contemplation, knowledge
of other minds arises. There being internal mind-object contemplation, knowledge of
other mind-objects arises”. I am grateful to Rod Bucknell for providing this reference and
translation from the Chinese. Cf. Schmithausen, 2012: 292, note 9.
112 Ditrich
This sutta explicates, among other attributes, two aspects for each of the five
aggregates, internal or external (ajjhattaṃ vā bahiddhā vā) but does not talk
about their contemplation. Only in the context of contemplation is the third
mode (ajjhattabahiddhā) given as well; for example, in the Satipaṭṭhānasutta,
in the section on dhammas,17 the contemplation of the arising and passing
away of each aggregate is to be conducted in one of the three modes, as reiter-
ated in the refrain.
Similarly, in the Bojjhaṅgasaṃyutta the five hindrances are described as
internal and external:
15 The translations “oneself” and “others” are more common rendering for the pair whereas
“externally” and “internally” are used for the triad; e.g. in the English translation of the
Satipaṭṭhānasutta (DN II 290–315) by Walshe, 1995: 335–350, the triad is translated
ajjhattaṃ “internally,” bahiddhā “externally,” and ajjhattabahiddhā “internally and exter-
nally;” whereas in the same Nikāya, in the Sangītisutta (DN III 249) the pair ajjhattaṃ and
bahiddhā is translated as “yourself” and “others” respectively (ibid., p. 500).
16 MN III 19; trans. by Ñāṇanamoli and Bodhi, 1995: 888: “Bhikkhu, any kind of material form
whatever, whether past, future, or present, internal or external, gross or subtle, inferior or
superior, far or near – this is the material form aggregate”.
17 MN I 60–63.
18 SN V 110; trans. by Bodhi, 2000: 1603: “And what, bhikkhus, is the method of exposition by
means of which the five hindrances become ten?” Whatever sensual desire there is for the
internal is a hindrance; whatever sensual desire there is for the external is also a hin-
Interpretations of the Terms ajjhattaṃ and bahiddhā 113
But when instructions on the contemplation of the hindrances are given these
are to be contemplated in one or another of the three modes.19 The pair
ajjhattaṃ and bahiddhā in the Nikāyas most commonly refers to a range of
pairs in opposition, such as internal-external, inward-outward, inside-outside.20
The English rendering of the terms as pairs of the opposites personal–imper-
sonal or oneself-others seems to draw mainly on the interpretation of the
aṭṭhakathā texts, to be discussed below (2.2). It has to be noted here that since
the pair and the triad occur in specifically different contexts, the semantic
ranges inferred from the usage of the pair (ajjhattaṃ, bahiddhā) should not be
simply, and without caution, inferred and applied to the interpretations of the
meaning of the triad (ajjhattaṃ, bahiddhā, ajjhattabahiddhā), especially of the
compound ajjhattabahiddhā which seems to refer to some other, a specific
mode or method of contemplation.
2.1.3 bahiddhā
The term bahiddhā is most frequently attested in a pair with ajjhattaṃ. When
bahiddhā occurs on its own, it usually refers to things being external, outside,
or aside.21 Another common occurrence of bahiddhā is in conjunction with ito,
referring to those people that are outside the Buddhist teachings or commu-
nity (ito bahiddhā); e.g. in the Brahmajālasutta:
2.1.4 ajjhattaṃ
The term ajjhattaṃ occurs more frequently on its own than bahiddhā does. It
usually signifies “inwardly, internally,” often referring to states of mind; e.g.
drance. Thus what is spoken of concisely as the hindrance of sensual desire becomes, by
this method of exposition, twofold”.
19 For example, MN I 60–63.
20 Cf. PED s.v. ajjhatta.
21 For example, in the Janavasabhasutta: … na cassa bahiddhā ghoso niccharati… “… its
sound does not carry outside …” (DN II 211).
22 DN I 16; trans. by Walshe, 1995: 75: “And whatever ascetics or Brahmins are Eternalists and
proclaim the eternity of the self and the world, they do so on one or other of these four
grounds. There is no other way”. Cf. DN II 151, 283; AN III 206, 372; AN IV 25, 27, 136; MN I
323; MN II 121, 122; Nidd I 249, 397; SNII 133; SN V 229, 230; Th I 141.
114 Ditrich
23 MN I 8.
24 DN III 49.
25 MN I 181.
26 Namely, sensual desire (kāmacchanda), aversion (byāpāda), sloth and torpor (thīna
middha), restlessness and worry (uddhaccakukkucca) and doubt (vicikicchā).
27 Then the sutta continues instructions on how an unarisen hindrance can arise, an arisen
hindrance can be removed and the future arising of the removed hindrance can be pre-
vented.
28 MN I 60; trans. by Ñāṇanamoli and Bodhi, 1995: 151: “Here, there being sensual desire in
him, a bhikkhu understands: ‘There is sensual desire in me’; or there being no sensual
desire in him, he understands: ‘There is no sensual desire in me”.
29 iti ajjhattaṃ vā dhammesu dhammānupassī viharati, bahiddhā vā dhammesu dhammānu
passī viharati, ajjhattabahiddhā vā dhammesu dhammānupassī viharati (MN I 60, DN II
301).
Interpretations of the Terms ajjhattaṃ and bahiddhā 115
Similarly, ajjhataṃ occurs on its own under the contemplation of the seven
factors of enlightenment (bojjhaṅgapabbam)30 where for each factor a monk
is instructed to know whether the factor in question is present in him (atthi me
ajjhattaṃ) or not (natthi me ajjhattaṃ).31 Once again, each of the seven factors,
perceived as internal phenomena, are to be contemplated in one or another of
the three modes, as instructed in the refrain following this section.
2.1.5 ajjhattikabāhira
The two stems also appear frequently in the adjectival forms ajjhattika
“internal” and bāhira “external”. They are attested either on their own or
in a dvandva compound, in each case with specific meanings and usages of
the terms, depending on the context. When compounded in a dvandva they
always refer to the six internal and external sense-spheres (āyatana);32 e.g. in
the Satipaṭṭhānasutta the compound ajjhattikabāhira occurs in the section
on sense-spheres (āyatanapabbaṃ), referring to the six internal and external
sense-spheres (āyatana) respectively. The sutta lists all the six senses and their
objects (cakku and rūpa, etc.) and the fetters arising dependent on them; e.g.
for the eye sense:
The Satipaṭṭhānasutta does not directly explain the meaning of the two adjec-
tives; however, ajjhattika “internal” is commonly interpreted as referring to the
six senses and bāhira “external” to their respective objects.34 In the case of the
sixth sense-sphere, the internal sense-sphere is the mind (mano) and the
external sense-spheres are dhammas which are the objects arising through the
mind-door (mano).35 The sutta instructs that the entire set of sense-spheres –
internal (the senses) and external (objects of the senses) – are to be
contemplated either internally, or externally, or both internally and externally;36
e.g. in the case of the sixth sense-sphere, both the internal sense, the mind
(mano), and its corresponding external sense-sphere, phenomena (dhamma),
are to be contemplated either internally, or externally, or internally and
externally.
When the two adjectives are not compounded but occur in close or distant
collocation they are usually used as attributes describing or explaining various
aspects of the phenomena; e.g. in the Mahāhatthipadopamasutta, the adjec-
tives refer to two aspects of the element of earth as internal or external:
pathavīdhātu siyā ajjhattikā, siyā bāhirā “The earth element may be internal,
may be external”.37
2.1.6 ajjhattika
The adjective ajjhattika always occurs in close or distant collocation with
bāhira; e.g.:
34 Namely, internal: eye (cakkhu), ear (sota), nose (ghāna), tongue (jivhā), body (kāya), mind
(mano); external: visible form (rūpa), sound (sadda), odor (gandha), flavor (rasa), tangi-
ble object (phoṭṭhabba), phenomena (dhamma); cf. DN III 243; Ps I 287.
35 The later subcommentary, the Līnatthapakāsanā Ṭīkā by Dhammapala, further explicates
that the sixth internal sense-base is the bhavaṅga, the so-called life-continuum: chaṭṭassa
pana bhavaṅgamanasaṅkhāto manāyatanekadeso uppatti dvāraṃ (Soma, 1981: 133). Mod-
ern interpreters such as Ñāṇamoli, 1980: 159, suggest that ajjhatikāyatana represents the
“organisation of experience” and bahiddhāyatana the “experience as organised”. Cf.
Anālayo, 2006a: 215.
36 iti ajjhattaṃ vā dhammesu dhammānupassī viharati, bahiddhā vā dhammesu dhammānu
passī viharati, ajjhattabahiddhā vā dhammesu dhammānupassī viharati (MN I 61).
37 MN I 185.
38 AN II 212; trans. by Bodhi, 2012: 586: “There are, bhikkhus, these eighteen kinds of craving
related to the internal and eighteen kinds of craving related to the external”.
Interpretations of the Terms ajjhattaṃ and bahiddhā 117
2.1.7 bāhira
The adjective bāhira, when attested on its own, does not refer to the senses but
mostly refers to “external, outside, outward;” e.g. bāhiraṃ cammakāyaṃ “outer
hide;”39 bāhirāni vā kammantāni paṭivekkhituṃ “or attending to outside
works;”40 bāhirehi paccatthikehi paccāmittehi “by external adversaries and
enemies”.41 The adjectival stem bāhira appears very frequently in compounds
(e.g. bāhira-assāda, bāhira-samaya, santara-bāhira) and in secondary nominal
derivations (e.g. bāhiraka, bāhirima). As an adverb bāhire, it usually refers to
those who are outside the Buddhist order or teachings.
Some semantic inferences may be drawn for the context dependent use of
the terms ajjhattaṃ and bahiddhā in the Nikāyas, studied at the syntagmatic
level. The usage of the triad is very specific, with a very restricted semantic
range: it is attested only in the instances where instructions are given on the
three modes of contemplations and this is also the only context for the occur-
rence of the compound ajjhattabahiddhā. Similarly, the use of the compound
ajjhattikabāhira is restricted only to the sense-spheres. The use of the pair
(ajjhattaṃ and bahiddhā) is less restricted, appearing in a range of contexts,
explicating various aspects of the teachings. The stylistic patterns of formulaic
repetitions of the terms – in dvandva, asyndeta or constructions with the con-
junction vā – seem to reflect the style of the ancient Indian tradition of oral
transmission of sacred texts, particularly the transmission of the Vedas where
the specific stylistic and linguistic patterns in the Vedic language act as semi-
otic markers, related to the religious beliefs and ritualistic acts.42 The Indian
Buddhist oral transmission may have adopted the already established stylistic
and linguistic patterns and markers, such as the usage of dvandva compounds,
for indicating specific contexts such as the modes of contemplation, especially
in relation to the four satipaṭṭhānas.43
Although the study of lexical collocational patterns informs on the use of
the two terms in distinct contexts, most evidently in the case of the triad, the
Nikāyas do not explicate the meanings of the terms in question, nor do they
39 MN III 275.
40 AN I 69.
41 AN IV 106.
42 Ditrich, 2010; and forthcoming.
43 Although Buddhist oral literature may not have fully followed the rather precise conven-
tions of the Vedic oral transmission and exhibits, through variations and divergences,
many characteristics similar to ancient epic transmissions (as pointed out by Cousins,
1983: 1–11), it is likely that many mnemonic formulae, stylistic and linguistic devices were
adopted and adapted by early Buddhists from the already well established Vedic tradition.
118 Ditrich
49 Vism 473; trans. by Ñāṇanamoli, 1994: 537: “It is internal in the sense of one’s own that
should be understood here as internal and that of another person as external”.
50 Cf. As 46.
51 Vism 440; trans. by Ñāṇanamoli, 1994: 484: “In the fourth triad, insight-understanding ini-
tiated by apprehending one’s own aggregates is interpreting the internal. That initiated by
apprehending another’s aggregates or external materiality not bound up with the facul-
ties, is interpreting the external. That initiated by apprehending both is interpreting the
internal and external. So it is of three kinds as interpreting the internal, and so on”.
120 Ditrich
52 The largest number of attestations of the three terms in collocation is in the Vibhaṅga,
followed by the Dhammasaṅgaṇi, and a very few instances in the Dhātukathā and the
Paṭṭhāna.
53 Dhs 2.
54 Trans. by Rhys Davids, 2012: M2: “States that belong to one’s self; are external to one’s self;
are belonging or external to one’s self. States that have for an object one’s self; an object
external to one’s self; an object that is both”. This translation is based on paragraphs 1044–
1049 (Dhs 250).
55 Frauwallner, 1964: 67, comments on the list of the attributes (“Eigenschaften Mātṛkāḥ”) as
being artificially contrived, and claims that ajjhattabahiddhā was added later as the third
aspect to the original pair. This explanation is problematic in relation to the evidence in
the suttas (see 2.1) as well as in the Vibhaṅga where the group of the three terms (as
adverbs) is consistently incorporated into the texts on the four satipaṭṭhānas.
Interpretations of the Terms ajjhattaṃ and bahiddhā 121
objects of contemplation (e.g. the sixteen states of the mind, five hindrances,
five aggregates, seven factors of enlightenment etc.),56 indicating that these
were viewed as either essential, most relevant or sufficient for the develop-
ment of meditation practice. It is noteworthy that out of the twenty-two
triplets listed in the mātikā there is only one triplet (ajjhattā, bahiddhā,
ajjhattabahiddhā) that occurs adverbially, as the three modes of contempla-
tion, in the refrain of the Satipaṭṭhānasutta, indicating that for some reason it
was viewed as an essential aspect of the satipaṭṭhāna practice; this may be fur-
ther underlined by the positioning of the three modes at the forefront in the
beginning of the refrain in the Satipaṭṭhānavibhaṅga (Vibh 193–207) (dis-
cussed in 3.2.1 below).
The two triplets listed in the mātikā are further explained in the Nikkhepa
kaṇḍa section of the Dhammasaṅgaṇi:
The internal dhammas are explained as those of one’s own (sattānaṃ) which
are internally, individually bound – the five khandas. By way of contrast, the
external dhammas are presented as those for other beings, other people
(parasattānaṃ parapuggalānaṃ), and dhammā ajjhattabahiddhā are simply
presented as those dhammas that are both (tadubhayaṃ). Although the
Dhammasaṅgaṇi, as distinct from the Nikāyas, comments on the meaning of
59 This explanation is adopted in the commentaries, e.g. Vism 440; As 46, 325.
60 Dhs 188; trans. by Rhys Davids, 2012: 251: “Which are the states that have an object of
thought concerning the self? Conscious states and their mental properties, which arise in
connection with states of the self. Which are the states that have an object of thought
concerning that which is external [to the self]? Conscious states and their mental proper-
ties, which arise in connection with that are external [to the self]. Which are the states
that have an object of thought concerning that which is ‘personal-external’? Conscious
states and their mental properties, which arise in connection with states that are ‘per-
sonal-external’”.
61 Dhs 42–53; trans. by Rhys Davids, 2012: 54: “When, that he may attain to the heavens of
Form, he cultivates the way [thereto], and unconscious of any part of corporeal self, but
seeing external objects to be limited …”. This translation which follows Buddhaghosa’s
comments (As 188, 189, 191) is problematic and a clearer rendering of ajjhattaṃ may be
simply “internally”.
Interpretations of the Terms ajjhattaṃ and bahiddhā 123
they are listed among attributes of the elements, e.g. in the description of
the earth element: … kakkhaḷattaṃ kakkhaḷabhāvo ajjhattaṃ vā bahiddhā vā
upādiṇṇaṃ vā anupādiṇṇaṃ vā – idaṃ taṃ rūpaṃ pathavīdhātu.62 The term
bahiddhā has several occurrences on its own in the Nikkhepakaṇḍa,63 referring
to people outside the Buddhist teachings, whereas ajjhattaṃ does not appear
outside its collocation with bahiddhā.
The adjectives ajjhatika and bāhira are listed in the mātikā section among
twofold designations (dukamātikā): ajjhattikā dhammā, bāhirā dhammā.64
The two terms are attested only in the Rūpakaṇḍa section, in the same seman-
tic range as in the Nikāyas (2.1.5–2.1.7): they refer to internal and external
sense-spheres respectively,65 or appear among the pairs of attributes related to
rūpa.66 In distinction to the Nikāyas, ajjhatika and bāhira do not appear com-
pounded (ajjhattikabāhira) but only in asyndetic collocation.
3.1.2 Commentaries
In the last section of the Dhammasaṅgaṇi, in a supplement of commentarial
nature, called the Atthuddhāra67 or Aṭṭhakathākaṇḍa,68 the terms ajjhattā,
bahiddhā and ajjhattabahiddhā are further discussed:
It is stated that form (rūpa) which is not bound up with the faculties, and
nibbāna are external whereas all other dhammas may be internal (siyā ajjhattā),
external (siyā bahiddhā), or internal and external (siyā ajjhattabahiddhā).
Then dhammas in relation to objects are explained:
62 Dhs 177; trans. by Rhys Davids, 2012: 222: “That which is hard, rough, hardness, rigidity,
whether it be of the self, or external, or grasped at, or not grasped at”.
63 For example, Dhs 183.
64 Dhs 5.
65 Dhs 129.
66 Dhs 125.
67 As 6, 38.
68 As 409, 422.
69 Dhs 241; trans. by Rhys Davids, 2012: 340: “With the exception of form which is not bound
up with faculties, and Nirvana, all states may be personal or external or personal-external.
[Material] form which is not bound up with faculties, and Nirvana, are both external”.
124 Ditrich
70 Dhs 241; trans. by Rhys Davids, 2012: 340: “Which states have a personal [internal] object
of thought? The sphere of infinite consciousness and the sphere where there is neither
perception nor non-perception – these are states that have a personal object of thought”.
71 Dhs 241–242; trans. by Rhys Davids, 2012: 340: “Which states have an external object of
thought? The threefold and fourfold Jhāna relating to the heavens of Form, whether it
arise as good (karma), as result (of good karma), or as completed thought, also results of
Fourth Jhana, the sphere of infinite space, the four Paths that are the Unincluded and the
four Fruits of the life of the recluse: these states have an external object of thought”.
72 Apariyāpannā means here: not including the worlds of the senses, form and the formless.
73 Dhs 242; trans. by Rhys Davids, 2012: 340–341: “Excepting form, states, good, bad and inde-
terminate relating to the sensuous universe, and the fourth Jhāna relating to the worlds of
Form, whether it arise as good (karma), or as completed thought: all these may be either
a personal [internal] object of thought, or an external object of thought, or a personal-
external object of thought”.
Interpretations of the Terms ajjhattaṃ and bahiddhā 125
The section concludes that it should not be said that the sphere of nothingness
has an internal, external or both internal and external object.75 Form and
nibbāna are without an object.
To summarize, in the commentary on the first triplet, form (rūpa) and
nibbāna are external, all other dhammas may be internal, external, or inter-
nal and external. Apart from form, all dhammas related to the desire realm
and the fourth jhāna (related to the form realm) may have an internal, exter-
nal, or internal and external object. Dhammas having an internal object are
the sphere of infinite consciousness and the sphere of neither perception nor
non-perception. Dhammas having an external object are the threefold and
fourfold jhānas, the result of the fourth jhāna, the sphere of infinite space,
and the four paths and fruitions. Although the text does provide more details
on the meaning of internal and external, it does not give any new informa-
tion on ajjhattabahiddhā and ajjhattabahiddhārammaṇa dhammas; apart
from rūpa and nibbāna, all cittas and cetasikas may be ajjhattabahiddhā, and
all dhammas, apart from rūpa and the fourth jhāna, may have as an object
ajjhattabahiddhārammaṇa.
In the Atthasālinī, the commentary on the Dhammasaṅgaṇi by Buddha
ghosa,76 the comments on the first triplet (dhammā ajjhattā, dhammā
bahiddhā, dhammā ajjhattabahiddhā) focus almost entirely on ajjhatta:
74 Dhs 242; trans. by Rhys Davids, 2012: 341: “But it is not proper to say that the sphere of
nothing whatsoever is all three. Form and Nirvana are without objects of thought”.
75 Cf. Ajjhattadhammāpagamamattatova ākiñcaññāyatanārammaṇassa ajjhattabhāvampi
bahiddhābhāvampi ajjhattabahiddhābhāvampi ananujānitvā ākiñcaññāyatanaṃ na
vattabbaṃ ajjhattārammaṇantipītiādi vuttaṃ (As 423).
76 Three commentaries on the Abhidhamma texts are attributed to Buddhaghosa:
Atthasālinī, Sammohavinodanī and Pañcappakaraṇāṭṭhakathā; cf. von Hinüber 1996: 149–
153; Norman, 1983: 120–130.
126 Ditrich
77 As 46; trans. by Maung Tin, 2013: 60: “In the triplet of ‘personal,’ this word refers to states
which occur after making a locus of selves as though with the understanding ‘we shall
consider or take things thus existing to be we ourselves’. The word ‘personal’ (ajjhatta) has
a fourfold content, namely personal in field, in self-reference, (just) personal, personal in
range”.
78 Ten’ Ānanda, bhikkhunā tasmiṃ yeva purimasmiṃ samādhinimitte ajjhattam eva cittaṃ
saṇṭhapetabbaṃ. Ajjhattarato samāhito ti ādīsu hi ayaṃ gocarajjhatte dissati (As 46).
Trans. by Maung Tin, 2013: 60: “In such sentences as, ‘Ānanda, mind should be well
focussed by that bhikkhu as ajjhattam, namely, only in that symbol of concentration
which has been practised before;’ ‘inwardly rapt (ajjhattarato) and concentrated’ ajjhatta
means ‘personal in field’”.
79 Cf. samādhinimitta (MN I 249) is explained in the commentary (Ps II 292) as suññata
phalasamāpatti, fruition of arahatship; see also MN III 112.
80 Cf. DN II 107–108; SN V 263; AN IV 312; Dhp 101; Ud 64; Th 89.
81 ajjhattaṃ sampasādanaṃ occurs in numerous places in the Tipiṭaka in the context of
jhānas (e.g. DN I 37; DN I 74; MN I 21; SN II 211; AN I 53).
82 Ajjhattaṃ sampasādanaṃ ajjhattaṃ vā dhammesu dhammānupassī viharatī ti ādīsu niya-
kajjhatte (As 46). Trans. by Maung Tin, 2013: 60: “In such passages as, ‘He lives contemplat-
ing states, even among states which are pleasing as ajjhatta, ajjhatta means ‘subjective’”.
83 Cha ajjhattikāni āyatanānī ti ādīsu ajjhattajjhatte (As 46); trans. by Maung Tin, 2013: 60:
“[In] the six ajjhatika sense-organs, ajjhatta means ‘personal’”.
84 Cf. the Mahāsuññata Sutta (MN III 111) where ajjhattaṃ suññataṃ leads to the fruition of
arahatship.
Interpretations of the Terms ajjhattaṃ and bahiddhā 127
85 Ayaṃ kho pan’ Ānanda vihāro Tathāgatena abhisambuddho yad idaṃ sabbanimittānaṃ
amanasikārā ajjhattaṃ suññataṃ upasampajja viharatī ti visayajjhatte issariyaṭṭhāne ti
attho. Phalasamāpattīhi buddhānaṃ issariyaṭṭhānaṃ nāma (As 46). Trans. by Maung Tin,
2013: 60: “In such passages as, ‘This, Ānanda, is the life fully attained by the Tathāgata, to
wit, that he, by disregarding all provocative signs and symbols, has reached the ajjhatta
Void and therein abides, ajjhatta means ‘range’ in the sense of ‘dominion’. The attainment
of Fruition is named the dominion of the Buddhas’”.
86 Tasmā attano santāne pavattā pāṭipuggalikā dhammā ajjhattā ti veditabbā (As 46). Trans.
by Maung Tin, 2013: 60–61: “Hence states occurring in one’s own continuity and pertain-
ing to each individual are to be understood as ‘personal’”.
87 Tato bāhibhūtā pana indriyabaddhā vā anindriyabaddhā vā bahiddhā nāma (As 46).
Trans. by Maung Tin, 2013: 61: “But states outside that personality, whether bound up with
the controlling faculties or not, are termed ‘external’”.
88 Tatiyapadaṃ tadubhayavasena vuttaṃ (As 46). Trans. by Maung Tin, 2013: 61: “The third
term is spoken by virtue of both”.
89 For example: Rūpāni passatī ti bahiddhā pi nīlakasiṇādirūpāni jhānacakkhunā passati.
Iminā ajjhattabahiddhāvatthukesu kasiṇesu jhānapaṭilābho dassito (As 191).
90 Dhs 2.
91 Anantarattiko te yeva tippakāre pi dhamme ārammaṇaṃ katvā pavattanavasena vutto (As
46). Trans. by Maung Tin, 2013: 61: “The immediately following triplet refers to states
(dhamme) occurring in the act of attending to just these three kinds of states (i.e. per-
sonal, external, externo-personal) as objects”.
128 Ditrich
ence to those of another (parassa), and those having internal and external
object (ajjhattabahiddhārammaṇā) with reference to both (tadubhaya).92
Ajjhattabahiddhārammaṇā is consistently interpreted in the commentary as
“both,” except in one instance when it is said that it refers to objects sometimes
internally, sometimes externally (kālena ajjhattaṃ kālena bahiddhā, pavattiyaṃ
ajjhattabahiddhārammaṇaṃ).93 In other commentaries by Buddhaghosa the
same interpretations are given.94
92 For example: … ime terasa cittuppādā attano rūpādīni ārabbha pavattiyaṃ ajjhattāram
maṇā, parassa rūpādīsu pavattā bahiddhārammaṇā, tadubhayavasena ajjhattabahiddhā
rammaṇā (As 425).
93 This passage discusses powers (iddhi) manifesting internal and external objects: Iddhi
vidhacatutthaṃ kāyavasena cittaṃ cittavasena vā kāyaṃ pariṇāmanakāle bahiddhā
rammaṇaṃ attano kumārakavaṇṇādinimmānakāle ca sakāyacittānaṃ ārammaṇakaraṇato
ajjhattārammaṇaṃ, bahiddhā hatthi assādidassanakāle bahiddhārammaṇaṃ, kālena
ajjhattaṃ kālena bahiddhā, pavattiyaṃ ajjhattabahiddhārammaṇaṃ (As 426); cf. Vibh-a
375.
94 For example, in Ps I 249: ajjhattabahiddhā refers alternatively to oneself and to another
(kālena attano, kālena parassa); in Vism 440: ajjhattabahiddhā refers to both (ubhayaṃ).
95 Budddhaghosa lists the texts in the following order: Dhammasaṅgaṇi, Vibhaṅga, Dhātu
kathā, Puggalapaññatti, Kathāvatthu, Yamaka, and Paṭṭhāna (As 3, 21–23).
96 Frauwallner, 1971: 106–121, suggests that the Abhidhamma texts were written in the period
between 200 BCE and 200 CE; he proposes a relative chronology with the Vibhaṅga as the
earliest and the Dhammasaṅgaṇi as the latest text. The supposition by Frauwallner, 1964:
73–80, that the Vibhaṅga may have developed out of an earlier text which was also the
foundation of the Dharmaskandha of the Sarvāstivādins, is developed further by Bronk-
horst, 1985: 305–320, who argues, with rather convincing evidence, that “original” layers of
the Vibhaṅga may well stem back to before the Sūtrapiṭaka that we know at present.
97 See Kragh, 2002: 123–168; cf. Frauwallner, 1971: 69–12; Willemen, Dessein, and Cox, 1998:
10–16, 139–145; Wynne 2005: 35–70.
Interpretations of the Terms ajjhattaṃ and bahiddhā 129
The text presents the four satipaṭṭhānas in five sections (as opposed to 23 in
the Satipaṭṭhānasutta99): contemplation of the body comprising one section
only, i.e. the repulsiveness of the body (paṭikkūlamanasikāra); contemplation
of feelings; contemplation of the mind; contemplation of phenomena com-
prising two sections, i.e. the hindrances (nīvaraṇa) and the factors of
enlightenment (bojjhaṅga). Under each section, the three terms are explained:
98 Vibh 193.
99 MN I 55–63; DN II 290–315.
100 Vibh 193–194.
101 Vibh-a 219.
102 For example, the translation by Thiṭṭila, 2010: 251–252.
130 Ditrich
A monk contemplates the mind internally by knowing the lustful mind “my
mind is lustful” (sarāgaṃ me cittam); externally by contemplating the mind of
this one (or his) (assa cittam) “the mind of this (one) is lustful” (sarāgamassa
cittam); and internally and externally by contemplating “the mind is lustful”
(sarāgaṃ cittam), without relating it to anyone.
In the section on dhammas, only hindrances and factors of enlightenment
are discussed, starting with the hindrances:
108 Thiṭṭila (2010: 255) omits ajjhattabahiddhā in his translation of the sentence kathañca
bhikkhu ajjhattabahiddhā vedanāsu vedanānupassī viharati: “And how does a Bhikkhu
dwell contemplating feeling in feelings?”
109 Vibh 197–198.
132 Ditrich
The internal rūpa is explained as one’s own (sattānaṃ), the external as for any
other beings, other people (parasattānaṃ parapuggalānaṃ).118 The semantic
context for the occurrences of the terms in pair is different to the context for
the triad which is consistently within the framework of contemplation.
3.2.2 Sammohavinodanī
The commentary on the Vibhaṅga, the Sammohavinodanī, ascribed to
Buddhaghosa, gives similar interpretations of the three terms as those in
the Atthasālinī. The three terms are discussed in the commentary on the
Satipaṭṭhānavibhaṅga, starting with the three modes of contemplation of the
body:120
atthi kāye ti idaṃ yasmā na ekantena attano kāyo, nāpi parass’eva kāyo
adhippeto, tasmā vuttaṃ.127
The commentary states that neither one’s own body (attano kāyo) nor anoth-
er’s (parass’eva kāyo) is meant exclusively. Similarly, the Sammohavinodanī
comments on the section on feelings:
The third mode is presented here as neither oneself (neva attā) nor another
(na paro). There is no comment in the Sammohavinodanī on the three modes
in the sections on the mind (citta) and phenomena (dhamma).
The interpretation saying for the third mode to be neither one’s own nor
another’s exclusively is, to my knowledge, found only in the Sammohavinodanī
and is not further discussed or followed up by any other Pāli text of that period.
In other commentaries ascribed to Buddhaghosa the usual interpretation of
the third mode is “sometimes one’s own, sometimes another’s,” or “both”.
Other comments on the three terms in the Sammohavinodanī are presented
in a similar way as in other commentaries by Buddhaghosa. In the commen-
tary on the Khandhavibhaṅga the aggregates are discussed, having internal
and external objects (ajjhattabahiddhārammaṇā) as being sometimes inter-
nal, sometimes external: kālena ajjhattaṃ kālena bahiddhā dhammesu evaṃ
pavattentassa ajjhattabahiddhārammaṇā.129 The adjectival stems ajjhatika
and bāhira are attested in the Sammohavinodanī in reference to internal and
external senses respectively,130 or occur among the pairs of attributes.131
To summarize then, several interpretations for the three terms are
given in the Abhidhamma. In the last section of the Dhammasaṅgaṇi, the
Aṭṭhakathākaṇḍa, the terms are related to different stages of high absorptions:
e.g. dhammas having an internal object are the sphere of infinite consciousness
and the sphere of neither perception nor non-perception; dhammas having an
external object are the threefold and fourfold jhānas, the result of the fourth
jhāna, the sphere of infinite space, and the four paths and fruitions (3.1.2). The
third mode is given a different interpretation in the Vibhaṅga: ajjhattabahiddhā
128 Vibh-a 268; trans. by Ñāṇamoli, 1996: 330: “Internally and externally: he applies the mind
at one time to his own and at another time to another’s feelings. In the [last] section,
because neither self nor other is specified, therefore in order to point out the mere laying
hold of feeling …”.
129 Vibh-a 44; Cf. As 426, Ps I 249, Vibh-a 375.
130 See Vibh-a 51.
131 See Vibh-a 55, 276.
136 Ditrich
132 Gethin, 2001: 92–99; Anālayo, 2006a: 92–99; Kuan, 2008: 117–119; Schmithausen, 2012: 291–
303.
133 D III 103, noted by Anālayo, 2006a: 93; cf. Schmithausen, 2012: 294–295.
134 Anālayo, 2006a: 93–94; cf. Schmithausen, 2012: 295.
135 Anālayo, 2006b: 247.
136 Idha, bhikkhave, bhikkhu araññagato vā rukkhamūlagato vā suññāgāragato vā nisīdati (M
I.56).
Interpretations of the Terms ajjhattaṃ and bahiddhā 137
indicates that there would be no other people to observe, especially since the
monks were instructed to sit alone in meditation and develop concentration
and mindfulness.137 The three modes of contemplation are an integral compo-
nent of most texts on contemplation, retained to a large extent in the early
schools of Buddhism, and seem to have originated from a very early stage of
Buddhist tradition and hence, it can be presumed that they were initially
addressing wandering mendicants rather than large groups in monasteries.
Thus the instructions on the three modes would probably target primarily
ascetics who would practice in silence and solitude, not observing other
people.
The interpretation of the term ajjhattabahiddhā is even more puzzling: nei-
ther the Nikāyas not the commentaries by Buddhaghosa shed any light on
what this mode of practice involves. Interpretation of the three modes, espe-
cially of the third one, seems to have been problematic already at early stages
of canonical transmission, as reflected not only in the Pāli Canon and its com-
mentaries but also in the Chinese translations of the Sarvāstivāda and
Dharmagupta texts.138 The overview by Schmithausen of several texts attesting
the three modes shows that their interpretations vary; e.g. in the Śrāvakabhūmi
the three modes are presented is several ways: internally as referring to one’s
live body, externally to observation of a dead body, and both internally and
externally referring to insight that the dead body was once alive and that one’s
own live body will also become a corpse; or internally refers to one’s live body,
externally to non-living matter, and internally and externally to the body of
another living being; or internal contemplation of feelings refers to one’s own
feelings emerging in one’s own body, external to one’s own feelings emerging
in relation to non-living matter, and internal and external to those emerging in
relation to another.139
A different interpretation of the two modes may be drawn from the Chinese
translation of the *Mahāprajñāpāramitāśāstra by Kumārajīva.140 In the section
on mindfulness of the body, one’s own body (svakāya) is inner, another’s body
(parakāya) is outer; the five organs (indriya) are the inner body whereas the five
137 There are many instances in the suttas where the delight of seclusion is recommended, e.
g. AN V 134–135.
138 For example, the Sarvāstivāda version of the Satipaṭṭhānasutta in Chinese, translated
from the Madhyamāgama by Kuan, 2008: 146–151, prescribes, unlike the Pāli Canon, only
two modes – internal and external; e.g. “Thus a monk contemplates body as a body, con-
templates the external body as a body” (2008: 147).
139 Schmithausen, 2012: 296–299.
140 Lamotte, 1970: 1173–1175 (cited by Gethin, 2001: 54), Anālayo, 2006a: 97, Schmithausen,
2012: 296.
138 Ditrich
objects of the senses (viṣaya) are the outer body. The text does not elucidate the
third mode of contemplation, however, it states that the body can be examined
internally and externally simultaneously (yugapat), at the same time (ekakāle),
or as two distinct operations (bhinna).141 In comparison, Buddhaghosa also
interprets the third mode as “at times internally, at times externally” (kālena
attano, kālena parassa)142 which may refer to two distinct operation, or as
“both” (ubhayam),143 however, it is unlikely that this would refer to simultane-
ous contemplation since he states that both cannot occur at once (ekasmiṃ
kāle panidaṃ ubhayaṃ na labbhati).144 The *Mahāprajñāpāramitāśāstra dis-
tinguishes bodily feeling as external and mental feeling as internal, and the
feelings associated with the [first] five consciousnesses as external, and those
linked to the mind consciousness as internal. The mind is external when it has
for its object an external dharma, and internal when its object is an internal
dharma. Mental consciousness (manovijñāna) is the internal mind, and the
[first] five consciousnesses (pañcavijñāna) are the external mind. Notably, the
text does not elucidate the third mode (external and internal).145
Most modern scholars and practitioners, attempting to interpret ajjha
taṃ and bahiddhā, especially in terms of meditation practice,146 follow
Buddhaghosa, often with some modifications: e.g. only the internal mode
can be practiced as direct contemplation;147 the external mode is to be con-
templated by inference;148 the external mode refers to bodily feelings at skin
level, and internal to the feelings deeper in the body.149 The meaning of the
5 Conclusion
Abbreviations
Abbreviations and the quotation system of Pāli sources follow the Critical Pāli
Dictionary (Epilegomena to vol. 1, 1948, pp. 5*–36*, and vol. 3, 1992, pp. II–VI). The
numbers in the quotations of Pāli sources refer to the volume and page of the Pāli Text
Society edition (e.g. MN I 21 refers to the Majjhimanikāya, vol 1, p. 21).
Bibliography
Hinüber, Oskar von (1996). A Handbook of Pali Literature (Indian Philology and South
Asian Studies 2). Berlin: De Gruyter.
Kragh. Ulrich L. (2002). “The Extant Abhidharma-Literature”, The Indian International
Journal of Buddhist Studies 3, pp. 123–168.
Kuan, Tse-fu (2008). Mindfulness in Early Buddhism: New Approaches through Psychology
and Textual Analysis of Pali, Chinese and Sanskrit Sources. London and New York:
Routledge.
La Vallée Poussin, Louis de, and Thomas, Edward J. (eds.) ([1916–1917] 2001). Mahāniddesa.
London: The Pāli Text Society.
Lamotte, Étienne (1970). Le Traité de la Grande Vertu de Sagesse de Nāgārjuna
(Mahāprajñāpāramitāśāstra). Vol. 3. Louvain-la-Neuve: Université de Louvain,
Institut Orientaliste.
Mahāsi, Sayadaw (1994). The Progress of Insight: A Modern Pali treatise on Buddhist
Satipaṭṭhāna Meditation. Trans. by Nyānaponika. Kandy: Buddhist Publication
Society.
Maung Tin, Pe ([1920–1921] 2012–2013). The Expositor (Atthasālinī): Buddhaghosa’s
Commentary on the Dhammasaṅgaṇi, the first book of the Abhidhamma Piṭaka. Rev.
by Mrs Rhys Davids. 2 vols. Bristol: The Pāli Text Society.
Morris, Richard and Hardy, Edmund (eds.) ([1885–1900] 1999–2013). Aṅguttaranikāya.
5 vols. London: The Pāli Text Society.
Müller, Edward (ed.) ([1897] 2011). Atthasālinī. Rev. by L.S. Cousins. London: The Pāli
Text Society.
Müller, Edward (ed.) ([1885] 2001). Dhammasaṅgaṇi. London: The Pāli Text Society.
Ñāṇamoli Bhikkhu (1980). A Thinker’s Notebook. Kandy: Buddhist Publication Society.
Ñāṇamoli Bhikkhu ([1956] 1994). The Path of Purification (Visuddhimagga) by Bhantā
cariya Buddhaghosa. Kandy: Buddhist Publication Society.
Ñāṇamoli Bhikkhu (1996). The Dispeller of Delusion (Sammohavinodanī). Rev. by
L.S. Cousins; Nyanaponika Mahāthera and C.M.M. Shaw. 2 vols. Oxford: Pāli Text
Society.
Ñāṇanamoli Bhikkhu and Bodhi Bhikkhu (1995). The Middle Length Discourses of the
Buddha: A New Translation of the Majjhima Nikāya. Boston MA: Wisdom Publication.
Ñāṇapoṇika Thera (1962). The Heart of Buddhist Meditation. London: Rider.
Norman, Kenneth R. (1983). Pāli Literature (A History of Indian Literature VII, 2).
Wiesbaden: Otto Harrassowitz.
Oldenberg, Henrik and Pischel, Richard (eds.) ([1883] 1966). Theragāthā and Therīgāthā.
London: Pāli Text Society.
Rhys Davids, Caroline A.F. ([1900] 2012). A Buddhist Manual of Psychological Ethics.
Bristol: Pāli Text Society.
Rhys Davids, Caroline A.F. (ed.) ([1904] 2003). Vibhaṅga. London: The Pāli Text Society.
Interpretations of the Terms ajjhattaṃ and bahiddhā 145
Rhys Davids, Caroline A.F. (ed.) ([1920–1921] 1975). Visuddhimagga. London: The Pāli
Text Society.
Rhys Davids, Thomas W., and Carpenter, Joseph E. (eds.) ([1890–1911] 1995–2007).
Dīghanikāya. 3 vols. London: The Pāli Text Society.
Rhys Davids, Thomas W., Carpenter, Joseph E. and Stede, William (eds.) ([1929–1932]
1968- 1971). Sumaṅgalavilāsinī, London: Pāli Text Society.
Schmithausen, Lambert (2012). “Achtsamkeit ‘innen’, ‘außen’ und ‘innen wie außen’”, In
Achtsamkeit: Ein buddhistisches Konzept erobert die Wissenschaft – mit einem Beitrag
S.H. des Dalai Lama. Edited by Michael Zimmermann Christof Spitz and Stefan
Schmidt. Bern: Hans Huber, pp. 291–303.
Soma Thera ([1941] 1981). The Way of Mindfulness. Kandy: Buddhist Publication Society.
Stede, William (ed.) ([1918] 1988). Cullaniddesa. London: The Pāli Text Society.
Steinthal, Paul (ed.) ([1885] 1982). Udāna. London: The Pāli Text Society.
Sujato Bhikkhu (2005). A History Of Mindfulness: How Insight Worsted Tranquility in the
Satipatthana Sutta. Taipei: The Corporate Body of the Buddha Educational
Foundation.
Takakusu, Junjirō 高楠順次郎, Watanabe, Kaigyoku 渡邊海旭, and Ono Gemmyō 小野
玄妙 (eds.) (1924–1934). Taishō shinshū daizōkyō 大正新脩大藏經, Tokyo: Taishō
issaikyō kankōkai.
- T.1.1: Dīrghāgama, Chang ahan jing 長阿含經, Buddhayaśas and Zhu Fonian 竺佛
念
Thiṭṭila, Sayadaw U. (2010). The Book of Analysis (Vibhaṅga). Pāli Text Society (PTS
Translation Series No. 39). Bristol: Pāli Text Society.
Trenckner, Vilhelm and Chalmers, Robert (eds.) ([1888–1902] 2013). Majjhimanikāya.
3 vols. London: The Pāli Text Society.
Walshe, Maurice (1995). The Long Discourses of the Buddha: A Translation of the Dīgha
Nikāya. Boston MA: Wisdom Publications.
Willemen, Charles, Dessein, Bart and Cox, Collett (1998). Sarvāstivāda Buddhist Scholas
ticism (Handbook of Oriental Studies 2/11). Leiden: Brill.
Woods, James H., Kośambi, Dharmananda, Horner, Isaline B. (eds.) ([1922–1938] 1976–
1979). Papañcasūdanī, Majjhimanikāyāṭṭhakathā of Buddhaghosa. 5 vols. London:
The Pāli Text Society,
Wynne, Alexander (2005). “The Historical Authenticity of Early Buddhist Literatue:
A Critical Evaluation”, Wiener Zeitschrift für die Kunde Südasiens 49, pp. 35–70.
146 Kramer
Chapter 4
This chapter is concerned with two main topics: The first part provides a com-
parison of the four proofs for the existence of the ālayavijñāna found in
Vasubandhu’s Pañcaskandhaka and Sthiramati’s Pañcaskandhakavibhāṣā with
the eight proofs included in the “Proof Portion”1 of the Yogācārabhūmi and the
six proofs of the Mahāyānasaṃgraha. The second part deals with the concept
of manas, trying to trace its development from early Buddhism to the Yogācāra
theory of kliṣṭamanas.
* I would like to thank Ralf Kramer, Constanze Pabst von Ohain, Alexander von Rospatt,
Lambert Schmithausen, and Robert Sharf for offering very helpful comments and corrections
to previous drafts of this chapter. I am also grateful for the support received from the German
Research Foundation (DFG), which enabled me to complete this article.
1 The name “Proof Portion” for this part of the Yogācārabhūmi was introduced in Schmithausen,
1987: 299, note 226.
2 A detailed study of the four proofs is included in Kramer, 2014b: 316–319.
4. Without the ālayavijñāna an individual could not (a) arise in and (b) be
liberated from saṃsāra, because, on the one hand, the process of rebirth
would not be possible and, on the other, the contaminations (kleśa)
could not be removed.
Apart from these two obviously different traditions there is a third alternative
enumeration found in the Mahāyānasaṃgraha,4 which includes the following
six5 proofs:
3 The edition of the Tibetan text of the “Proof Portion” is found in Hakamaya, 1978: 7–15. The
Sanskrit text is preserved in ASBh 11,15–13,20. See also Yid kun 40a-48b. For investigations of
these eight proofs, see Griffiths, 1986: 96–104 and 130–138; Schmithausen, 1987: 194–196; and
Waldron, 1995: 16–18. Moreover, a very insightful study of the eight proofs and their possible
meditative implications has recently been published by Nobuyoshi Yamabe (see Yamabe,
2015).
4 MSg I.29–55.
5 As already pointed out in Schmithausen, 1987: 402, note 710, the sixth proof is not included in
the systematic enumeration of five proofs (the first three concerning “defilement” [saṃkleśa]
and the latter two relating to “purification” [vyavadāna]) stated in advance in MSg I.29.
148 Kramer
Although these three lists of proofs are formally different, the arguments
they include are partly similar and may be classified into three main
categories.
6 Schmithausen, 1987:195.
7 PSkV 55b6ff. See also Kramer, 2014b: 318.
8 For a detailed discussion of these arguments, see Waldron, 1995: 26f. Notably, the reasons given
in the Mahāyānasaṃgraha to support the idea that the ālayavijñāna is necessary for the
process of rebirth in all the three spheres of the world and for sustaining the body during the
whole life are more multifaceted than those in the Pañcaskandhaka(vibhāṣā). MSg I.37, for
instance, mentions the concept of vijñāna “nourishing” the body, i.e. securing the subsistence
of corporeal matter (on this function, see also Schmithausen, 1987: 70), as one of the four
“nourishments” (āhāra).
Some Remarks on the Proofs of the “Store Mind” 149
9 PSkV 55a6ff.
10 Schmithausen, 1987: 44.
11 Griffiths, 1986: 102f. See also Yamabe, 2015: 153–155, and Schmithausen, 2014: 17–21.
12 On this function, see Schmithausen, 1987: 44.
150 Kramer
the “Proof Portion” emphasizes the fact that the ālayavijñāna keeps the body
alive during these states, whereas the Pañcaskandhaka(vibhāṣa) focuses on the
assumption that actual perception can reappear after a person has risen from
unconscious states only on the basis of the store mind. It is therefore possible
that the eighth proof of the “Proof Portion” (i.e. the indispensability of the
store mind in the moment of the mind gradually leaving the body at the point
of death) is not mentioned in the Pañcaskandhakavibhāṣā because this issue is
only related to the vitalizing mechanism (and not to continuity).
Summarizing the results of our investigation so far, the following proofs
have been subsumed under the first category concerned with the ālayavijñāna’s
relevance for the presence of the mind in the body:
13 PSkV 55a1f.
14 PSkV 55a4f. See also Kramer, 2014b: 318.
Some Remarks on the Proofs of the “Store Mind” 151
ted issue, namely the condition that only an entity having the nature of the
ālayavijñāna can be the recipient of the imprints (vāsanā) made by the actual
sense perceptions (MSg I.30).
As mentioned above, some of the proofs which I have assigned to this sec-
ond group – because they are also concerned with the pravṛttivijñānas – refer
to a slightly different issue, namely the simultaneous arising of several vijñānas.
These proofs are only dealt with in the Yogācārabhūmi, namely in proofs 2, 3,
and 5, and have no parallels in the other texts. As already indicated by Lambert
Schmithausen, the arguments presented in these “proofs” are not directly prov-
ing the existence of the ālayavijñāna but are aiming to show that several
vijñānas, i.e. the pravṛttivijñāna(s) and the ālayavijñāna, may occur at the same
time.15 According to the first of these proofs, more than one actual perception
may occur at the same time because there are cases when someone wants to
see, hear, taste etc. simultaneously. Neither the meaning of this argument nor
the reasons for its being indicated as “beginning” (ādi) in the summarizing
verse at the beginning of the “Proof Portion” are entirely clear. Paul Griffiths
seems to take the latter to mean that ālayavijñāna is necessary as the basis for
the first moment (ādi?) of a particular actual perception. His example is “the
occurrence of visual consciousness in the mind of a man whose eyes have been
closed for hours”.16 Moreover, Griffiths understands the argument in the sense
that ālayavijñāna is needed as the immediate condition for simultaneously
arising sense perceptions.17 In contrast, Lambert Schmithausen considers it
not to be intended as a direct proof of the ālayavijñāna but of the appropriate-
ness of the assumption that the pravṛttivijñāna(s) and the store mind can exist
simultaneously since even several pravṛttivijñānas can occur at the same time.18
The second of this category’s proofs (i.e. proof 3) is adduced to show that
mental perception (manovijñāna) could not function clearly if it could not
arise at the same time as the other sense perceptions, the sense data of which
the manovijñāna analyses. In order to illustrate this issue, the thinking about
present objects of the sense perception (i.e. analyzing them by manovijñāna)
is compared with the thinking about remembered objects. While mental per-
ception is explained to function clearly with regard to the first, it is considered
to be unclear with regard to the latter objects.19 The last of the proofs of this
20 See Griffiths, 1986: 101f, 135f., and Schmithausen, 1987: 90, 196, 386f., note 631a. Griffiths
considers this proof and the fourfold vijñapti mentioned in it to be parallel with a passage
in the Madhyāntavibhāga and seems to understand vijñapti here in the context of the
vijñaptimātratā (see especially p. 102). In contrast, Schmithausen points out that this pas-
sage in the “Proof Portion” does not presuppose the “representation only” concept and
refers merely to the cognitive functions of the eight vijñānas.
21 The fact that it was important in the early developmental stages of Yogācāra thought is
discussed in Schmithausen,1987: 45f.
22 PSkV 49b4.
23 TrBh 102, 9ff.
24 See, for instance, MN I 112.
25 See Schmithausen, 1967: 122f.
Some Remarks on the Proofs of the “Store Mind” 153
sampaṭicchana
The sources compared above all assume the manovijñāna to arise subsequent
to the sense perception, the data of which it investigates, determines, and
conceptualizes.32 However, in a number of other Yogācāra sources, such as
the Viniścayasaṃgrahaṇī, the Saṃdhinirmocanasūtra, and, as already indi-
cated above, the Triṃśikā and its commentary by Sthiramati, an alternative
system is being taught. According to these sources, all six kinds of perception
may arise at the same time and manovijñāna is always active (except for some
unconscious and meditative states) with or without any of the other five sense
perceptions.33
Remarkably, the two views of a “multilayered” manovijñāna on the one hand
and the simultaneously arising pravṛttivijñānas on the other seem to exclude
each other. Texts which list the various stages of manovijñāna do not mention
the simultaneity of the perceptions, and works mentioning the simultaneity
do not analyze manovijñāna into various stages.34 What is even more strik-
ing is the fact that Sthiramati uses both systems in his commentaries without
discussing this obvious contradiction explicitly. As already mentioned, in the
Pañcaskandhakavibhāṣā he sticks to the first model whereas in his Triṃśikā
commentary he states the simultaneity of perceptions. In the latter he even
explains the reason why others do not accept the possibility of more than
one pravṛttivijñāna at the same time, without indicating that he himself
also propagates the concept of succeeding single moments of mind in the
Pañcaskandhakavibhāṣā.35 The argument that is adduced against the simul-
taneity of several pravṛttivijñānas says, according to Sthiramati, that in some
cases there would not be enough immediate conditions (samanantarapraty-
aya) for more than one vijñāna.36 This refers to the possibility that there is, for
instance, only one visual perception in the present moment, but two or more
perceptions in the following moment. In such a case, the samanantarapraty-
aya for only one perception would be available and the second would lack an
immediate condition. Sthiramati objects to this opposing view by stating that
the whole cluster of the preceding vijñānas functions as the samanantarapra-
tyaya for the following cluster of vijñānas, no matter how many perceptions it
includes.37
33 TrBh 104,13–15.
34 However, it should be noted that, for instance, in the Saṃdhinirmocanasūtra the mano
vijñāna that accompanies the sense perceptions is characterized as vikalpaka (Saṃdh V.4:
rnam par rtog pa’i yid kyi rnam par shes pa).
35 It should be noted, however, that Sthiramati does not explicitly state in his Pañcaskand-
haka commentary that the simultaneous existence of several vijñānas is impossible. For
passages in which such a simultaneous existence is explicitly denied, see Schmithausen,
1987: 16f., note 303.
36 TrBh 102,9–104,10.
37 Interestingly, both models of perception consider manas to be the samanantarapratyaya
for the following mental moment. Thus, the question arises whether the concept of
manas is used with two different connotations in the present context. For a more detailed
discussion of the concept of manas, see below.
156 Kramer
Coming back to our investigation of the second category of proofs for the
existence of the ālayavijñāna, it can be stated that the following proofs are
concerned with the ālayavijñāna’s relation to the pravṛttivijñānas:
The fact that the “Proof Portion” does not include a discussion of the topic of
purification and liberation in connection with the ālayavijñāna probably
shows that the debates surrounding the concept of the store mind have
changed in the course of time and that the aspect of purification became rel-
evant only at a later developmental stage.
The arguments of this category of proofs can be subdivided into two sub-
groups. The first of them is concerned with the problem that the storage of
seeds of contaminations would be impossible without the ālayavijñāna. It is
discussed in the first proof of the Mahāyānasaṃgraha and in the second part
of the fourth proof of the Pañcaskandhakavibhāṣā. The problem originates
from the assumption that the contaminations and their seeds cannot exist in a
moment of mind that counteracts the contaminations. Thus, as soon as the
counteracting mind has arisen in a person on the path to liberation, there is,
without the existence of the ālayavijñāna, no possibility to store the seeds of
those contaminations which have not been removed yet. Thus, a multi-layered
mind stream is needed for the removal of the contaminations, allowing a par-
allel existence of seeds and their antidotes. This seems to be the main concern
of the argument in the Mahāyānasaṃgraha. The argumentation in the Pañca
skandhakavibhāṣā appears to have a slightly different focus or at least to
include another closely related but somewhat different point, namely the con-
cept that in order to remove the contaminations, the seeds of the latter and
their antidotes have to be existent in the person simultaneously (otherwise
Some Remarks on the Proofs of the “Store Mind” 157
they could not be removed by the antidotes).38 Sthiramati explains that the
contaminations have to be abandoned either in their active forms or as seeds.
The first alternative is not acceptable since, as mentioned above, a moment of
the counteracting mind cannot contain a contamination. Thus, it must be the
seed of the contamination that is present, but the seed can neither exist in the
counteracting mind itself. Therefore a parallel “store” is needed that allows the
simultaneous existence of the seed of the contamination and its antidote.39
The second subgroup of arguments assigned to this third category com-
prises proofs 4 and 5 of the Mahāyānasaṃgraha. These proofs are also
associated with the process of liberation but go beyond the explanations pre-
sented in the context of the first subgroup. Proof 4 deals with the assumption
that a moment of mind belonging to the desire realm (kāmadhātu) cannot
be the seed of the mind belonging to the form realm (rūpadhātu). Therefore
there must be a seed for the latter stored in the ālayavijñāna. The fifth proof
of the Mahāyānasaṃgraha is concerned with the question where the supra-
mundane forms of the mind originate from, and it has already been discussed
in some detail by Lambert Schmithausen.40 In this context, the store mind is
regarded to be necessary in order to store the (mundane) “imprint of listen-
ing” (śrutavāsanā) which originates from the true reality (dharmadhātu) and
later becomes the seed of supramundane insight (or factors leading to it). This
imprint is left in the mental series of a person who listens to the Buddhist doc-
trine and “coexists” with the ālayavijñāna without becoming part of it like milk
coexists with water.41
1.4 Summary
Summing up the results of the comparison of the three lists of proofs the fol-
lowing distribution across the three categories emerges:
38 PSkV 57a3ff.
39 See also Kramer, 2014b: 318f.
40 Schmithausen, 1987: 79f.
41 Schmithausen, 1987: 370, note 580.
158 Kramer
As indicated above, the second part of this chapter is concerned with the
concept of manas. In the Pañcaskandhaka and in Sthiramati’s commentary,
manas is first explained as the basis for the following moment of mind.42
However, Vasubandhu also mentions that the main function of manas in the
context of the vijñānaskandha is to be the vijñāna that has the ālayavijñāna
as its object and that is always associated with the contaminations of “wrong
attitude towards the self” (ātmamoha), “false view of the self” (ātmadṛṣṭi),
“conceited conception of the self” (ātmamāna), and “self-love” (ātmasneha).43
In his comments on this passage, Sthiramati further specifies this vijñāna as
the “contaminated mind” (kliṣṭaṃ manaḥ) and explains that the latter con-
tinuously takes as its object the ālayavijñāna in the form of the self (ātman).44
According to Sthiramati, the manas which is the basis for mental perception
(manovijñāna) is different from the manas which has the nature of conceit
(manyanā).45
In canonical sources manas seems sometimes to be regarded as a kind of
a “mental sense faculty” that experiences its objects. But the main function
of this faculty was probably to reflect on the data provided by the other sense
faculties.46 In Abhidharmic context this view of manas as a mental sense organ
corresponding to the five material sense faculties has been given up, possibly
because of its close affinity to the concept of a permanent self or ātman. The
fact that such a mental sense organ was not accepted is obvious for instance
from statements in the Abhidharmakośabhāṣya. There it is explained that
there is no manas (or manodhātu) besides the six vijñānadhātus and that
manas is only the designation for one of the six vijñānas that has just passed
(anantarātīta).47 Moreover, the text states that manas is included in the
42 PSk 16,8: manaḥsanniśrayatām; PSkV 49a3f.: ṣaṇṇāṃ cakṣurādivijñānānāṃ yad yan nirud-
hyate / tat tad anantarasya vijñānasyotpadyamānasyāśrayabhāvena vyavatiṣṭhata ity ataḥ
samanantaracittasanniśrayatām upādāya mana ity ucyate.
43 PSk 17,7f.
44 PSkV 59a5f.
45 PSkV 59b1. See also Kramer, 2014b: 313.
46 See Vetter, 2000: 64, note 135, 101, note 7 and Boisvert, 1995: 120f.
47 AKBh 11,18–21.
Some Remarks on the Proofs of the “Store Mind” 159
48 AKBh 11,25.
49 AKBh 11,26f.
50 MSg I.6.1.
51 An example for an explicit definition of manas as the immediately preceding basis
(samanantarāśraya) for the next moment of perception in general (i.e. of all six vijñānas)
is found in Y 4,6f., 6,5, 7,6 and 22, 8,14, and 11,9f. See also Schmithausen, 1987: 110f., 124.
160 Kramer
they think that the manovijñāna analyses the sense-data perceived in the last
moment while the other perceptions are busy perceiving the objects of the
present moment? Or did they assume that the manovijñāna is able to “access”
the sense-data immediately, that is in exactly the same moment as it is per-
ceived by the sense perceptions? Unfortunately, these questions have to remain
unanswered since there is no explanation given in this regard in the texts
under discussion here.52 It must also be taken into account that the differentia-
tion between the two concepts of a succeeding and a simultaneous manovijñāna
was not always considered in detail when the often stereo-typed definitions of
manas were created. In general, the statements regarding the nature of manas
found in Abhidharmic and Yogācāra works give the impression that the con-
cept was not thought through thoroughly and that several slightly inconsistent
models of manas existed concurrently.
The problem of a missing basis for the manovijñāna was only partly solved
by the assumption of manas as having this function in the form of the imme-
diately preceding mental moment. In the course of further organizational
development another desideratum became manifest in connection with the
manovijñāna, namely the need of a “simultaneously present basis” (sahabhūr
āśrayaḥ) or “dominant condition” (adhipatipratyaya) for the latter. While the
five indriyas provide a continuously present support for the five sense facul-
ties, the manovijñāna’s basis in the form of manas is present only in the instant
immediately before the occurrence of the manovijñāna itself. This inconsis-
tence must have been regarded a weak point of the system – especially as
manas was a mere designation and not a “real” vijñāna or mind on its own.
Some texts, as for instance the Pañcaskandhakavibhāṣā, seem to have ignored
this problem. Sthiramati classifies the four conditions necessary for a per-
ception in the following way: the imprints stored in the ālayavijñāna are the
“causal condition” (hetupratyaya), the vijñāna that has just passed, i.e. manas,
is the “immediate condition” (samanantarapratyaya), the sense objects, visible
matter etc., are “object conditions” (ālambanapratyaya), and the sense facul-
ties (indriya) are the “ruling condition” (adhipatipratyaya). Since the indriyas
most probably do not include manas but only the five sense faculties which
are mentioned in the rūpa chapter of the Pañcaskandhaka,53 Sthiramati does
52 As mentioned above, the third proof of the Yogācārabhūmi seems to presuppose the idea
of manovijñāna accessing the data it analyses directly. However, since the argument is not
explained in detail, it appears difficult to say with certainty if this statement is really aim-
ing at the concept of perception and analysis of a particular object happening at exactly
the same time.
53 PSk 2,3f.
Some Remarks on the Proofs of the “Store Mind” 161
54 Another example is Y 4,6ff., which specifies a sahabhūr āśrayaḥ for the five sense percep-
tions but omits this category silently with regard to the manovijñāna (Y 11,9). Note that the
*Pañcaskandhavivaraṇa (PSkViv 25a6) and the *Pañcaskandhabhāṣya (PSkBh 95a6), two
further commentaries on the Pañcaskandhaka, mention six indriyas as adhipatipratyaya.
55 MSg I.7.2. See also Schmithausen, 1987: 124.
56 See Schmithausen, 1987: 147–152.
162 Kramer
priate way.57 From the point of view of organization, the term manas might
have appeared suitable insofar as the triad of vijñāna, citta, and manas was
already an established structure in the Abhidharmic context.
The analogy of names of the “traditional” and the “new” manas resulted in an
increasing overlapping of their functions. Already in the context of what might
be the first explicit occurrence of the new manas in the Viniścayasaṃgrahaṇī,
the new manas with its specific functions of conceiving and establishing the
feeling of identity is also claimed to be the basis of manovijñāna.58 The text
explains that this is stated in the sense of the manovijñāna’s being bound
to cognizing its object as long as manas is not interrupted. A similar idea is
expressed in another part of the Yogācārabhūmi, in which manas is taught to be
the basis for the conceptualizing aspect (vikalpa) of manovijñāna,59 and in the
Mahāyānasaṃgraha, which describes manas as the basis for the defilement
(saṃkleśa) of vijñāna.60 All these statements are clearly based on the assump-
tion of a manas which has the nature of the “new” manas, being responsible
for the concept of “I” and “mine” in a person. This manas was obviously con-
sidered an important constituent of the individual’s conceptualizing process
of everyday phenomena and thus as accountable for the contamination of the
mind. At the same time these characteristics of the new manas were combined
with the traditional function of manas of being a basis, in particular a basis
of manovijñāna. Finally, all these developments and probably also the wish
for giving the kliṣṭamanas an appropriate position within the system of the
four conditions (pratyaya) resulted in the equation of the kliṣṭamanas with the
sahabhūr āśrayaḥ, setting the kliṣṭamanas on one level with the sense faculties
(indriya). Thus, it seems that the final concept of (kliṣṭa)manas resembles the
original idea of manas as a “real” mental faculty.
Finally, it should be mentioned that the Mahāyānasaṃgraha seems to be
the oldest (available) source containing proofs for the existence of the
kliṣṭamanas. The text provides six arguments which aim at showing that the
It is notable that these arguments either are not mentioned at all in later
works or are adopted from the Mahāyānasaṃgraha without any substantial
modifications,63 whereas the proofs of the ālayavijñāna are presented in at
least three different systems. In the Pañcaskandhaka(vibhāṣā), kliṣṭamanas is
treated far less exhaustively than the ālayavijñāna, and neither Vasubandhu
nor Sthiramati mention the proofs for its existence.
3 Conclusion
In conclusion, it can be stated that there are (at least) three different tradi-
tions of ālayavijñāna proofs to be found in Yogācāra texts, which, taking into
consideration the differences in terminology and in the way the arguments
are presented, do not seem to be directly related. Nonetheless, the argu-
ments are dealing with similar issues in general and most of the proofs of
the Pañcaskandhaka(vibhāṣā) have parallels in the other lists. In some cases
61 See MSg I.7. The six proofs have been investigated by Lambert Schmithausen in an unpub-
lished study entitled “Satkāyadṛṣṭi, asmimāna und kliṣṭamanas”.
62 Sthiramati explains the ignorance characterized as āveṇika to be the “inborn” (sahaja)
ignorance, which is not associated with the other contaminations (PSkV 31a5f.).
63 See, for instance, Yid kun 48b-52a.
164 Kramer
the arguments appear to be similar at the first glance, but on closer look it
becomes obvious that their focus is different, as for instance in proofs 1 and
6–8 of the “Proof Portion,” which seem to be mainly concerned with vitalizing
aspects, whereas the corresponding passages in the Pañcaskandhakavibhāṣā
emphasize the aspect of continuity. All these divergences might reflect various
stages of development of the concept of ālayavijñāna. This probably results
from the fact that slightly different functions of the store mind were under
discussion when the relevant Yogācārabhūmi passages were composed and
when the Pañcaskandhaka and its commentary were produced. This idea is
also supported by the fact that some of the proofs, as for instance the proofs
of the “Proof Portion” concerning the simultaneous arising of various vijñānas
are not mentioned at all in the lists of the Pañcaskandhaka(vibhāṣā) and the
Mahāyānasaṃgraha.
The twofold understanding of the concept of manas which manifests itself
in the Pañcaskandhaka(vibhāṣā) is the result of the combination of various
lines of thought that developed within Buddhist philosophy with regard to the
basis of thinking, the contamination of the mind, and the notion of “I”. While
manas was probably considered to be a really existing “mental sense faculty” at
its earliest developmental stage, it was reduced to a mere designation of the
immediately preceding moment of mind in Abhidharmic sources. However, it
never lost its relevance as the basis for the arising of manovijñāna and finally
became, in combination with the Yogācāra concept of kliṣṭamanas (which,
according to Lambert Schmithausen, most probably developed from the desire
to remodel the concepts of asmimāna and the sahajā satkāyadṛṣṭiḥ into a sub-
liminal form of mind), the “simultaneously present basis” (sahabhūr āśrayaḥ)
of the manovijñāna.
Abbreviations
Bibliography
Lamotte, Étienne (1938). La Somme du Grand Véhicule d’Asaṅga. vol. 1, Louvain: Bureaux
du Muséon.
Li, Xuezhu and Steinkellner, Ernst (ed.) (2008). Pañcaskandhaka. Beijing/Vienna: China
Tibetology Publishing House/Austrian Academy of Sciences Press.
Pradhan, Prahallad (ed.) (1975). Abhidharmakośabhāṣyam of Vasubandhu (Tibetan
Sanskrit Works Series 8). Patna: K.P. Jayaswal Research Institute (rev. 2nd ed.).
Rhys Davids, Caroline A.F. (ed.) ([1920–1921] 1975). Visuddhimagga. Pāli Text Society,
London: Routledge & Kegan Paul.
Schmithausen, Lambert (1967). “Sautrāntika-Voraussetzungen in Viṃśatikā und
Triṃśikā”, Wiener Zeitschrift für die Kunde Süd- und Ostasiens 11, pp. 109–136.
Schmithausen, Lambert (1967; unpublished paper). “Satkāyadṛṣṭi, Asmimāna und
Kliṣṭamanas”.
Schmithausen, Lambert (1987). Ālayavijñāna: On the Origin and the Early Development
of a Central Concept of Yogācāra Philosophy. Tokyo: The International Institute for
Buddhist Studies.
Schmithausen, Lambert (2014). The Genesis of Yogācāra-Vijñānavāda: Responses and
Reflections. Tokyo: The International Institute for Buddhist Studies.
Tatia, Nathmal (ed.) (1976). Abhidharmasamuccayabhāṣya. Patna: K.P. Jayaswal Research
Institute.
Trenckner, Vilhelm (ed.) (1888). Majjhimanikāya. London: The Pāli Text Society.
Vetter, Tilmann (2000). The ‘Khandha Passages’ in the Vinayapiṭaka and the Four Main
Nikāyas. Wien: Verlag der Österreichischen Akademie der Wissenschaften.
Waldron, William (1995). “How Innovative Is the Ālayavijñāna: The Ālayavijñāna in the
Context of Canonical and Abhidharma Vijñāna Theory (Part II)”, Journal of Indian
Philosophy 23/1, pp. 9–51.
Yamabe, Nobuyoshi (2015). “A Reexamination of on Being Mindless: Possible Meditative
Implications of the Eightfold Proof of Ᾱlayavijñāna”, In Buddhist Meditative Praxis:
Traditional Teachings and Modern Applications. Edited by KL Dhammajoti. Hong
Kong: Centre of Buddhist Studies, the University of Hong Kong, pp. 137–176.
Yid dang kun gzhi’i dka’ ba’i gnas rgya cher ’grel pa (Tsong kha pa), reproduced in Tsultrim
Kelsang and Nobuchiyo Odani (1986). アーラヤ識とマナ識の研究:クンシ・カン
テル (*A Study of Tsong kha pa’s Treatise on Ālayavijñāna and Manas, the Kun gzhi
dka’ ’grel), Kyoto: Bun’eido, pp. 147–184.
Some Remarks on the Proofs of the “Store Mind” 167
Part 2
Intellectual History
∵
168 Kramer
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 169
Chapter 5
1 Introduction
1 Cousins, 2011.
language of the day. It seems likely that the Sanskrit language played a
greater role in the history of the Theriya school of Southern India and
Ceylon than has been hitherto realized. This no doubt accounts for
some of the increasing North Indian awareness of Theravāda during this
period.
2 Bechert, 1998: 2.
3 Lamotte, 1958: 597; Li, 1996a: 320 and 331; Li, 1995: 119.
4 A different view: Deeg, 2012.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 171
should take it that this Chinese evidence indicates that the Mahāvihāra tradi-
tion was at this time predominant among ‘Sthaviras’ in South India (excluding
Mahāyānist Mahāsaṃghikas) and weak in its homeland or possibly strong only
in its hinterland in the south of the island. It is perhaps worth noting that
Xuanzang is using the terminology of the four fundamental Nikāyas; so when
he refers e.g. to Sāmitīyas, this probably means any of the four or so Pudgalavādin
schools. However, by this date the name ‘Sthavira’ or similar seems to be used
exclusively of the three schools which originate from the three major monas-
teries of Anurādhapura.
Xuanzang’s account is supported by the data provided by another Chinese
pilgrim, some decades later. In particular, Yijing 義淨 notes the complete rejec-
tion of the Mahāsaṃgha in Ceylon and does not record any presence in the
south of the four nikāyas which are grouped as the Mūlasarvāstivādin funda-
mental nikāya. The three ‘Sthavira’ schools are mentioned as predominating in
the south, especially Ceylon.5
According to the Nikāyasaṃgraha, Jotipāla came from Jambudīpa. That
is perhaps also implied in the Cūlavaṃsa, if we understand that the king is
specifically mentioned as arranging for him to dwell in the Mahāvihāra. It is
likely then that he came from the major center of non-Mahāyāna Theravādin
tradition in the Tamil country. Given the obvious similarity of the names
Dhammapāla and Jotipāla we should look to some kind of pupil-teacher or
monastic relation between the two.6
Because the two elements that have the [heart-] base as their support
operate bound up with rūpa,
the heart is certainly the support for rūpa in two [kinds of] existence in
dependence upon those two [elements]. (cp. Abhidh-av 674; Sacc 12)
The second topic concerns higher knowledge (abhijñā). Again at Vism-sn III
1098 (to Vism 456) the following gāthā are attributed to Jotipāla-māhimiyo:
Higher knowledge does not give results because it is the reward in this
existence of dhyāna just as the obtaining of wealth [is the reward of] giv-
ing and the rest; also because it is impossible.
The reason that it is impossible is that karma does not give fruit on a dif-
ferent level [to its own] and the object of resultant [citta] of the rūpa
[level] is [sign of] the karma, not such objects as small dhammas.
In AS I a number of closely related discussions in the ṭīkās are cited and some
are translated.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 173
The third topic dealt with in AS I concerns five Sanskrit stanzas on the
pratisaṃvid. Again, in Sumaṅgala’s discussion prior to his final mention of
Jotipāla, we find:
For ease of reference I add a continuous numbering of all the Sanskrit verses
which I attribute to Jotipāla. The first two pādas of stanza nine correspond to
the line cited in Abhidh-av-ṭ. I translate:
The discriminations number four [and] are divided in sequence into the
knowledges of artha, dharma, nirukti and the knowledge of those [knowl-
edges]. (5)
Artha is fivefold by way of (a) what has arisen by a condition; (b) the
meaning of speech; (c) liberation; (d) resultant [citta]; (e) kriyā [citta].
Furthermore, it has three kinds because it may be the result (artha)
which follows from doing, the meaning (artha) which has to be made
known or the goal (artha) which must be attained. (6)
Dharma is fivefold: (a) condition; (b) speech; (c) the path; (d) the white
and (e) the black. Furthermore it is the three kinds of cause: that which
does, that which finds out and that which attains.(7)
7 On the paṭisambhidā/pratisaṃvid, see: Pagel, 1995: 272ff.; 359ff. Also, Samtani, 1971: 53f.; 115ff.;
275 (refs); Griffiths, 1994: 116 note.
174 Cousins
Nirukti is the Māgadhī language; this means that it is what causes one to
understand names.8 Some say that it is sound joined with a form of com-
munication. (8)9
The obtaining [of the discriminations] is [nothing but] the śaikṣa or the
aśaikṣa path because it is due to the absence of confusion at that moment.
But [a discrimination] is subject to āśravas in operation (prayoge)
because it occurs taking artha and the rest as its object. (9)
must have been written either late in the twelfth century or early in the thir-
teenth, i.e. before Vism-sn.
Vism-sn was not the first sanne on Vism. Sumaṅgala refers explicitly to a
Sinhalese commentary on Vism and evidently understands Dhammapāla to
refer to it.11 So some or all of the Sanskrit passages in Vism-sn are likely to be
taken from there and may therefore be much earlier in date. Certainly, there is
nothing which has so far been dated with certainty to a period after the sev-
enth century. The same seems to be the case with the Pāli quotations in
Vism-sn, although there are a small number whose source cannot be identi-
fied.12 This kind of work is of course not intended as an original composition in
the modern sense. Rather the author seeks to present the tradition with subtle
modifications of detail on specific points.
Most of these Sanskrit passages are in verse. Examining these verses we find
three which are unequivocally from brahmanical literature in general and two
which are from grammatical writing.13 These do not name their source, pre-
sumably because it was obvious. From other Buddhist literature, there is one
stanza which apparently derives from the Saundarananda of Aśvaghoṣa, one
from the Vyākhyāvyukti of Vasubandhu, one attributed to Diṅnāgapāda (cor-
rected from Dīghanāgapāda), and two explicitly cited from the Abhidharmakośa.
I omit one other which could be based upon the Kośa.14
There remain fifty-four ślokas (including those already attributed to Jotipāla)
which mostly concern specifically Buddhist topics. There has been a tendency
to assume that these will be from otherwise unknown Buddhist Sanskrit, i.e.
non-Theravādin, authors. This must be mistaken. Many of them describe posi-
6 The Jñeya-saptati-ṭīkā
In his last paper on Buddhist Sanskrit literature in Ceylon, Heinz Bechert refers
to the quotations from Sanskrit works found in Vism-sn as follows:
15 Vism-sn II 479; III 1119 (×3); 1136 (×2); 1190 (×3); 1250; 1344; IV 1688. The first of these is the
grammatical stanza attributed to Rāhulapāda. Assuming that pāda is honorific, his name
would be Rāhula. See Cousins, 2012: 91ff.
16 See previous note.
17 Vism-sn III 1075. Nāgabodhi’s verses reject the notion that the water element is fluid-
ity (dravatā). cf. Mahāvyutpatti (cited LVP) p. 101; Śikṣāsamuccaya p. ; Buddha-c 9.56
& 13.57: apāṃ dravatvaṃ. Abhidh-k I 13: sneho ’bdhātuḥ; Abhidh-dīp p. 12: abdhātunā
dravasnehalakṣaṇena; Abhidh-sam 3: abdhātuḥ katamaḥ niṣyandatā.
18 Vuttañ h’ etaṃ Purātanehi: Vism-mhṭ II 109 = Abhidh-s-mhṭ 151, etc.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 177
Kārikā:
Ojas tu yāpana-hetu snehaḥ praśyâdi-vastunām22 (10)
nutritive essence on the other hand is a liquid which is the reason why
such things as edibles and so on nourish’, (10)
JS-bh:
iti tu-śabdaḥ kṛtyenābdhātos tasya viśeṣa-darśanârthaṃ prayuktaḥ; na hi
tasya yāpanaḥ kṛtyam ābandhana-kṛtyatvāt.
the expression ‘on the other hand’ is employed in order to show that it [i.e.
nutriment] has a different function to the water element; for nourishing
is not the function of the water element, since that has binding together
as its function.23
JS-ṭ:
Praśyādi-vastunām: praśyam annam. Ādi-śabdena pānaṃ parigṛhyate
ubhayenāpi yat kabalaṃ kṛtvâb[h]yavaharyyate,24 bhūta-bhautika-
saṅkhyātaṃ vastu tat pradarśayati. Yasmāt tat sarvaṃ rūpotpādanena
Such things as edibles and so on: an edible is food. Drinks are included
under the heading of ‘and so on’. By means of both he shows that that
which is eaten after making it into lumps is accounted among those
things (vastu) classed as elements and derivative elements. Because all of
that supports the body with its sensory faculties by arousing rūpa, there-
fore the liquid which appears as its sāra is called ‘nutritive essence’
because it is the reason why they nourish.
The work called the Jneya-sampatti-ṭīkā (sic) which has been cited sev-
eral times, as revealed from the quotations, appears to be a text-book of
this school [i.e. the Vijñānavādins]. A bhāṣya and a bhāṣya-ṭīkā of this
work are also cited.
Has the characteristic of nutritive essence: the sāra of the rasa which flows
throughout the body is a specific nutritive essence dependent on the ele-
ments which gives support and strength.
But this was said to show nutritive essence together with its material
basis. Only after obtaining external food as its condition does internal
food arouse materiality. But it is called food (āhāra) because it nourishes
(āharati) materiality. That is why he said: it has the function of nourishing
materiality. After that [he said]: it has the manifestation of supporting this
body because it arouses [kalāpas of] materiality with nutriment as its
eighth component. Because nutritive essence’s function of nourishing
materiality is dependent on external [nutriment] he said: its footing is a
material basis to be nourished.
Abhidh-s-mhṭ 152:
Idañ ca savatthukaṃ katvā āhāraṃ dassetuṃ vuttaṃ. S’endriya-
kāy’opatthambhana-hetu-bhūtā pana aṅgamaṅgānusārī
rasa-hara-saṅkhātā ajjhoharitabbâhāra-sineha-bhūtā ojā idha āhāra-
rūpaṃ nāma. Tathā h’etaṃ
s’endriya-kāy’opatthambhana-hetu-bhāva-lakkhaṇaṃ, oj’-aṭṭhamaka-
rūpâharaṇa-lakkhaṇaṃ vā.
cause of sustaining the body with its faculties, or of nourishing [the clus-
ters of] materiality which have nutriment as their eighth [constituent].26
This comes just after a śloka and a partial śloka on the subject of the relation
between nirvāṇa and the aggregates. The first of these is attributed to a work
named as the Āryasatyāvatāra ‘Entrance to the Noble Truths’.
There is division of materiality and the other [four] into past, [present
and future]. When [past, present and future] have been grouped together,
each is designated as being an aggregate in the specific sense of being a
grouping (rāśi). (11)
This is part of discussions on the nature of the aggregates found both in the
Pāli commentarial sources and in the Abhidharmakośa (see nn. 31 & 32). The
latter cites an authoritative source which states that the meaning of aggregate
is ‘heap’ (rāśi). This must be a citation from a recension of the Peṭaka, since
exactly this statement is found there.27 It also justifies this with reference to a
passage in a sūtra which states that the rūpa aggregate is so-named after group-
ing together (ekadhyaṃ abhisaṃkṣipya) all rūpa – past, present and future,
26 Abhidh-av-ṭ II 140f. (to Abhidh-av 675): Idañ ca savatthukaṃ katvā āhāraṃ dassetuṃ
vuttaṃ. Kabaḷī-kārâhārassa pana ojā idha āhāro nāma. Tenāha: yāya ojāyā ti-ādi. So ca
rūpāharaṇa-sabhāvo upatthambhana-bala-karo aṅga-m-aṅgânusārī rasâharaṇa-bhūto
bhūta-nissito eko viseso. Bāhirāhāra-paccayaṃ paṭilabhitvā eva ajjhattikāhāro rūpaṃ
uppādetī ti ayaṃ ajjhattikāhārassa upanissaya-bhāvenâpi rūpaṃ āharatī ti āha: “oj’-
aṭṭhamakaṃ rūpaṃ āharatī” ti. Ojā aṭṭhamī yassa, taṃ oj’-aṭṭhamakaṃ. Yāya ojāyā ti
attano udayânantaraṃ rūpa-jananato ojā-saṅkhātāya odana-kummāsâdi-vatthu-gatāya
yāya pharaṇa-ojāya. Yattha yatthā ti yamhi yamhi jana-pade, nagarādīsu ca kabaḷīkāro
āhāro ti pavuccati; tab-batthukattā ti adhippāyo.
27 Peṭ 112: rāsattho khandhattho. The Peṭaka was known in the north. See Zacchetti, 2002a,
Zacchetti, 2002b.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 181
within or without, gross or subtle and so on. Most of this passage is found as
part of the stock description of the aggregate of materiality in many Pāli
sources, but the specific reference to ‘grouping together’ is found only in the
Vibhaṅga.28 Nothing in this stanza indicates whether we are dealing with a
Sanskrit śloka based on the Kośa or something which is rendering into Sanskrit
ideas from the Pāli commentaries. However, a few lines later we have another
Sanskrit stanza, probably also from the Āryasatyāvatāra:
Appended to this is a piece of commentary. This suggests that the name of the
work in kārikās may in fact be Āryasatyāvatāra (ĀS). Jñeya-saptati (JS) could be
an alternative name with the Jñeya-saptati-bhāṣya (JS-bh) being its ‘auto-com-
mentary’. Or, the name Jñeya-saptati-ṭīkā might strictly apply only to the
sub-commentary. There could well have been confusion as to this, as manu-
scripts of JS-ṭ would no doubt contain the kārikā together with the commentary.
As to the exact name of the work, it is sometimes given as Jñeya-sampatti in
editions of Vism-sn, but the more probable ‘saptati’ seems to be confirmed by
the form in Sacc-ṭ: Aññeyya-sattati- (see below). The form aññeyya- ‘to be
understood’ may be correct, although unexpected as a rendering of jñeya-.
Perhaps the technical uses of the Pāli word ñeyya- were felt to constrict the
intended meaning in some way.
kārikā = Āryasatyāvatāra
Yet even although it is a real entity, nirvāṇa is not spoken of [as one of the
aggregates] because the meaning is [not] applicable [and] because the
Commentary = Jñeya-saptati-ṭīkā
i.e. because the meaning of grouping is applicable; both because it is not
a real entity and because the meaning is [not] applicable. ‘And’ is in the
conjunctive sense.
But if aggregate has the meaning of separate constituent, would not the
(unwanted) possibility then arise of nirvāṇa being another aggregate?
There is no such possibility because the aggregates are connected with
that which is temporally divided.
Sacc 5–6:
Param’-attho sa-nibbāna-pañca-kkhandh’ ettha rāsito
khandh’-attho ca samāsetvā, vutto ’tītâdi-bhedanaṃ.
Vedanâdīsu p’ ekasmiṃ, khandha-saddo tu ruḷhiyā
samuddâd’-eka-dese tu, samuddâdi-ravo yathā.30
The supreme meaning is nibbāna and the five aggregates. But here such
divisions as past, [present and future] have been combined and the
meaning of ‘aggregate’ has been given as ‘heap’. (5)
Even when there is only a single item as in the case of [the aggregates of]
feeling and so on, the word ‘aggregate’ [has the meaning of ‘heap’] by
convention, just as in such cases as the sound of the sea referring to one
part of the sea. (6)
30 cp. Vibh-a 2: svāyam idha rāsito adhippeto. ayañ hi khandhaṭṭho nāma piṇḍaṭṭho pūgaṭṭho
ghaṭaṭṭho rāsaṭṭho. tasmā ‘rāsilakkhaṇā khandhā’ ti veditabbā. koṭṭhāsaṭṭho ti pi vattuṃ
vaṭṭati; lokasmiñ hi iṇaṃ gahetvā codiyamānā ‘dvīhi khandhehi dassāma, tīhi khandhehi
dassāmā’ ti vadanti. iti ‘koṭṭhāsalakkhaṇā khandhā’ ti pi vattuṃ vaṭṭati. evam ettha
rūpakkhandho ti rūparāsi rūpakoṭṭhāso.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 183
Both these senses are allowed in the Pāli commentaries. The Abhidharmakośa
is familiar with this second sense but rejects it:
Alternatively the meaning is ‘piece;’ for there are those who say ‘we will
give what is to be given in three parts’. That is outside of the Sūtra; for the
Sūtra mentions only the meaning of ‘heap’ with the words: whatever
materiality past, present and future and so on.31
31 Abhidh-k-bh p. 13: pracchedārtho vā; tathā hi vaktāro bhavanti tribhiḥ skandhakair deyaṃ
dāsyāma iti. tad etad utsūtram. sūtraṃ hi rāśyartham eva bravīti “yat kiṃcid rūpam
atītānāgatapratyutpannam” iti vistaraḥ.
32 The notion of khandha as equivalent to rāsi is found in Pāli before Buddhaghosa. See Peṭ
113: Tattha katamo khandh’-attho: samūh’-attho khandh’-attho, puñj’-attho khandh’-attho,
rās’-attho khandh’-attho, taṃ yathā dabba-kkhandho vana-kkhandho dāru-kkhandho aggi-
kkhandho udaka-kkhandho vāyu-kkhandho. Iti evaṃ khandhesu sabba-saṅgaho va evaṃ
khandh’-attho. In fact, it is probably implied by the treatment of the khandhas in the
Yamaka.
184 Cousins
Given that eye discrimination depends on both [the eye and visual
forms], (because of the canonical statement: ‘eye discrimination arises,
conditioned by eye and visual forms’), why is it designated ‘eye discrimi-
nation’ and not ‘form discrimination’?
He replied: what is called designation is made on the basis of a cause
which is not shared, as in the case of a shoot of barley; for the eye is the
unshared cause of eye discrimination. The form base is a shared [cause]
because it is the cause of mind discrimination also.
Closely related to this is another passage from earlier in Vism-sn, also cited by
Junkichi Imanishi:35
For one whose opinion is that the eye sees, it is not every eye that sees.
How so? It is just [eye that is] directed by consciousness. But for one
whose opinion is that consciousness sees, it is not every consciousness
that sees. How so? It is just [consciousness that is] supported by the eye.
And that consciousness does not arise when there is something inter-
posed where there is obstruction of light in such cases as a wall. But
where there is no obstruction [of light] it does arise even when there is
something interposed as in such cases as crystal or a thin layer of cloud.
Therefore it is because consciousness has not arisen that when there is
something interposed the eye does not see, being restricted by what is
interposed.36 Therefore it is said that eye consciousness sees. For eye con-
sciousness has the function of seeing.
Both these passages concern the controversy over vision.37 That seems to have
originally been a debate with the Mahāsāṃghikas who apparently held that it
is the sensitive matter in the physical eye that actually sees.38 Buddhaghosa
and other commentators cite a stanza from the old commentaries:
dvārârammaṇa-saṅghaṭṭ[an]e pana
[cakkhu]-pasāda-vatthukena cittena passati.39
Īdisī pan’ esā ‘dhanunā vijjhatī’ ti-ādīsu viya sasambhāra-kathā nāma hoti.
Tasmā cakkhu-viññāṇena rūpaṃ disvā ti ayam ev’ ettha attho” ti.
Seeing a visible object with the eye: seeing a visible object with eye
consciousness that is capable of seeing visible objects and which is
referred to as ‘eye’ because [the eye] is the cause [of vision]. But the
Porāṇas declared:
Matter does not see40 because it is not part of the mind.
Mind does not see because it is not part of the eye.
But when the object impacts on the [eye] door,
one sees by means of mind whose base is the sensory matter [of the
eye].
But that is what is known as speaking in reference to everything con-
nected41 as in such cases as when one says ‘he wounds with a bow’.
Therefore the only meaning in this case is: ‘seeing a visible object with
eye consciousness’.
Because [the eye] is the cause [of vision]: because it is a particular cause.
For the result is indicated by a particular cause as with a shoot of corn
[or] the sound of a drum. Or, this refers to what is supported by a word for
a support, as [in the expression] ‘the seats make a noise’. Seeing a visible
object with the eye: as to this if the eye could see a visible object, those
experiencing other consciousnesses could see [it]. But that is not the
case. Why? Because the eye is not conscious. Therefore [the Porāṇas]
said: The eye does not see a visible object because it does not have mind. If
consciousness could see a visible object, it could see it even in such con-
cealed places as the other side of a wall because there is no sensory
39 Pādas a & b are tuṭṭhubha; pādas c & d are jagatī. Cakkhu in pādas a & d is intrusive (prob-
ably from a commentary). It is omitted in the second case in some editions. In pāda c read
°saṅghaṭṭane pana with Nidd-a III 141 and Paṭis-a III 676. There is resolution of the first
syllable. In pāda d emend to °vatthukacittena.
40 Or, one does not see a visible object.
41 cf. Kv-a 178: sasambhārakathānayena vuttaṃ. yathā hi usunā vijjhanto pi ‘dhanunā vijjhatī’
ti vuccati, evaṃ cakkhuviññāṇena passanto pi ‘cakkhunā passatī’ ti vutto.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 187
resistance. This is not the case because all consciousness lacks [the abil-
ity to] see. Therefore [the Porāṇas] said: The mind does not see because it
is not an eye. As to this, just as it is consciousness supported by the eye
which sees, not any consciousness at all (and it does not arise when any
such thing as a wall is in the way where there is an obstruction to light,
but where there is no obstruction in such cases [of transparency] as the
interior of a membrane42 or a crystal, it does arise even when there is
something in the way), in the same way it is eye supported by conscious-
ness which sees, not any eye. Hence the words: Seeing a visible object with
the eye were said with reference to the eye supported by consciousness.43
This establishes that the issue was alive in the writings of the early ṭīkā writers
in a form very similar to that we find in JS-ṭ. But it is even closer in Sumaṅgala’s
twelfth century ṭīkā to the Abhidhammāvatāra and there it is associated with
the name of Jotipāla.
Although the eye does not see a visible object, why in that case is it just
consciousness supported by the [eye]. Accordingly it is said that ‘he sees
by means of the sensitive matter of the eye’ because the expression arises
taking the activity of what is supported as like the supports, as in such
[expressions] as ‘the seats make a noise’. For this was said also by the
Elder Ācariya Jotipāla with just this specific meaning. The following is the
agreed answer to the question: does the eye see a visible object or is it
consciousness [which sees a visible object]?
What is more, as to this [it may be objected]: if the eye were to see, the
eye of a person experiencing another kind of consciousness would see.
And consciousness would see even what is concealed by a wall or other
obstruction because there is no sensory resistance. This is not a valid
objection. For someone who holds that the eye sees, it is not every eye
that sees, but only an eye that is directed by consciousness. But for some-
one who holds that consciousness sees, it is not every consciousness that
sees, but only a consciousness that is supported by the eye and that does
not arise in relation to something concealed in a place where there is
obstruction of the light by such things as walls. But in a place where there
is no such obstruction as in such cases as crystal or a thin layer of cloud,
it does arise even in relation to something concealed. Therefore [con-
sciousness] does not see what is concealed because it has not arisen. The
following is the agreed answer to the question: that is to say, eye con-
sciousness sees; for that has the function of seeing.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 189
[Objection] If so, how come that it is stated [in the texts]: ‘seeing a visual
object with the eye’?
[Reply] The meaning in these cases is ‘by means of the [eye] door which
is the cause’. Or rather, the activity of what is supported is applied to the
supports, as in such [expressions] as ‘the seats make a noise’.44
[By a kamma other than the one which gives paṭisandhi] there are caused
[resultants] too, i.e. the tadārammaṇa is caused resultant because of the
words (in the Paṭṭhāna): ‘caused bhav’-aṅga is a condition for causeless
bhav’-aṅga by succession condition’. But some say that in this context
‘caused’ should be taken as referring to two-caused [resultant] only.
However, in the Aññeyya-sattati-ṭīkā it is said that, since ‘caused’ does not
specify [either two-caused or three-caused], it should be taken as both
47 The notion that a caused tad-ārammaṇa is possible for a being with a causeless recon-
necting by a kamma other [than the reconnecting] is found at Dhs-a 288, but it does not
seem that the fifth or sixth century commentaries ever specify whether caused here
means ‘two-caused’ or both ‘two-caused and three-caused’. Probably Jotipāla intends to
argue that if the author of the Abhidhamma Commentary had intended to make it
restricted, he would have specified ‘two-caused’. Since he did not so specify, his intent was
to allow ‘three-caused’ also. Abhidh-av 206f. and 215f. allow only 37 cittas for those reborn
in the four apāya – a figure which excludes both kinds of caused resultant. This would still
leave open the case of a causeless rebirth as a human being and in any case Abhidh-av-ṭ
takes the figure as excluding ‘other kamma’.
48 The Burmese edition must be in error here, by omitting a na or something similar. Com-
pare Abhidh-s-mhṭ 121 where Sumaṅgala gives a briefer version of the same debate: idha
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 191
objection was rejected:] ‘ask the author of the Atthasamāsa himself’. But
others say that the [causeless being] does not get a three-caused
tadālambaṇa because his fundamental connecting (i.e. paṭisandhicitta)
lacks intelligence.
But the Elder Ācariya Jotipāla said that there is a three-caused tadā
rammaṇa even for causeless [beings] because ‘caused bhav’-aṅga’ is
given [in Paṭṭhāna] without specifying. Then he said the following: ‘ask
the author of the [Atthasamāsa] himself,’ as a humorous rejection of the
claim that there are no resultants joined with knowledge in the case [of
causeless beings].
Sumaṅgala goes on to point out that, although this was said humorously, in
fact the right thing to do is to consult an (or the) ācariya i.e. Anuruddha, the
author of both Abhidh-s and Pm-vn and quotes from Pm-vn by name.50 He
then adds that others comment: “just as there is a caused tadārammaṇa for
causeless beings, similarly there is a three-caused tadārammaṇa for two-
caused beings. And in compliance with their understanding, people say that
the rejection of resultants joined with knowledge at this point [in Abhidh-s]
applies only to [the case of] the causeless”.
He then comments that, since there is no authoritative text for this, their
words should be accepted after investigation, as the ācariya has classified the
cittas in exactly the same way (samakam eva) after giving a reason common to
both [causeless and two-caused beings] for the absence of resultants joined
49 As Oskar von Hinüber points out (Hinüber 1996, §346), Abhidh-s-mhṭ was “finished
within the astonishingly short time of 24 days”. This is no doubt best accounted for by
supposing that Sumaṅgala is translating his teacher’s sanne into Pāli and adding material
from an already written Abhidh-av-ṭ.
50 Taṃ pana parihāsa-vasena vuttam pi ācariyaṃ pucchitvā va vijānan’-atthaṃ vutta-
vacanaṃ viya ṭhitaṃ. Tathā hi ācariyen’ev’ ettha kāraṇaṃ Param’-attha-vinicchaye
vuttaṃ: ñāṇa-pākā na vattanti, jaḷattā mūla-sandhiyā ti (Pm-vn 271).
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 193
The five aggregates were taught in this order to reveal: the vessel, the
food, the spice, the maker [of the meal] and its consumer. (13)
(cp. Abhidh-k-bh I 22: bhājanabhojanavyañjanakarttṛbhoktṛbhūtā hi
rūpādayaḥ skandhāḥ; Mahāvibhāṣā).
This comes closely after the passage attributed to the Āryasatyāvatāra and
seems likely to be part of an account of the first noble truth in that work. The
actual simile involved is found in both Sarvāstivādin and Theriya sources. But
if the stanza is indeed part of a work of Jotipāla, its immediate source might be
expected to be the commentaries of the school of Buddhaghosa. But there we
have:
But why are the aggregates spoken of just as five and just in this order? It
should be understood that that it is to show them variously as the vessel,
the food, the spice, the cook and the one who eats, and to expound the
way in which they are gross and subject to defilement. Feeling and label-
ing are taught as separate aggregates, separating them from the aggregate
of constructions, considering them as being the cause of the bases of dis-
pute, the cause of saṃsāra and the cause of kamma [respectively].
If we tabulate the different versions, they seem fairly close and it does not
seem possible to identify a source. The concept seems likely to predate both
Buddhaghosa and the Kośa; so the Sanskrit stanza may be quite old. Or, it could
be a version produced by Jotipāla or elsewhere.
Abhidh-s-sn 106:
Suptaś cūtasya mūle phala-patana-ravāt kaścid unmīlya cakṣur
dṛṣṭvā câdāya caitat phalam atha paritā pīḍayitvā kareṇa
jñatvā pakva-sva-bhāvaṃ kila paribubhuje tad rasâkṛṣṭa-cetāḥ
paścāt tālv-ādi-lagnaṃ rasam api ca saha-śleṣmaṇâsvādya śiśye.
This is a version of the mango simile, used to explain the consciousness pro-
cess in Pāli sources. It is quite close to the version given in the Abhidhamma
Commentary (Dhs-a 271) and even closer to the versions given in the Pāli ṭīkās.
It differs even from the version in the Chinese translation of the Vimuttimagga.
Nothing like it is known from any non-Theravādin source and if the indication
of the Vimuttimagga is anything to go by, it probably differs from any version
current in the Abhayagiri school.
So whether or not Saddhātissa had any source for the attribution to the
Āryasatyāvatāra, it seems certain that these two stanzas are of Mahāvihāravāsin
origin and they might well be the work of Jotipāla.
53 I do not know whether he had a source for this or merely inferred it.
54 Gethin reads paritaḥ.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 197
The syllable dus has the sense of despicable, while the syllable kha is a
synonym for the empty. Combining both together, he spoke of the truth
of duḥkha first among the four.56 (14)
The Lord spoke of the net [of craving] as samudaya, referring to three
things: sam and ud have the two senses of conjunction (saṃyoga) and
arising (utpatti), while aya is a term for cause.57 (15)
Vism-sn 1190:
Nirabhāvo nistṛṇavad bhavati ca rodha iti cārakasyākhyā
tāv abhisamīkṣya tādṛn nirodha ity āha nirvāṇam. (16)
55 Kahrs, 1998.
56 Vism 494: Idha du iti ayaṃ saddo kucchite dissati, kucchitaṃ hi puttaṃ dupputto ti vadanti,
khaṃ-saddo pana tucche, tucchaṃ hi āsaṃ khan ti vuccati. Idañ ca paṭhama-saccaṃ
kucchitaṃ aneka-upaddavâdi-ṭṭhānato tucchaṃ bāla-jana-parikappita-dhuva subha-
sukh’-atta-bhāva virahitato, tasmā kucchitattā tucchattā ca dukkhan ti vuccati.
57 Vism 495: Saṃ iti ca ayaṃ saddo samāgamo sametan ti ādisu saṃyogaṃ dīpeti, u iti ayaṃ
saddo uppannaṃ uditan ti ādisu uppattiṃ, aya-saddo kāraṇaṃ dīpeti, idāñ câpi dutiya-
saccaṃ. Avasesa-paccaya-samāyoge sati dukkhass’ uppatti-kāraṇaṃ, iti dukkhassa
saṃyoge uppatti-kāraṇattā dukkha-samudayan ti vuccati.
198 Cousins
Two stanzas give the reason for the selection of craving as the cause of suffer-
ing. Buddhaghosa simply states that it is the chief cause (mūlakāraṇa), but the
ṭīkās elaborate exactly as in these Sanskrit stanzas:59
We can compare the closely parallel statement in the ṭīkā to the Abhidhamm’-
atthasaṅgaha (Abhidh-s-mhṭ 178):
58 Vism 495: Tatiya-saccaṃ pana yasmā ni-saddo abhāvaṃ, rodha-saddo ca cārakaṃ dipeti,
tasmā abhāve ettha saṃsāra-cāraka-saṅkhātassa dukkha-rodhassa sabba-gati-suññattā.
Samadhigate vā tasmiṃ saṃsāra-cāraka-saṅkhātassa dukkha-rodhassa abhāvo hoti tap-
paṭipakkhattā ti pi dukkha-nirodhan ti vuccati.
59 Vism-mhṭ I 229 = Sp-ṭ I 231: Tattha mūla-kāraṇa-bhāvenā ti santesu pi avijjâdīsu aññesu
kāraṇesu tesam pi mūla-bhūta-kāraṇa-bhāvena. taṇhā hi kammassa vicitta-bhāva-hetuto,
sahāya-bhāvûpagamanato ca dukkha-vicittatāya padhāna-kāraṇaṃ. Vism-mhṭ I 204 = Sp-ṭ
III 177: Kasmā pan’ettha taṇhā va samudaya-saccaṃ vuttā ti? Visesa-hetu-bhāvato. Avijjā hi
bhavesu ādīnavaṃ paṭicchādentī, diṭṭhi-ādi-upādānañ ca tattha tattha abhinivisamānaṃ
taṇhaṃ abhivaḍḍhentī dosâdayo pi kammassa kāraṇaṃ honti, taṇhā pana taṃ-taṃ-
bhava-yoni-gati-viññāṇa-ṭṭhiti-sattâvāsa-satta-nikāya-kula-bhog’-issariyâdi-vicittataṃ
abhipatthentī, kamma-vicittatāya upanissayataṃ kammassa ca sahāya-bhāvaṃ upa
gacchantī bhavâdi-vicittataṃ niyameti, tasmā dukkhassa visesahetu-bhāvato aññesu pi
avijjā-upādāna-kammâdīsu Sutte Abhidhamme ca avasesa-kilesâkusala-mūlâdīsu vuttesu
dukkha-hetūsu vijjamānesu taṇhā va samudaya-saccan ti vuttā ti veditabbaṃ.
60 Two lines of this are cited also at Vism-sn II 480 in a different order.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 199
But why, since there are other causes of suffering, is craving alone stated
as the origin of suffering? Because it is the principal cause. For since it is
the reason for a variety of suffering by being the motivation for a variety
of kamma and by following as the companion of kamma, craving espe-
cially is the cause of suffering.
Coming to the third truth, we find three stanzas related to the debates as to
whether nirvāṇa is a really existing entity in some sense or just the mere
absence of greed, hate and delusion. These too are closely related to the discus-
sions of this topic in the Visuddhimagga and the Abhidhammāvatāra, as well
as in the ṭīkās to those works.61 The first of these is:
Since the existence of the supreme object has been established on the
authority of one who has realized it, the rejection of nirvāṇa by others is
[like] the denial by one born blind of the [existence of] the sun. (19)
At this point Buddhaghosa simply states that it cannot be said that what a fool-
ish puthujjana does not perceive, therefore does not exist (Vism 507).
Sumaṅgala, however, does give the same simile.62 Since he is clearly influenced
by Jotipāla, his source is most probably the Satyāvatāra.
This second of our Sanskrit stanzas makes two points. The first is that
nirvāṇa is strictly speaking not the destruction of greed, hate and delusion,
but rather the cause of that.63 The second is that the same statement is made
about arahatship; so if that statement implies non-existence, arahatship
too would be non-existent. This second point is already made in the discus-
sion in the Visuddhimagga and the Abhidhamma Commentary.64 In fact, the
Visuddhimagga account is for the most part a rewritten version of the material
in the latter. Indeed, the account in the fourth or fifth century Abhidhamma
Commentary is almost certainly copied from earlier material, giving refuta-
tions of the position of the Vitaṇḍavādin(s). The first point, however, does
not seem to be explicitly made in this context until the account in Kassapa’s
Mohavicchedanī (ca. twelfth century).65
This third stanza is responding to the argument that the statement in the
Jambukhādakasutta is an authoritative one, i.e. the sutta is nītārtha – one
whose meaning is explicit. This point does not seem to be spelt out in the com-
mentarial literature, although it is certainly implied by their arguments. It
seems to be first spelt out in the Mūla-ṭīkā.66 It is expanded in the Mahāṭīkā.67
This again seems to fit with an attribution to Jotipāla.
Another stanza that could be part of a work on the truths is given at the
conclusion of the discussion of the aggregate of materiality:
66 suttato muñcitvā ti suttapadāni muñcitvā. añño suttassa attho “mātaraṃ pitaraṃ hantvā”
ti ādīsu (Dhp 294f.) viya āharitabbo, natthi suttapadeh’ eva nīto attho ti attho.
67 Vism-mhṭ II 208: tam pi hi … pe … niddiṭṭhaṃ: tattha yathā rāgādīnaṃ khīṇante
uppannattā arahattaṃ rāgakkhayo ti ādinā niddiṭṭhan ti suttassa neyyatthattā. evaṃ
nibbānaṃ āgamma rāgâdayo khīyantī ti nibbānaṃ rāgakkhayo ti ādinā niddiṭṭhan ti
imassâpi suttassa neyyatthatā veditabbā. cp. Abhidh-av-ṭ II 191.
68 Variant reading: śāstre kṛtaṃ. pāda a: read ajātatvaṃ; pāda c: read lakṣ[y]atvāt.
69 SN II 26: jāti, bhikkhave, aniccā saṅkhatā paṭiccasamuppannā; Paṭis I 51: jāti
paṭiccasamuppannā.
70 Dhs-a 341f.
71 Vism-mhṭ II 112: yadi evaṃ, kathaṃ “jarā-maraṇaṃ paṭicca-samuppannan” ti (SN II 26)
vuttaṃ? yasmā paṭicca-samuppannānaṃ dhammānaṃ paripāka-bhaṅgatāya tesu san-
tesu honti, na asantesu. na hi ajātaṃ paripaccati, bhijjati vā, tasmā taṃ jāti-paccayataṃ
sandhāya “jarā-maraṇaṃ paṭicca-samuppannan” ti pariyāyena suttesu vuttaṃ.
202 Cousins
there is no way to be sure that this stanza comes from Jotipāla’s work. But it is
perhaps more likely than not.
Vism-mhṭ II 132:
Visaya-visayī-bhāva-siddhāya dhammatāya ārammaṇassa abhimukhī-
bhāvena pasādassa tāva ghaṭṭanā hotu, añña-sannissitassa pana
bhav’-aṅgassa calanaṃ kathaṃ hotī ti? Taṃ-sambandha-bhāvato.
bheri-tale ṭhapitāsu sakkharāsu ekissā sakkharāya ghaṭitāya tad-añña-
sakkharāyaṃ ṭhita-makkhikā calanaṃ c’ettha udāharaṇan ti.
72 Cousins, 1981.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 203
existence of the objective field and the possessor of the objective field,
but how does there come to be disturbance (movement) of the life-con-
tinuum that has a different support? Because it is connected with it. And
here the example is this: when grains of sugar are put on the surface of a
drum and one of the grains of sugar is tapped, a fly sitting on another
grain of sugar moves.
Vism-sn III 1106 (cf. Pāli version Abhidh-s-mhṭ 119 = Abhidh-s-sn 119f.):
javanaḥ tat-samāno vā tad-ālambana-cetasā
javo ’nubaddhyate nânyo bāla-dāraka-līlayā (24)
bījâbhāvāt na cordhveṣu tad-ālambanam asya tu
bījaṃ kāmâpta-pākâkhyaṃ pratisandhi-mano matam (25)
pravarttamāna-javanam ucite sthāna eva tat
anubadhnāti dṛṣṭântam āhur atrâpi dārakam (26)
The illustrations given here are specific to the ṭīkā literature. In fact the
Abhidhamm’-atthasaṅgaha Commentary has what looks like a rewritten ver-
sion of these verses in Pāli:
Abhidh-s-mhṭ 119:
janakaṃ taṃsamānaṃ vā javanaṃ anubandhati,
na tu aññaṃ tadālambaṃ bāladārakalīlayā. ||
bījassābhāvato natthi brahmānam pi imassa hi;
paṭisandhimano bījaṃ kāmāvacarasaññitaṃ. ||
204 Cousins
But these illustrations are already found much earlier in the Mahāṭīkā.73 They
seem unlikely to be significantly earlier than the seventh century in date. Two
stanzas on the same general topic follow shortly after.
After small active minds which are pure or rooted in greed and delusion
or kriyā with pleasant feeling,
[there arises] one corresponding retention with pleasant feeling from
among the five [possible]. (27)
After small impure, pure or kriyā minds which are without happiness,
[there arises] one appropriate retention from among the six with neutral
feeling. (28)
The account here is not different in essence to the commentarial account, but
the phrasing is quite close to the Mahāṭīkā.75
Two further verses concern the consciousness process at the time of rebirth:
When there is a pure active mind before decease for one being reborn in
a fortunate destiny,
the very wise decide that there is only a dark active mind for one being
born in a bad destiny. (29)
at the time of [death]76 the mind of one with craving is said to have like a
dream,
the mark of the destiny, the karma or the sign of karma due to his karma.
(30)
In this case the ideas found do not go beyond those found in the commentaries
of the school of Buddhaghosa, but they remain specific to the Theriya tradition.
76 Possibly emend to cutikāle, but if so it is used loosely. The reference is to the last javana
process before death.
206 Cousins
Abhidh-av-ṭ 315:
yathāhu Porāṇā:
jigucchati nāhiriko, pāpā gūthā va sūkaro.
na bhāyati anottappo, salabho viya pāvakā” ti ||
The comparisons with the pig and the moth are already found in the Mahāṭīkā.77
I have not so far found them in any Sanskrit source, but they clearly could
come from a non-Theriya source. But, given the passage in the Mahāṭīkā, a
Theriya origin again seems the most likely.
Three stanzas in the āryā meter are concerned with faith (śraddhā/saddhā).
They give comparisons of faith with a hand, wealth and seeds. These similes
are quite old with canonical roots and various mentions in the commentarial
literature.78 They are further expanded in the works of Dhammapāla, the ṭīkā
writer. I have not come across them so far in Northern Buddhist literature as
similes applied to śraddhā, but they may well be found there somewhere.79
Just as one who has vision does not take hold even of what is given as
food as wealth,
so likewise he does not take hold of the ārya wealth given by the hand of
faith. (32)
Just as one who eats the seeds will have no shoot even in a well-ploughed
area of land,
so one who consumes the seeds of faith will have no skilful root. (34)
Two stanzas in the āryā meter concern jealousy (īrṣyā/issā) and miserliness
(mātsarya/macchariya). The first is:
Those for whom jealousy’s flame ignites the fuel of other people’s gains,
do not find the elixir of the gladness which is to be enjoyed by?brahmas.
(35)
Abhidh-av-ṭ I 319:
ten’ev’ etaṃ vuccati:
issânala-sikhā yesaṃ hadaye jalatîdha te
n’eva vindanti pāmojjaṃ sambuddhâdīhi sevitan ti.
Those for whom the flame of the fire of jealousy burns here in the heart,
do not find the gladness practiced by the Fully Awakened Ones, etc.
The mind accomplishes nothing even in dream with the joy that springs
from relinquishing for those in whose mind the unlucky throw of miserli-
ness is obtained. (36)
208 Cousins
The beautiful [occurs] with wished for things [as its object] due to appro-
priate directing towards the object and due to [familiarization] and due
to restricting the mind to the skilful and changing it [to that]. (37)
But surely this wished for object is the basis for greed? How has this citta
which is called ‘skilful’ arisen? [The answer is:] due to restricting (niyam-
ita), due to changing (pariṇāmita), due to familiarization (samudācāra),
and due to directing (ābhujita). Someone’s mind is restricted to the per-
formance of the skilful with the idea ‘I should do only what is skilful,’ after
turning from the occurrence of the unskilful it is changed into doing just
the skilful, it makes just the skilful familiar by doing it again and again
and appropriate directing [of the mind] occurs due the prior conditions
(upanissaya) of dwelling in a suitable place, attendance upon good peo-
ple, hearing the saddhamma, having merit performed previously and so
on. Because of this that which is called ‘skilful’ has arisen to him by means
of restricting, changing, performing and directing.81
The passage is slightly rewritten in the Mahāṭīkā with ‘directing’ in first place
as in our Sanskrit verse.82 It is again slightly rewritten by Sumaṅgala, returning
to the order of the Abhidhamma Commentary. But he then adds a Pāli version
of our Sanskrit verse:
Abhidh-av-ṭ I 182f.:
Hoti c’ettha:
niyāmapariṇāmehi, samudāciṇṇatāya ca
ñāṇapubbaṅgamābhogā iṭṭhe pi kusalaṃ siyā” ti ||
It is striking that the version we have here in Sanskrit is closest to that in the
Mahāṭīkā.
Two stanzas address the question of precisely how many resultant cittas
arise as the results in another life of the various skilful cittas. Part of the issue
turns precisely on the difference between skilful cittas with three causes (hetu)
and those with two. In the Northern Abhidharma systems wisdom is a univer-
sal accompaniment of all cittas and the idea of skilful states without wisdom is
81 Dhs-a 75: Nanu c’etaṃ iṭṭhārammaṇaṃ lobhassa vatthu? Kathaṃ etaṃ cittaṃ kusalaṃ
nāma jātan ti? Niyamitavasena pariṇāmitavasena samudācāravasena ābhujitavasenā ti.
Yassa hi ‘kusalam eva mayā kattabban’ ti kusalakaraṇe cittaṃ niyamitaṃ hoti, akusalap-
pavattito nivattetvā kusalakaraṇe yeva pariṇāmitaṃ, abhiṇhakaraṇena kusalasamudācāren’
eva samudācaritaṃ, patirūpadesavāsasappurisūpanissayasaddhammasavanapubbekata
puññatādīhi ca upanissayehi yoniso ca ābhogo pavattati, tassa iminā niyamitavasena
pariṇāmitavasena samudācāravasena ābhujitavasena ca kusalaṃ nāma jātaṃ hoti.
82 Vism-mhṭ II 115: nanu ca iṭṭhārammaṇaṃ lobhassa vatthu, kathaṃ tattha kusalaṃ
hotī ti? na-y-idam ekantikaṃ iṭṭhe pi ābhogādivasena kusalassa uppajjanato. yassa hi
catusampatticakkasamāyogâdivasena yoniso va ābhogo hoti, kusalam eva ca mayā kat-
tabban ti kusalakaraṇe cittaṃ niyamitaṃ, akusalappavattito ca nivattetvā kusalakaraṇe
eva pariṇāmitaṃ, abhiṇhakaraṇavasena ca samudācaritaṃ, tassa iṭṭhe pi ārammaṇe
alobhādisampayuttam eva cittaṃ hoti, na lobhādisampayuttaṃ.
210 Cousins
not admitted. So this discussion too can only derive from the Theriya schools.
In fact, we can go further than this.
When commenting on the eight mahākusala cittas at their first occurrence
in the Dhammasaṅgaṇi, the Abhidhamma Commentary offers what it refers to
as the mātikā of the discourse setting out resultant [citta]. This is part of a
series of named topics in a sequence already established in an earlier com-
mentary or commentaries. In the course of this sequence the positions
attributed to three named elders are set out in brief.83 The positions of the first
two are set out in these two Sanskrit verses. The three Elders in question appear
to have lived in Ceylon during the first century BCE or just after.84 Their views
will have been known to the author of the Abhidhamma Commentary from
sources dating from the early centuries CE. These two stanzas also can only
have come from the Theriya tradition:
The superior kind which has three causes would result in both the sets of
eight which have been mentioned together with the connexion [citta].
But the inferior kind of citta would result in [cittas] with two causes lack-
ing those with intelligence. (38)
The superior kind of citta lacking knowledge is like that. But the inferior
kind would result in the set of eight without causes. Others declare that
their results are twelve, ten and eight. (39)
Sacc 124–126:
Ñāṇa-yutta-varaṃ tattha, datvā sandhiṃ ti-hetukaṃ
pacchā paccati pākānaṃ, pavatte aṭṭhake duve ||
tesu yeva nihīnaṃ tu, datvā sandhiṃ du-hetukaṃ.
deti dvādasa pāke ca, pavatte dhī-yutaṃ vinā ||
evaṃ dhī-hīnam ukkaṭṭhaṃ, sandhiyañ ca pavattiyaṃ
hīnaṃ pan’ ubhayatthâpi, hetu-hīne va paccati ||
83 Dhs-a 267–288; Abhidh-av 378ff.; Paṭis-a III 574 cites Paṭis against both the first two.
84 Mori, 1989: 126f., 129f., 134f.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 211
Among these the excellent citta joined with knowledge, after giving
three-caused connexion, subsequently in process gives two sets of eight
results.85
But a lower kind of citta among those, after giving a two-caused connex-
ion, subsequently in process gives twelve results (without the [four] that
have intelligence).
This distinguishes inferior and superior kinds of these cittas. The supe-
rior kind can give results of its own kind, whereas the inferior kind cannot
give the best results.
The superior [kind of citta] lacking intelligence [gives results] in the
same way both in connexion and in process, but the inferior kind [of
ñāṇavippayuttacitta] gives in both [connection and process] only the
[eight results] without causes.
85 i.e. eight great resultants and eight skilful resultant causeless cittas.
86 Vism-mhṭ II 118: Ta-y-idaṃ aṭṭha-vidham pi kāmâvacaraṃ kusala-cittaṃ rūpârammaṇaṃ
yāva dhammârammaṇan ti chasu ārammaṇesu yaṃ vā taṃ vā ālambitvā upekkhā-
saha-gatâhetuka-kiriyā-mano-viññāṇa-dhātânantaraṃ kāya-dvārâdīhi tīhi dvārehi
kāya-kammâdi-vasena uppajjatī ti veditabbaṃ. Tattha ñāṇa-sampayuttāni cattāri
yadā ti-hetuka-paṭisandhiṃ uppādenti, tadā soḷasa vipāka-cittāni phalanti. Yadā pana
du-hetukaṃ, tadā dvādasa ti-hetuka-vajjāni. Ahetukaṃ pana paṭisandhiṃ ti-hetukāni
na uppādent’eva; du-hetukāni pana du-hetuka-paṭisandhi-dāna-kāle dvādasa, ahetuka-
paṭisandhiṃ dāna-kāle aṭṭha phalanti. Ti-hetukā pana paṭisandhi duhetukehi na hoti
yeva. ‘Aṭṭha phalantī’ ti ce taṃ paṭisandhiṃ janaka-kamma-vasena vuttaṃ. Aññena pana
kammunā “sa-hetukaṃ bhav’-aṅgaṃ ahetukassa bhav’-aṅgassa anantara-paccayena
paccayo” ti (Paṭṭh III 42) vacanato sa-hetukam pi vipāka-cittaṃ ahetuka-paṭisandhikassa
hoti yeva. Imasmiṃ ca pakkhe balavatā paccayena uppannaṃ asaṅkhāraṃ, dubbalena
sasaṅkhāran ti veditabbaṃ. Ye pana āgamanato ca vipākassa asaṅkhāra-sasaṅkhāra-
bhāvaṃ icchanti, tesaṃ matena dvādasa, aṭṭha ca phalantī ti yojetabbaṃ. Evaṃ tidhā
phalaṃ dadantañ c’etaṃ kāmâvacara-sugatiyaṃ upapattiṃ, sugati-duggatīsu bhoga-sam-
padañ ca karoti.
212 Cousins
too does not seem to be something which could have come from a Northern
Abhidharma tradition:
But they explain that there is [no] fruit of the citta which is accompanied
by agitation,
since it is said with reference to agitation that [discrimination of] the
meaning is [knowledge of] its result. (40) (cf. Vibh 297)
In its present form the stanza seems corrupt. The reference is the commen-
tarial discussions concerning the results of this eleventh kind of unskilful citta.
The discussion presupposes the statement in the Dhammasaṅgaṇi (Dhs §1391)
that classifies this citta arising (cittuppāda) as exclusively to be abandoned by
bringing into being (bhāvanā). In other words it can only be removed by the
development of the path after the stage of stream-entry. Most other citta aris-
ings are partially eliminated at this stage and partially eliminated in the
preceding stage of vision (dassana). So this is generally understood as meaning
that agitation consciousness is not able to give result as the connexion citta
which would initiate a future life (in one of the four apāyas).88
It is explicitly understood that it can give results during the course of such a
life. The basis for this is a statement in the Vibhaṅga. At Vibh 297, when clas-
sifying the four discriminations (paṭisambhidā), it is specifically indicated that
the discrimination of attha is knowledge of the result in respect of all twelve
unskilful cittas. It is this that is being referred to in the second line of this
Sanskrit stanza. We must therefore emend the first line to remove the
negation.
It is not specifically indicated in the stanza, but we may suspect that what
is being referred to is a debate in the ṭīkā literature, which refers to a modi-
fied view which allows for the existence in puthujjanas of a type of agitation
consciousness which can give reconnection in a future life.89 They seem to
have understood that it was already abandoned prior to the stage of stream-
entry. According to Sumaṅgala this was the position of Buddhamitta and
others. However, the name Buddhamitta does not seem to have been unusual
and no Pāli text authored by a Buddhamitta appears to be extant. Indeed it
is quite possible that what is being referred to is a position espoused by non-
Mahāvihāravāsin writers.
8 Conclusion
From the Cūlavaṃsa we learn that Jotipāla played a major role in the
Buddhism of Ceylon around the seventh century. From the citations in the
work of Sumaṅgala we know that he was considered a major authority in the
Polonnaruva period and we can deduce that he was the author of an Anuṭīkā
on the Abhidhammapiṭaka. From the citations of verses specifically attributed
to him it is clear that he also wrote in Sanskrit.
90 Ce1: skandhenaiva.
91 em. to samuccaya iti and to arthâyog- in the first and third occurrences.
92 Ce1: Jñeya-sampatti-ṭīkā.
93 Ce1: vastuna.
94 Ce1: prayuktā.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 215
Vism-mhṭ 243:
Jambu-dīpo sakaṭa-saṇṭhāno. Apara-goyānaṃ ādāsa-saṇṭhāno. Pubba-videho
addha-canda-saṇṭhāno. Uttara-kuru pīṭha-saṇṭhāno. Taṃ-taṃ-nivāsīnaṃ,
taṃ-taṃ-parivāra-dīpa-vāsīnañ ca manussānaṃ mukham pi taṃ-taṃ
saṇṭhānan ti vadanti.
The meaning is not made known through sounds which have not become
sense-objects.
Nor do those who have not apprehended meanings make [anything] known
from the mere [apprehension of] a being. (42)
95 Ce1: pānaḥ.
96 Ce1: ubhayenāpy akabalaṃ kṛtvāvahāryyate; cf. Abhidh-k-bh I 30; III 39.
97 cf. Vism-mhṭ I 243 = Ss 332 ≠ Sp-ṭ 273; also Lok-d 184.
98 (Ce1): a(rḍḍ?)endv-.
99 Sadd: pada-mattato.
216 Cousins
Abhidh-av-ṭ I 192:100
hoti c’ettha:
kāme ’vacaratîty etaṃ, kāmāvacara-saññitaṃ.
sese avacarantam pi, saṅgāmâvacaro yathā ti
The reason they are given as five is the result of dividing painful and pleasant
[feeling] in two as mental and bodily,
because they differ in the performance of their function, but that is not the case
with neutral feeling. (44)
Vism-mhṭ II 136:
Kiṃ pana kāraṇaṃ mānas’etara-sātâsāta-vasena sukhaṃ, dukkhañ ca vibhajitvā,
vuttaṃ: sukhaṃ somanassaṃ dukkhaṃ domanassan ti, upekkhā pana mānasī,
itarā ca ekadhā va vuttā ti? Bhedâbhāvato; yathā hi anuggahûpaghātakatāya
sukha-dukkhāni aññathā kāyassa anuggaham upaghātañ ca karonti, aññathā
manaso, na evam upekkhā. tasmā bhedâbhāvato ekadhā va upekkhā vuttā ti.
100 Dhs-a 62 ≠ Paṭis-a I 234; cf. I 296: Kiñcâpi hi etaṃ rūpârūpa-bhavesu pi avacarati, yathā
pana saṅgāme avacaraṇato saṅgāmâvacaro ti laddha-nāmako nāgo nagare caranto pi
‘saṅgāmâvacaro’ tv eva vuccati, thala-jala-carā ca pāṇā athale ajale ca ṭhitā pi ‘thala-carā
jala-carā’ tv eva vuccanti, evam idaṃ aññattha avacarantam pi kāmâvacaram evā ti
veditabbaṃ.
101 Or em. to ghātyānugrahayor ‘because they differ in harming and facilitating’.
Sanskrit Abhidharma Literature of the Mahāvihāravāsins 217
The chariot of birth mounted by foolish folk is this wheel of afflictions and
actions.
After destruction of just the one wheel of afflictions, it does not operate, there
being [only] the single wheel of actions. (47)
Abhidh-av-ṭ II 104:
Tenâhu Porāṇā:
bāla-jan’-ajjhāruḷho
jāti-ratho kamma-kilesa-cakko ’yaṃ
kilesa-cakka-vidhamano
na yāti kamm’-eka-cakke pī ti.
?The reality of a shared birth is seen by one investigating the reality or unreality
of something in a stream of continuity and its unreality [is seen] due to the
difference of the moment. (48)
Appendix 3: Unclear
Abbreviations
Abbreviations used in this chapter are those of the Critical Pāli Dictionary. Page refer-
ences to Vism-sn are to the edition of Saddhātissa, 1949–1955. Citations are to the page
numbers of the editions published by the Pāli Text Society. Page numbers for other
texts are as given for the Burmese edition on the Vipassanā Research Institute (VRI) CD,
unless otherwise indicated.
108 Sv-pṭ III (Be) 49 = III (Ee) 62 (vll): Micchā-diṭṭhi-kammassā ti “esa saddhādhigato deva-
yāno, yena yanti puttino visokā” ti-ādinā pavattitassa micchā-diṭṭhi-saha-gata-kammassa.
Mss to Ee variously have: panethāvitato; pan’ esā vitato; saddhâdhigato which Ee recon-
structs to: esa pan’ evâvitatho. Many read vipattino for puttino.
109 Vism-mhṭ (Ce): saddo.
110 Abhidh-av-ṭ (Ce): vitato; (Be): vihito for vitatho; Vism-mhṭ inserts: pag eva before vitatho.
111 Abhidh-av-ṭ (Ce): visokataṃ.
112 Abhidh-av-ṭ (Ce): passanti taṃ.
113 Abhidh-av-ṭ (Ce): te is omitted in the old Be.
220 Cousins
Bibliography
Chapter 6
1 Introduction
I was not surprised to notice that the French translation […] based mainly
on the Chinese version of Hiuan-tsang, adhered literally to the original
Indian text. The honor of such an exceptional success must be given both
to the Chinese master of the seventh century and the Belgian master of
the twentieth century […].1
When we collate Xuanzang’s translation of the AKB and the Ny, we can find
evidence, at least in some places, that Xuanzang’s understanding of the AKB
was guided by the Ny. Moreover, several of his pupils – Puguang 普光, Fabao
法寶, Shentai 神泰 and Huaisu 懷素 – had all received instructions from him
on the AKB and wrote important commentaries on the text (Huaisu’s is no lon-
ger extant), citing the Ny extensively and indispensably. Puguang, for instance,
quotes it by name more than 2,600 times, and generally takes Saṃghabhadra
as representing the correct Vaibhāṣika exponent. In the following, I will give
two of many similar examples that show the influence of the Ny on Xuanzang’s
pupils, and the central position which Saṃghabhadra often holds in the con-
text of Abhidharmic controversial debates.
(1) On the nature of the non-defiled nescience (akliṣṭājñāna), Puguang first
enumerates eleven Indian opinions.2 Information on opinions (1) to (4) must
have been derived from his teacher, Xuanzang. For opinions (5) to (11), he cites
Saṃghabhadra’s Ny. He then proceeds to refute them one by one, the first four
with his own reasoning basing himself on Abhidharma tenets.3 For the refuta-
tion of (5) to (11), he simply quotes Saṃghabhadra’s arguments, offering some
elaboration in each case.4 Finally, he cites, as the “correct meaning” (zheng yi
正義), Saṃghabhadra’s own interpretation that the akliṣṭājñāna has, as its
intrinsic nature, all the inferior prajñās (understanding) that are impure
(sāsrava) and non-defiled (akliṣṭa), which are present in one who has not yet
attained buddha-hood.5
Puguang’s enumeration of the first four opinions (the third is ascribed to a
“western master” (pāścātīya), named Deguang 德光, *Guṇaprabha) shows that
the views of other Abhidharma masters were known to the Xuanzang’s tradi-
tion – which is only to be expected since Xuanzang studied in India for many
years. Besides, Sthiramati’s Abhidharmkośaṭīkā Tattvārthā (= Tattvārthā) was
translated into Chinese by Xuanzang.6 In fact, both Puguang and Fabao occa-
sionally also quote its rejoinder to Saṃghabhadra’s objection to Vasubandhu’s
explanation, as can be seen from the next example.
(2) Vasubandhu defines the avijñapti thus:
6 Only a small portion of this translation is preserved. For the past many years, Professor Odani
Nobuchiyo, et al., have been working on an edition of the Sanskrit manuscript recently ob-
tained from China.
7 T.29.1562: 335b27–29: “ 又謂等言通無心者, 此言無用; 前已攝故: 亂心等言, 已攝一切
餘有心位; 第二等言, 復何所攝? …” The criticism covers more than what I have discussed
here.
8 T.41.1821: 22a15–28.
9 T.41.1821: 22b2–22.
228 Dhammajoti
2 Yogācāra Commentaries
In their Yogācāra commentaries too, the Xuanzang tradition as well as
Sthiramati quote Saṃghabhadra as the authoritative representative of the
Vaibhāṣikas. (See also below). Thus, Kuiji 窺基, considered as Xuanzang’s
chief pupil, in his commentary on Xuanzang’s Cheng weishi lun 成唯識論
(CWSL Shuji 述記), describes Sthiramati as the master who “compiled the
*Abhidharmasamuccayavyākhyā, rescued/defended the AKB and refuted the
*Nyāyānusāra master (jiu jushe po zhengli shi 救俱舍破正理師)”.14 Even in this
description, the indication is evident that the Yogācāra master, Sthiramati, is
10 T.29.1562: 348c12–349b7.
11 Tattvārthā(C), 222 f. See also Dhammajoti, 2012: 244f.
12 Dhammajoti, 2012: 216f.
13 Tattvārthā(C), 239.
14 T.43.1830: 321c19–23: “ 梵云悉恥羅末底。唐言安慧。即糅雜集。救俱舍論破正理
師 …”.
The Contribution of Saṃghabhadra 229
Vasubandhu speaks the truth when he says that his book is an explana-
tion of the doctrines of the Vaibhāṣikas of Kaśmīr […]. But the Kośa, as
we know, is also the refutation of several essential doctrines of the
Vibhāṣā. Vasubandhu, who the orthodox Vaibhāṣika Saṃghabhadra calls
“the Sautrāntika”, sets the view of the Sautrāntikas or his own views
against the views of the Sarvāstivādin-Vaibhāṣikas: here we have very lit-
tle information on the sources that he uses; he makes use of data foreign
to and undoubtedly later than the Vibhāṣā, the Dārṣṭāntikas–Bhadanta–
Dharmatrāta cited by the Vibhāṣā. We will see more clearly when we have
read Saṃghabhadra who gathers information on Śrīlāta, who must be
one of the Sautrāntika sources of Vasubandhu.23
Indeed, among the extant Abhidharma texts, it is in the Ny that we find detailed
doctrinal descriptions and arguments of the Dārṣṭāntika-Sautrāntikas, along
with Saṃghabhadra’s lengthy refutations. Of particular importance are
Sthavira Śrīlāta’s doctrines, such as: the anudhātu (or *pūrvānudhātu) theory,
the cognitive process, the citta-caitta doctrine, the doctrine that the sense-ele-
ments (dhātu) alone are real existents (and not the aggregates (skandha) or the
sense-spheres (āyatana); the process of direct perception (pratyakṣa); etc.24 It
is also in this work that we are introduced to Śrīlāta’s explanation of the eight-
moment process of direct realization (abhisamaya).25
The Ny also clarifies the relationship between the Dārṣṭāntikas and the
Sautrāntikas. Basing ourselves on the MVŚ, the AKB and the Ny, we can see that
the early Dārṣṭāntikas figuring in the MVŚ were Sarvāstivādins: They shared the
school’s central thesis of the tri-temporal existence of dharmas. At least around
the time of the AKB, we see one sector of them evolving into Sautrāntikas, com-
ing to hold the present-only-exists standpoint. Śrīlāta, usually described – as in
Ny – as a Dārṣṭāntika master, was also sometimes called a Sautrāntika.
Xuanzang informs us that he composed the *Sautrāntikavibhāṣā (Jing bu pipo-
sha 經部毘婆沙), which Xuanzang himself studied in India. From this, we can
23 Lodrö Sangpo, 2012: 78. Had Louis de La Vallée Poussin lived several years longer, we
would have had the good fortune of his study of the Ny. Although he quotes Saṃghabhadra
frequently in his annotations, it appears that he probably had not made any detailed
study of the text. We find him – at an earlier stage, for instance – at a couple of places
attributing the anudhātu doctrine of the “Sthavira”, in the Ny, to the “the school of the
Sthaviras”! (See his French translation of the AKB: de La Vallée Poussin, 1971: 173, note 5,
and 246, note 2). These were later amended by him in his “Additions and Corrections”.
24 We have been able to make some study of these doctrines to some extent. See Dhamma-
joti, 2007a; 2007b; 2011b; 2013.
25 Cf. T.29.1562: 684a-b.
232 Dhammajoti
understand that he was a Sautrāntika leader at the time, for which reason
Saṃghabhadra refutes his doctrines extensively when refuting Vasubandhu. In
fact, as is well known, Yaśomitra states that “sautrāntika” means the same as
“dārṣṭāntika”.26
Thus, in this later period, the two labels, “Dārṣṭāntika” and “Sautrāntika” had
come to be interchangeable, and no pejorative implication (contrary to Jean
Przyluski’s theory) is detectable in the first label.
We can also observe the same in the “what sees” debate. Vasubandhu attri-
butes the criticism that the whole debate is in vain to the Sautrāntika (AKB, 31:
sautrāntikā āhuḥ …); but both the Abhidharmadīpa-Vibhāṣāprabhāvṛtti of the
Abhdharmadīpa and the Ny attribute it to the Dārṣṭāntikas.27
The Ny, which predates the Vy, provides further instances showing the
interchangeability between the two labels. For example, Saṃghabhadra
states that Śrīlāta and his pupils are those who assert that all sūtra is explicit
meaning (nītārtha).28 This is none other than the fundamental stance of the
“Sautrāntikas”.
On the issue of nītārtha and neyārtha sūtras (sūtras of implicit meaning),
Saṃghabhadra states more explicitly that Śrīlāta – and along with him
Vasubandhu as well – is not worthy of the Sautrāntika claim that they take the
sūtras as the final authority; thus implying that Śrīlāta was identified as a
Sautrāntika during Saṃghabhadra’s time:
[…] the Sthavira himself, from various perspectives, interprets the pro-
found implicit meanings of that sūtra. […] Which wise person is so
attached [as he], as to call a sūtra having an implicit meaning a nītārtha
sūtra? […] It is extremely illogical [for him] to claim: “we take the sūtras
as the decisive authority” (我用經為定量; *vayaṃ sūtraprāmāṇikāḥ29) –
while not discerning the difference between the nītārtha and neyārtha
sūtras. […] In this connection, the Sūtrakāra (i.e., Vasubandhu) asserts
thus: “…” This [assertion] should be fully refuted as in the case of the
Sthavira’s tenet.30
It is not the case that, in the Sautrāntika tenets, the paryavasthānas can
be abandoned (fei jing bu zong (chan) you duan yi 非經部宗 (纏) 有斷義),
because, [according to them,] they are conjoined with thought
(cittasaṃprayukta) and the past and future are non-existent […].33
31 Śrīlāta cites a sūtra (Cf. Mahāmāluṅkyasutta, Majjhima I, 433) to substantiate his claim
that “the Buddha, Bhagavat, himself explains the difference between the paryavasthānas
and the anuśayas: When a kleśa comes into manifestation, it is called a paryavasthāna,
because it presently binds the person’s serial continuity; the anudhātu of the kleśa is
called [its] anuśaya, for [its] causal efficacy always follows along and lays dormant”.
(T.29.1562: 597b27-c3: “ 上座於此, 謂佛世尊, 自說諸纏與隨眠異. 謂: 諸煩惱現起,
名纏; 以能現前縛相續故. 煩惱隨界, 說名隨眠; 因性恒隨而眠伏故. 以契經說:
幼稚童子嬰孩眠病, 雖無染欲, 而有欲貪隨眠隨增. 此唯說有諸隨眠性 …”).
32 T.29.1562: 598a1–5.
33 T.29.1562: 598a7–8.
34 T.29.1562: 398b12–17.
234 Dhammajoti
Significantly, this causal category came to be held by the Yogācāra school as the
only true type of causality – quite in spite of the dominant doctrinal influence
on this school by the Sautrāntikas who were (as evident in all the extant
Abhidharma texts) vehemently opposed to it – and, indispensably, was uti-
lized in their establishment of the vijñaptimātra thesis and of the doctrines of
bīja (seed), vāsanā (habitual residue), and the simultaneously occurring pro-
cess of bījas in the ālayavijñāna manifesting into phenomena and the
perfuming of these phenomena on the bījas.35
36 T.27.1545: 396b18–22.
37 T.29.1562: 631, b9–10.
236 Dhammajoti
This leads to our discussion of what are commonly called the “neo-Sarvāstivāda”
doctrines of Saṃghabhadra. Such a conception is very much influenced by
Vasubandhu’s exposition of the Sarvāstivāda-Vaibhāṣika doctrines, and often
carries the implication that these supposedly new or modified Sarvāstivāda
doctrines represent innovative views not held by the tradition prior to the AKB
– and generally attributed to Saṃghabhadra.
One example is the claim that under the criticism in the AKB, the Vaibhāṣikas
came to assert the simultaneous operation of examination (vitarka) and scru-
tiny (vicāra) (also so explained in the Ny). But, in actual fact, such an
explanation has long been presented in the MVŚ.45 Other similarly claimed
views include the above-mentioned definition of activity (kāritra);46 the defi-
nition of the cessation not due to deliberation (apratisaṃkhyānirodha);47 the
simultaneous operation of the three primary characteristics (saṃskṛtalakṣaṇa)
– enduring (sthiti), deterioration (jarā), passing away (vyaya); the contrast
between hehe 和合 and heji 和集; etc. Most of these doctrines have been dis-
cussed elsewhere. As to the simultaneous operation of the three lakṣaṇas:
Kuiji and others describe it as a neo-Sarvāstivāda doctrine of Saṃghabhadra.48
But it is in fact already found in the MVŚ.49
To save space, I shall here only discuss the last example, i.e. the contrast
between hehe and heji, which will also serve to further illustrate how
The gigantic MVŚ – whose compilation must have spanned over decades
and finally completed as early as circa the middle of the second century CE
– records Abhidharma doctrines which are extremely valuable for the study
of the emergence of the Mahāyāna. These include: the doctrine of the three
vehicles (yānatraya: śrāvaka-, pratyeka- and buddhayāna); of the Buddha
alone possessing perfect wisdom (sarvākārajñāna) and great compassion
(mahākaruṇā) whose intrinsic nature is understanding (prajñā); of the
Buddha’s fulfillment of the six pāramitās, of him alone being totally free from
contamination (kleśa) along with all habitual residues (vāsānā); of (contam-
ination-hindrance) kleśāvaraṇa and knowable-hindrance (jñeyāvaraṇa); of
the Buddha “journeying in the five planes of existence together with Ānanda,
incessantly day and night benefitting sentient beings”;54 etc.55
This text also provides considerable information on the early Dārṣṭāntikas
and the Sarvāstivādin Yogācāra masters, who are important for our under-
standing of the emerging Mahāyāna movement.
The Ny contains further elaboration, additions and articulation of many of
these doctrines, and in this way adds to the value of the Abhidharma textual
tradition as an indispensable source for the study of the doctrinal evolution of
the Mahāyāna.
An important instance is Saṃghabhadra’s elaborate discussion on the non-
defiled nescience (akliṣṭājñāna) and habitual residue (vāsanā) (see 2.1 above).
Such buddhological notions have influenced the whole of the Mahāyāna
development, particularly the Yogācāra.56 They also partly contributed to some
later Mahāyāna conceptions, such as that of the avidyāvāsabhūmi and the two
types of death – parichinnā cyuti and pariṇāmikī cyuti. This two-death doctrine
originated in the tathāgagarbha school of thought; but also came to be taken
in by the Yogācāra57 and the *Mahāprajñāpāramitopadeśa (Da zhidu lun 大智
度論)58 in their own ways.
We have already noted above the doctrinal importance of the doctrine of
co-occurring causality for the Yogācāras. Saṃghabhadra’s systematic and very
detailed exposition and brilliant defence of this doctrine in the Ny can be stud-
ied alongside with the Yogācāra discussion predating and post-dating the Ny
– for example, Asaṅga’s *Mahāyānasaṃgraha (and commentaries thereon by
Vasubandhu and Asvabhāva) and the CWSL, both of which claim this causal
category as the only valid basis for the establishment of causality.59 Such a
comparative study would yield useful information on the doctrinal interaction
between the Abhidharma tradition and the Yogācāra in this milieu. The same
remark applies to discussions and arguments in the Ny on the doctrines of the
Sautrāntika, the Vibhajyavāda and the Mahāsāmghika (such as the intrinsic
8 Conclusion
Even though space is limited here, I hope I have been able, from various per-
spectives, to further clarify the importance and need for studying the Ny
beyond the AKB and the Vy. But daring to be more concrete and more specific:
Within our modern world where English functions as the lingua franca, it
would be a great step forward in the advancement of Abhidharma studies, if a
team of scholars could jointly devote some effort to the systematic translation
of the Ny into the English language.
Abbreviations
Bibliography
Chapter 7
1 Introduction
1 On pratītyasamutpāda, the classic study of Louis de La Vallée Poussin, 1913 remains one of the
best. See also Johansson, 1979; Nakamura, 1980; Stalker, 1987: 47–96; Cox, 1993; Bucknell, 1999;
Jurewicz, 2000; Shulman, 2008.
2 Yo paṭiccasamuppādaṃ passati so dhammaṃ passati, yo dhammaṃ passati so
paṭiccasamuppādaṃ passatīti (MN, 1:190–191).
3 Little is known about An Shigao himself (see Forte, 1995 for some interesting theories) or his
translation methods. He may have simply been the reciter of the Indian texts, as was the
normal role of the non-Chinese “translators” in later times (Zürcher, 1972: 31). For the sake of
convenience, I shall refer to “An Shigao” as an individual person, but we should remember
that the works associated with this name may have had a collective authorship.
that were, in this same period of time, leading to the formation of the various
Abhidharma collections.
An Shigao’s translations are equally important – if so far underutilized by
scholars – from the perspective of Chinese Buddhism, as they allow us, in
some cases at least, to glimpse how key Buddhist ideas were initially under-
stood in China. Such questions can be asked, of course, about a broad range
of topics. However focusing our attention on a fixed list of terms such as the
pratītyasamutpāda formula allows us to be certain of the underlying Indic
vocabulary that An Shigao was translating, something difficult to achieve in
other cases.4 With this as a point of reference, we can then begin to observe
notable features of the Chinese terms An Shigao selected in his transla-
tions, as well as of the interpretations of these terms by subsequent Chinese
commentators.
As I will show, An Shigao had particular difficultly finding a suitable Chinese
translation for the term sparśa, “sense contact,” a concept that did not have a
ready equivalent within pre-Buddhist Chinese theories of sense perception.
The subsequent efforts of commentators to grasp this topic, which occupies a
central place within pratītyasamutpāda, shows that sense perception was an
area where pre-existing Chinese theories and technical vocabulary strongly
shaped the Chinese interpretation of Buddhist ideas.
2 The Terms
It will be convenient to first present a table with all known examples of the
members of the pratītyasamutpāda formula within the surviving translations
of An Shigao.5 In compiling this information, I have limited myself to the con-
4 Indic parallels have been identified for some of An Shigao’s translations (Harrison, 1997;
Harrison, 2002; Zacchetti, 2002). When possible I have used these to help establish An Shigao’s
usage (for a complete survey of known parallels to An Shigao’s translations, see Nattier, 2008:
35–72). But even without such parallels, it is usually possible to identify the pratītyasamutpāda
formula, safely allowing us to align the translated terms with the items in this list.
5 For the purposes of this chapter, “An Shigao’s translations” are those given by Nattier, 2008
and Zacchetti, 2010a. I cannot here address the important question of the reliability of the
received texts of An Shigao’s translations. Though there is no reason to doubt them on the
whole, there is some evidence that, within early Chinese Buddhist texts in general, later copy-
ists or editors occasionally replaced unusual technical terms with more familiar ones. Thus
in the Xiuxing benqi jing 修行本起經 (possibly a third-century text; Nattier, 2008: 104–109),
the second Kǒryo edition (the main Taishō text) gives shou 受 as the translation of upādāna,
which is standard in early translations (T.3.184: 470c2). However the Sixi and Puning editions
250 Greene
a Ren ben yu sheng jing 人本欲生經, T.1.14: 243b6 and passim. There is also a single case within this
text of the eventually standard translation ming se 名色 (T.1.14: 243b19); this may be a later
hyper-correction.
b Chang ahan shi bao fa jing 長阿含十報法經, T.1.13: 240a24–25 (reading 名字色 for 名字苦).
c T.1.13: 240a25; YCRJ, T.15.603: 174b25 (and throughout).
d T.2.150A: 876b20.
e YCRJ, T.15.603: 174c13–14; here followed by the more usual liu ru 六入. Note the similar ru shou 入
受 in the Dao di jing 道地經 (T.15.607: 231c5). On the prevalence of reversible compounds in Han-
dynasty Buddhist translations, see Zürcher, 1979: 180.
f T.1.14: 243a27-b3 (and throughout).
g On this meaning of geng 更, see below note 26.
h T.1.13: 240a25 (de Jong, 1979: 272).
i YCRJ, T.15.603: 174b25, b30.
j T.1.57: 853a3.
k T.1.57: 852a25; YCRJ, T.15.603: 174c14–17.
l T.2.150A: 875c1 and passim (see Vetter and Harrison, 1998: 205).
m YCRJ, T.15.603: 174c16 (and passim); T.1.13: 240a25; T.1.14: 243b4 and passim; T.1.57: 852a20–22 and
passim.
n YCRJ, T.15.603: 174b25.
o This is An Shigao’s standard translation (e.g. YCRJ, T.15.603: 175a13–14).
p T.1.14: 242a14–15 and passim.
Pratītyasamutpāda in the Translations of An Shigao 251
(Song 宋 and Yuan 元 in the Taishō footnotes) give qu 取, which later replaced shou 受 as
the normal translation of upādāna. At some point an editor thus “updated” some copies of
this text. Although in this case the (presumably) correct reading survives in the second Kǒryo
edition, this raises the possibility that in some texts archaic terminology may have been re-
placed by later copyists and editors. This possibility raises important questions, especially
since scholars have often relied upon the translations of technical terms as markers of the
authenticity and likely date of early Chinese Buddhist translations.
6 In particular I include the Lou fen bu jing 漏分布經 (T.1.57) and the Qi chu san guan jing 七
處三觀經 (T.2.150A), both of which discuss “sensation” (vedanā) as originating in “contact”
(sparśa), thereby clearly invoking their relationship in pratītyasamutpāda.
7 The texts An Shigao translated were certainly not written in Sanskrit. For the sake of conve-
nience, however, I will use Sanskrit forms throughout, unless directly citing a Pāli or other
middle Indic text.
8 The translations used are not, however, always the most common ones. Thus the AHKJ uses
“seeking” (qiu 求) for upādāna (T.25.1508: 53a10), found in only a single other An Shigao trans-
lation (9b in the table).
9 On these terms from the AHKJ, see Zacchetti, 2004: 198, note 4.
252 Greene
行);10 for vijñāna not simply “consciousness” (shi 識) but “what one becomes
conscious of” (suo shi 所識);11 and similarly for sparśa “what is suffered” (suo
geng 所更) rather than simply “to suffer” (geng 更).12 We also find the seem-
ingly new “name and appearance” (ming se 名色) for nāmarūpa,13 though this
differs only slightly from the mingzi se 名字色 that An Shigao uses elsewhere
(4b in the table above).
A few of the new translations in the AHKJ are worthy of note, such as “appro-
priating” (de 得) for bhava.14 Several novel translations of saṃskāra – “seminal
factors of rebirth” (sheng si jing 生死精),15 and “seminal activities of rebirth”
(sheng si jing xing 生死精行)16 – are perhaps combinations of An Shigao’s
standard translation of saṃskāra in the case of pratītyasamutpāda, namely
xing 行 (“activities”), with the different translation he usually uses to render
saṃskāra when it is one of the five skandhas, namely shengsi 生死 (“[factors]
leading to rebirth”).17
The above table already allows us to draw some interesting conclusions
about the issues that An Shigao faced when translating the pratītyasamutpāda
formula. We may first notice the relative uniformity of the translations. Seven
of the twelve terms are always translated in exactly the same manner (one of
these, bhava, finds a variant only in the AHKJ). Three other terms occur with
only slight variations: nāmarūpa translated as both “name and style” (ming zi
名字) and “name and appearance” (ming zi se 名字色),18 and vedanā, usually
simply “pain” (tong 痛) but on one occasion “pain and pleasure” (tong yang 痛
痒),19 which is An Shigao’s normal translation for vedanā in the context of
10 T.25.1508: 53a8. Cf. xing zuo 行作 in Taizi ruiying benqi jing 太子瑞應本起經, T.3.185:
478b7.
11 T.25.1508: 53a9.
12 T.25.1508: 53a10. The use of suo 所 here is peculiar (as noted by Zacchetti, 2007: 412–413).
13 T.25.1508: 53a10.
14 T.25.1508: 53a12. This translation of bhava does not seem to appear outside this text.
15 T.25.1508: 53b20.
16 T.25.1508: 53b7.
17 An Shigao uses sheng si 生死 to translate saṃskāra only in the context of the skandhas,
never in pratītyasamutpāda, where we always find xing 行. However conversely, xing 行
does occasionally appear in the context of the skandhas (T.1.13: 234c24; T.15.607: 231b29-
c1).
18 The translation of nāmarūpa as ming zi 名字 is mysterious. As Vetter suggests, we are
perhaps meant to read ming 名 and zi 字 as opposed categories, “true name” versus “pub-
lic appearance” (Vetter, 2012: 55–56). Alternatively, perhaps An Shigao’s lists had only
nāma rather than nāmarūpa (as we see in, e.g., Vibh: 138–139).
19 Several second- or third-century texts associated with An Shigao explain tong yang 痛痒
(var. 痛癢) as “pain” and “pleasure” respectively (Da anban shouyi jing 大安般守意經,
Pratītyasamutpāda in the Translations of An Shigao 253
either the five skandhas or the second smṛtyupasthāna.20 For the term upādāna,
we do find two different translations (“grasping” shou 受 and “seeking” qiu 求).
But even here, the meaning of the two words is rather similar.
Things become more interesting for link 5, the “six sense gates” (ṣaḍāyatana)
and link 6, “contact” (sparśa), two terms that are crucial for the Buddhist expla-
nation of sense perception. For āyatana, An Shigao used two quite distinct
Chinese expressions: the “six entrances” (liu ru 六入) and the “six destroyers”
(liu shuai 六衰). The first of these translations fits a common Indian Buddhist
explanation of the āyatanas as the gateways through which sense data arrive,21
an explanation that is expressed pictorially when the āyatanas are depicted as
the six doors or windows of a house in paintings of the so-called wheel of life
(bhavacakra).22
But the second of the two Chinese translations – the āyatana as the “six
destroyers” – is more mysterious. Certainly it is not a literal translation. Here
too, however, An Shigao may have been drawing on Indian commentarial tra-
ditions. Indeed, according to one well-known explanation, the six external
āyatanas (the six kinds of sense data) are like “village-raiding bandits,” “attack-
ers” of the sense organs that “destroy” them.23 The term liu shuai (“the six
T.15.602: 168a7–11; YCRJZ, T.33.1694: 11c2; SMJ, 297–298). This does seem like a good expla-
nation, as this would yield a typical Chinese abstract noun meaning “sensation”. However
some classical Chinese Buddhist lexicographers consistently explained yang 痒 to be syn-
onymous with tong 痛 (Yiqie jing yinyi 一切經音義, T.54.2128: 529a5). Modern dictionar-
ies too explain yang 痒 as only possibly meaning “sickness” or “unpleasant itching”
(GHYZD: 754; HYDZD: 2673; Morohashi: 8080). Furthermore this meaning is not necessar-
ily late, and we find examples from Han or Three Kingdoms period Buddhist texts where
tong yang 痛痒 must simply mean pain (Zhong benqi jing 中本起經, T.4.196: 152a8–10;
Bannihuan jing 般泥洹經, T.1.6: 186c6). It is not clear how we are to resolve this discrep-
ancy.
20 As he seems to have done with saṃskāra (see above note 17), An Shigao may have inten-
tionally distinguished vedanā in the context of pratītyasamutpāda from vedanā as a skan-
dha or a smṛtyupasthāna. With only one exception he uses tong 痛 in the former context
and tong yang 痛痒 in the later. Interestingly in that exceptional case (the YCRJ), a com-
parable distinction is still maintained, albeit reversed, with tong 痛 used for vedanā as a
skandha (T.15.603: 173b7) or a smṛtyupasthāna (T.15.603: 174a4), and tong yang 痛痒 in
pratītyasamutpāda (T.15.603: 174b25). It is unclear if there was here any intended differ-
ence in meaning. Given that the YCRJ preserves the distinction while reversing the terms,
the purpose may have been simply to mark the different contexts.
21 AKBh, 13.18
22 Hartmann, 1940: 359.
23 “The external [āyatanas] are to be likened to village-raiding bandits, because they are of
the nature to attack the internal [āyatanas].” Bāhirāni gāmaghātakacorā viya
254 Greene
how to best render this term into Chinese. On the one hand, this might indi-
cate that An Shigao attributed particular importance to this term – that for
him sparśa was the key concept within pratītyasamutpāda. But it is also pos-
sible that the notion of sparśa proved especially difficult for the Chinese to
grasp, and that the diversity of translations reflects a continuing struggle
to convey the meaning of this term. Indeed none of An Shigao’s choices for
sparśa ultimately proved satisfactory, and most were never used again by later
translators.25 And this, again, stands in sharp contrast with many of his other
translations from the pratītyasamutpāda formula – such as ai 愛 for tṛṣṇā and
shi 識 for vijñāna – that remained standard for the rest of Chinese Buddhist
history.
Looking at the specific Chinese terms in question, we can often make at least
some sense of them. The initially perplexing translation geng 更, for example,
likely derives from this word’s semi-colloquial meaning (during the period
when An Shigao worked) of “encounter something bad” (hence my translation
“to suffer”) – such as being smitten by disease, curses, or other misfortune.26
Meanwhile the puzzling translation “to plant” or “a sprout” (zai 栽), was per-
haps chosen based on an explanation of pratītyasamutpāda that An Shigao
himself presents in the AHKJ. As Stefano Zacchetti has observed, the AHKJ
draws an analogy between the twelve links of pratītyasamutpāda and an
“external” sequence of twelve causal factors in the growth of a seed.27 Thus
ignorance (avidyā), the first link, is correlated with the “external” condition
earth (soil), conditioned factors (saṃskāra) with water, and so forth.28 This
continues for all twelve links, finishing with “old age and death” as the fruit (shi
實). Notably within this sequence the sixth “external” factor is the “seed,” zhong
種, and this perhaps explains An Shigao’s use of “sprout” or “to plant” (zai 栽)
to translate the corresponding “internal” factor, sparśa.29
25 “To suffer” (更) was reasonably influential, but only until the late fourth century. Zai 栽
was occasionally used into the very late third century (e.g., Fangguang bore jing 放光般
若經, T.8.221: 12a22–23), but not beyond that.
26 Hu, 2002: 143–144. The basic meaning of geng 更 is “change” (the Shuo wen 說文 glosses
it as gai 改; see GHYZD: 447). The meaning is thus similar to bian 變, which also means
both change and, by extension, misfortune. That geng 更 was here being used in this
sense, and, not in a more neutral sense of “experience” or “undergo” (cf. Vetter, 2012: 146),
is suggested by the slightly later, and only occasionally used translation of sparśa as geng-
le 更樂 (e.g. Weimojie jing 維摩詰經, T.14.474: 536a16), seemingly an attempt to create an
abstract noun (“contact”) by joining two opposites (“suffering and pleasure”).
27 Zacchetti, 2004: 207–212.
28 T.25.1508: 53a18–24.
29 Zacchetti, 2004: 208, note 59. Cf. Yiqie jing yinyi 一切經音義, which correctly interprets
zai 栽 as meaning sparśa (chu 觸), but which explains it by analogizing sparśa, which
256 Greene
Here too, however, knowing the ultimate source of the explanation of sparśa
as a seed or sprout does not explain why An Shigao chose to base his Chinese
translation of this word (or rather, one of his translations) on this understand-
ing, as he certainly did not follow a similar strategy with his translations of the
other pratītyasamutpāda terms. As with An Shigao’s translation of āyatana as
“destroyers,” we see here evidence that the concept of sparśa was somehow felt
to require a different kind of explanation.
“grows into the mind and mental factors,” with a sprout, which grows into a tree (T.54.2128:
356c19).
30 I here use the title “canonical” to distinguish the transmitted text bearing this name from
that discovered at Kongō-ji 金剛寺, which is likely An Shigao’s original translation
(Deleanu, 2003; Zacchetti, 2008).
31 T.15.602: 169c29. The other senses are described similarly. See T.15.602: 167c11–12 (ear);
169c27 (nose).
Pratītyasamutpāda in the Translations of An Shigao 257
32 T.33.1694: 11a2–3. As noted by Zacchetti, 2010b: 144–147, the base text and inter-linear
commentary were transmitted together as a single unit, and the distinct title Yin chi ru
jing zhu 陰持入經註 is a creation of the editors of the (modern) Taishō canon. For ease
of reference, however, I will use this title to refer to the commentary.
33 That the division between “old” and “new” karma is the division between sense experi-
ence and one’s response to sense experience (or if not one’s response, then some other
mental factor that is distinguishable from sense experience itself) is already seen clearly
in the early scriptures (see for example SN IV:172).
34 The typical approach of the Pāli commentarial tradition marks a distinction between the
fruits (phala) of past karma (consciousness, name-and-form, the six sense gates, contact,
and sensation) and present karmic causes (hetu), namely craving, grasping, and becom-
ing (Vism: 579–580). The Vaibhāṣika system (AKBh, 134.6–135.6) similarly distinguishes
between consciousness, name-and-form, the six sense gates, contact, and sensation as
“bases” (vastu; explained as the results of past karma), and craving and grasping as “defile-
ments” (kleśa; see also Van Den Broeck, 1977: 131–132). The Abhidharmasamuccaya gives
yet another classification, distinguishing the “projected members” (ākṣiptāṅga) that are
the results of past actions (name-and-form, the sense gates, contact, and sensation) from
the “actualizing members” (abhinirvartakāṅga) that are causes for the future, namely
craving, grasping, and becoming (ADhS, 26.6–9; Kritzer, 1999: 27). For references to other
similar divisions, see de La Vallée Poussin, 1913: 35–36.
258 Greene
training, we are told in such texts, lies not in eradicating one’s present sensory
experience – which is the unavoidable fruit (vipāka) of past karma – but in
preventing the karmically productive mental defilements that arise in response
to whatever sensations (vedanā) one experiences.35 This formulation amounts
to a technical explanation of the basic Buddhist notion that it is possible to be
fully liberated while still alive (which was not something that all Indian ascetic
groups believed); that it is possible, in the language of pratītyasamutpāda, to
experience “contact” (sparśa) and “sensation” (vedanā) without then giving
rise to craving or attachment.
That we can separate sense perception into two such aspects is not neces-
sarily a phenomenological given. One of the primary functions of the central
section of the pratītyasamutpāda formula is thus to provide the vocabulary
that makes such a separation analytically possible. And from this point of view,
we may say that An Shigao’s use of overlapping Chinese terms for his transla-
tions would not have contributed to cementing a clear distinction between
sparśa as the fruit of past karma in contrast to upādāna as newly created
karma.
It is, further, interesting to observe that there are a number of other exam-
ples where An Shigao’s translations blur such distinctions in a potentially
similar way. Consider, for example, the seemingly strange translation of sparśa
as zai 栽, “to plant” or “a sprout”. As discussed above, this translation might have
been motivated by the analogy between “internal” pratītyasamutpāda and an
“external” sequence of the growth of a plant discussed in the AHKJ, in which
sparśa lines up with the “seed” (zhong 種)36 of the external sequence. However
within the AHKJ we find another passage where “planting” and “sprouts” are
associated not with sparśa, but with saṃskāra (xing 行), the second link of
pratītyasamutpāda.
This passage begins with reference to the previously given analogy of the
growth of a seed, and then comments on the first several links of the chain:
“The earth [in the previous analogy] corresponds to confusion (avidyā); activi-
ties (saṃskāra) refers to the flourishing of the dark elements (*upādānaskandha)”
(di ming chi xing wei sheng yin 地名癡行為盛陰). What is going on here is then
35 We thus find, in the early scriptures, innumerable statements to the effect that “the eye is
not the fetter of visible forms, nor are visible forms the fetter of the eye, but rather the
desire and lust that arise there in dependence on both: that is the fetter there”. (SN IV:163;
translation by Bodhi, 2000: 1230 with slight modifications). Conversely the Buddha is
reported to have criticized those who taught practices that merely involved ceasing sen-
sory perception itself (MN III:298–302). See also Bronkhorst, 1986.
36 Or perhaps, the “planting”; both 種 and 栽 functioning as either nouns or verbs.
Pratītyasamutpāda in the Translations of An Shigao 259
further explained: “[This means] one carries out ‘activities of birth and death’
(saṃskāra); to plant the sprouts [of rebirth] is called ‘activity.’” (geng zuo sheng
si xing. zi zhong zai ming wei xing 便作生死行。自種栽名為行).37
The activity of the saṃskāras is thus said to be like “planting sprouts” (zhong
zai 種栽), and this indeed fits with typical explanations of saṃskāra (in this
context) as the past karmic activities that have given rise to present existence.
However in the immediately preceding passage, the AHKJ also links “planting
sprouts” to a different section of pratītyasamutpāda:
十二因緣有三38 事。一者癡。二者生死精行。是前世因緣。三者識。
從識受身生。四者名色。色身復成五陰。是今世因緣。五者六衰。復
作生死精行種栽。是後世因緣。39
37 T.25.1508: 53b16.
38 Following the reading 三 given in the Sixi (宋), Puning (元), and Kunaichō (宫) editions.
39 T.25.1508: 53b6–10.
40 See above note 34.
260 Greene
a cause of future rebirth, as something that “gives rise to the seminal activities
of birth and death and the planting of the sprouts [of rebirth]”.41
Reading this passage in keeping with a more typical understanding of
pratītyasamutpāda, we might take “the seminal activities of birth and death
and the planting of the sprouts [of rebirth]” as an explanation of how the
karmic activities that condition one’s future life (craving, grasping, becom-
ing) arise in the wake of the entire process of sense activity (āyatana, sparśa,
vedanā). To refer to presently performed karma using the same language that
earlier characterized the saṃskāras (the “sprouts”) would make perfect sense,
since the saṃskāras are usually explained as, precisely, the karmic activities of
a past lifetime.
However from the Chinese text of the AHKJ alone, such an understanding
could probably only be reached by someone with a certain amount of back-
ground knowledge. Indeed not only is the word “sprouts” (zai 栽) explicitly
given earlier in the text as a translation of sparśa, but in the above passage it is
explicitly said that “planting the sprouts” is produced from the six sense gates,
precisely where we would expect sparśa to occur. In other words, as a Chinese
text the AHKJ seems here to assert that saṃskāra and sparśa are similar if not
identical things.42
41 A very similar discussion immediately follows this passage: “Once there is name-and-
form, there then comes to exist the six sense gates. There then again arises the flourishing
of the dark elements. After activities plant the sprouts [of rebirth], one will once again
undergo rebirth based on the unfurling of causes and conditions”. 已有名色,便有六
入。復作盛陰。行種栽後,當復受轉相因緣生死。(T.25.1508: 53b17–19).
42 It is interesting to note that a very similar ambiguity occurs in the Liao ben sheng si jing 了
本生死經, an early Chinese translation of the Śālistambasūtra, a text that, as Zacchetti,
2004, has noted, may have been important in the formation of the AHKJ (the translator of
the Liao ben sheng si jing is unknown, but the text is cited in the YCRJZ, so it must pre-date
the mid-third century). In a passage listing the entire pratītyasamutpāda formula,
saṃskāra is initially translated as xing 行. Links 5 through 9 are then described as follows:
“In dependence on name-and-form, there are the six sense gates. Because of a union of
the three [organ, object, and consciousness] there is contact. Because of the activity of
contact, there is sensation. Delighting in this sensation, there is craving. When craving
comes to fullness, there is grasping. When there is grasping there is again activity, and
therefore there is becoming”. 猗名色根故為六入。三合故為更樂。更樂 行 故為
痛。痛而樂故為愛。愛彌廣故為受。受當復有 行 ,故為有。 (T.16.708: 816a11–
17). Thus xing 行, which initially translates saṃskāra (as is usual), appears again in the
discussions of sparśa (geng le xing 更樂 行) and upādāna (shou dang fu you xing 受當復
有 行). In the case of upādāna, the point seems clear – grasping is precisely the occur-
rence of new karmic formations (saṃskāra). For sparśa it is much less easy to explain
things. Whatever the motivations of the translator, a Chinese reader here could easily see
Pratītyasamutpāda in the Translations of An Shigao 261
There is yet a third example where one of An Shigao’s translations ends up,
intentionally or not, associating sparśa not with karmic fruits from the past but
with karmic causes in the present. The word geng 更, “to encounter something
bad” (translated in the table above as “to suffer”) was one of An Shigao’s most
frequently used translations of sparśa. In the YCRJ, however, it translates not
sparśa but cetanā, “volition,” a concept usually associated with karmically pro-
ductive mental activity,43 that is to say, with the factors that are the cause of
future rebirth as opposed to those that are the fruits of past karma.
The context in the YCRJ is an analysis of the fourth skandha, “conditioned
factors” (saṃskāra; xing 行). The relevant passage of the YCRJ, and the Pāli par-
allel from the sixth book of the Peṭakopadesa,44 are as follows:
行種名為身六更。一色所更,二聲所更,三香所更,四味所更,五
通45 所更,六法所更。是為身六更。是名為行種。(T.15.603: 173b16–
19)
that the mysterious “activities” (行) were active during (in their Indic terminology)
saṃskāra, sparśa, and upādāna. This is precisely the same overlap that we see in the
AHKJ’s discussion of the “sprouts” (zai 栽).
43 “It is intention, O monks, that I declare to be karma. Having intended, one makes karma
through body, speech, and mind”. cetanāhaṃ, bhikkhave, kammaṃ vadāmi. cetayitvā
kammaṃ karoti kāyena vācāya manasā (AN III 415). On the interpretations of this oft-
cited passage, see McDermott, 1980: 181–189.
44 On the Peṭakopadesa as a parallel to the YCRJ, see Zacchetti, 2002.
45 The Taishō faithfully reproduces the second Kǒryo edition reading chu 觸. The first Kǒryo
edition reads the same (<http://kb.sutra.re.kr/ritk>, accessed July 2012; photographs of a
copy of the first Kǒryo edition of this text were available here from 2011–2013, but have
since been removed). However the Taishō records the variant tong 通 in the Sixi (宋) and
Puning (元) editions, and this is shared by the Jin 金 edition (ZHDZJ, 36:130). This must be
correct because the commentary, in all editions, comments on the character tong 通, not
chu 觸. The word here must correspond to spraṣṭavya, “tangibles,” translated as 更 in the
previous paragraph. Chu 觸 becomes the translation of sparśa and spraṣṭavya beginning
only with the translations of Kumārajīva in the late fourth century. The versions of the
YCRJ represented in the Kǒryo editions have thus probably been “corrected” based on this
later usage. This anachronistic chu 觸 occurs in two other places in the text (T.15.603:
174c7, c18). Though the Jin edition shares these readings, the first Kǒryo edition reads, in
both cases, shen 身, agreeing with the Sixi, Puning, and Kunaichō (宫) editions. I suspect
that shen 身 is again correct. This might be confirmed by the commentator’s initial gloss
262 Greene
The correspondence between the texts is, as elsewhere, nearly exact.46 The
meaning of the Pāli version is straightforward:
of tong 通 as shen 身 (T.33.1694: 10a19), which seems to assume that the reader is familiar
with shen 身 as a translation of spraṣṭavya.
46 Tattha cha cetanākāyā saṅkhārakkhandho (xing zhong ming wei shen liu geng 行種名為
身六更) rūpasañcetanā (yi se suo geng 一色所更) yāva (er sheng suo geng, san xiang suo
geng, si wei suo geng, wu tong suo geng 二聲所更,三香所更,四味所更,五通所更)
dhammasañcetanā (liu fa suo geng 六法所更) ime cha cetanākāyā (shi wei shen liu geng
是為身六更) ayaṃ saṅkhārakkhandho (shi ming wei xing zhong 是名為行種).
47 In the expression se suo geng 色所更, the word suo 所 (“place”) might be an attempt to
explain rūpasañcetanā as a locative or genitive dependent compound (“sañcetanā with
respect to rūpa”).
48 That geng 更 was seen by An Shigao as a plausible translation for a term like cetanā is
confirmed elsewhere by his translation of samyaksaṃkalpa, the second member of the
eightfold path, as zhi geng 直更 (T.1.57: 852a13). Saṃkalpa is usually understood as similar
in meaning to cetanā.
Pratītyasamutpāda in the Translations of An Shigao 263
彼癡因緣行為何等?為六望受。何等為六?色、聲、香、味、身50 、
法,是為身六望受。是名為行。51
What are these “activities” that are conditioned by confusion? They are
the six “yearnings”. What are the six? Visible forms, sounds, smells, tastes,
bodily [sensations], and dharmas. These are the six yearnings of the body.
These are called the “activities”.
In this case, we cannot rely on the parallel from the Peṭakopadesa because
the corresponding section is missing.52 However the definitions given
immediately after the above passage for vijñāna and vedanā (as members of
pratītyasamutpāda) are identical to those found in the earlier passage where
these same terms were defined as members of the five aggregates (skandha).53
This strongly suggests the definition of saṃskāra here was also identical to its
earlier definition as one of the skandhas (the passage discussed above), as the
six “collections of volition” (cetanākāyā). Indeed once we take the otherwise
obscure term wangshou 望受, translated above as “yearning,” simply as a differ-
ent translation of cetanā, the two passages are effectively identical.
It is quite remarkable that the YCRJ here repeats an evidently identical
passage but translates its key terminology differently. And while we cannot
know for certain what motivated An Shigao, as the list of pratītyasamutpāda
49 An Shigao uses shen 身 in this manner throughout the YCRJ, and elsewhere (see for exam-
ple Vetter and Harrison, 1998, passim). It would seem that An Shigao’s Chinese readership
did not understand his idiosyncratic (to put it mildly) usage (see the YCRJZ’s explanations
of the “six collections of consciousness” liu shen shi 六身識; T.33.1694: 14a18–19; this has
been noted by Zacchetti, 2010b: 157).
50 Here following the first Kǒryo, Sixi, Puning, and Kunaichō editions. The second Kǒryo
edition reads chu 觸 (which as I suggest above must be wrong; see note 45).
51 T.15.603: 174c6–7.
52 The absence does not appear to be intentional (Ñāṇamoli, 1964: 160). In light of the YCRJ,
Ñāṇamoli’s suggestion that the missing passage might have referred to the three-fold divi-
sion of saṅkhāra as body, speech, and mind is probably not correct.
53 In the Peṭakopadesa, the definitions of vijñāna and vedanā as a members of pratītyasa
mutpāda are abbreviated (as the “six collections” of each term), thus referring back to the
full definition given under the analysis of the skandhas (Peṭ, 116/459 and 463).
264 Greene
continues and we come to sparśa, we find that it too now receives a new
translation, namely siwang 思望, “distantly longing”.54 It is again striking that
cetanā (wangshou 望受) and sparśa (siwang 思望) have been translated into
reasonably similar Chinese words. By itself the similarity between these two
terms would perhaps not be enough to attract our attention. But considered
together with the earlier translation of cetanā as geng 更 that also overlaps
An Shigao’s most common translation of sparśa, and the similar overlapping
use of “sprouts” (zai 栽) elsewhere, we might be justified in seeing the pair-
ing wangshou 望受 and siwang 思望 as another example of the way that An
Shigao’s translations tend to blur the distinction between sense contact itself
(sparśa) as the fruit of past karma, and karmically productive factors such as
craving, grasping, or volition (cetanā).
An Shigao’s translations thus often seem to create the potential for a Chinese
reader to see sparśa, sensory “contact,” as something very similar or even iden-
tical to things such as upādāna, saṃskāra, or cetanā, terms that in Indian
interpretations of pratītyasamutpāda are usually presented as forms of karmic
activity, and hence as strictly different from sense contact itself (understood as
the fruit of past karma).
This blurring of the line between sparśa and karmically productive desire
was, I believe, part of a more general pattern in the earliest Chinese interpreta-
tions of Buddhist theories of sense perception. This pattern is revealed most
clearly through the vocabulary that appears in third-century Chinese composi-
tions as glosses on, or even as alternative translations of, the word āyatana. Our
most important source in this regard is the YCRJZ, one of our best windows
onto mid-third-century Chinese Buddhism, and a text that attests to the con-
tinuing influence and legacy of An Shigao’s translations during this time.55
As mentioned above, within An Shigao’s translations the word āyatana is
rendered in two distinct ways: as the “entrances” (ru 入) and the “destroyers”
(shuai 衰). Within translated texts, these remain the preferred Chinese terms
54 The word siwang 思望 is attested from the mid-Han (HYDCD, 2:4254). Its meaning in the
late second and early third centuries comes out clearly in a passage from the Taizi ruiying
benqi jing 太子瑞應本起經: “[Beings] are pulled about by desire, confused by material
acquisitions and sex, and long for comfort and pleasure. From this arises the root of all
evils” 牽於愛欲,惑於財色,思望安樂,從是生諸惡本。 (T.3.185: 478b9–10).
55 On the YCRJZ, see Zacchetti, 2010b.
Pratītyasamutpāda in the Translations of An Shigao 265
throughout the Three Kingdoms period (until ca. 280 CE). However the YCRJZ,
a Chinese composition rather than a translation, uses quite different terminol-
ogy. Both when glossing the more standard translations found in the root text
(the YCRJ), and in other, independent discussions, the “external” āyatanas (the
sense objects) are mostly commonly called the “six objects of desire” (liu yu
六欲)56 or the “six perversities” (liu xie 六邪),57 while the “internal” āyatanas
(the sense organs) are, almost invariably, termed the “six emotions” (liu qing
六情).58
The persistent use of the term “six emotions” to refer to the sense organs is
especially noteworthy because of the important role played by the concept of
qing 情 in pre-Buddhist Chinese discussions of human nature.59 At the risk of
glossing over some of the nuances, we may generally translate qing as “emo-
tion,” and early Chinese philosophical texts contain various lists of six or seven
such emotions such as joy, anger, sadness and so forth. There was much discus-
sion among early Chinese thinkers about the nature and status of the qing.
Most important for our purposes is that by the early Western Han dynasty (206
BCE–9 CE) the qing had come to be understood not simply as those human
emotions that were natural and unlearned (the primary meaning before this
time), but rather those desires that are problematic. Typical examples of the
qing thus came to include things such as “greed” (tan 貪) and other factors
associated with our baser, animal natures.60 By the Eastern Han (25–220 CE),
the qing had further come to be linked to the yin 陰 or “dark” aspects of human
qi 氣 (“vital energy”),61 which were understood as the factors within the body
that cause death and decay (this provided a physiological explanation for the
dangers of desire).62 The understanding thus came to be not simply that the
qing were human emotions in need of regulation or control, but that human
beings would be better off without them all together. This idea is embodied in
the famous dictum that “sages” (sheng 聖) such as Confucius were sages pre-
cisely because they were “without qing” (wu qing 無情), that is to say free of
problematic human desires and emotional responses.63
Qing 情 was thus a central concept in pre-Buddhist discussions of desire,
emotion, and human perfection, and its appearance as a technical term within
the YCRJZ reflects an effort to find ways of talking about Buddhist theories of
sense perception and desire that would resonate with a more general Chinese
audience. Many scholars have indeed noted the tendency of key third-century
Buddhist texts and authors – especially those associated with the Wu 吳
Kingdom period (220–280 CE), as is the YCRJZ – to freely draw from a range of
Chinese philosophical concepts in their attempts to explicate Buddhist ideas.
It is in Wu-Kingdom Chinese Buddhist texts that we thus find the first exam-
ples of interpretations of Buddhist idea of rebirth that make use of terms
derived from pre-Buddhist Chinese understandings of the fate of the spirit
after death such as shen 神, shishen 識神, hunling 魂靈, and shenling 神靈.64
Since qing referred, in the Chinese context, to the problematic desires that
perfected human beings have eliminated, in many ways it would have been a
perfect translation for terms such as tṛṣṇā or kleśa that refer to the craving and
other mental defilements that the practice of Buddhism is supposed to eventu-
ally eliminate. Yet qing was never used to translate or gloss such terms. When
used in a technical context, this word appears in early Chinese Buddhist litera-
ture exclusively as a collective designation of the six internal āyatanas, the
sense organs themselves. It is not clear how or when this word came to be used
in this manner, though it does not seem to have happened prior to the Wu
Kingdom period.65
Regardless of the details of its first appearance, what is most relevant about
the interpretation of the internal āyatanas as the “six emotions” is that it seems
63 The classic treatment of this subject remains Tang, [1957] 2001 (see especially p. 101–102).
See also Wang, 1987: 372–375; Wu, 2009: 9–30.
64 Zacchetti, 2010b: 171–176. Zacchetti aptly refers to all of these and other similar terms as
“soul-language.”
65 Using Nattier’s list of likely authentic Han and Three-Kingdoms Buddhist translations
(Nattier, 2008), there are only two possible examples earlier than the Wu Kingdom. First
is the Zhong benqi jing 中本起經 (T.4.196: 152a9), but as Nattier points out this text was
most likely revised in light of the other biographies of the Buddha that appeared during
the Three Kingdoms period (Nattier, 2008: 109). The other example is the AHKJ (T.25.1508:
53c13). However the AHKJ uses the expression liu qing 六情 not for the āyatanas, but for
the six sensory consciousnesses (vijñāna). It is possible that this usage eventually
morphed into what we find in later texts such as the YCRJZ.
Pratītyasamutpāda in the Translations of An Shigao 267
to accord with, and indeed further reinforces, the ambiguity between sparśa as
raw sense “contact” and karmic forces such as craving, an ambiguity that, as we
have seen, was a persistent feature of An Shigao’s translation of the terms of
the pratītyasamutpāda formula.
The YCRJZ, following the root-text’s listing of the six internal āyatanas and
their corresponding objects, thus explains their interaction by noting that “the
six internal emotions grasp the six external objects of desire” (nei liu qing shou
wai liu yu ye 內六情受外六欲也).66 As noted above, merely describing the
basic interaction between sense organs and sense objects as “grasping” (shou
受) already suggests a certain conflation of sense contact and karmic desires
because “grasping” was also the standard translation of upādāna. The glosses
of “six emotions” and “six objects of desire” for the sense organs and their
objects reinforces this perspective – under this description sense perception is
not an event that makes available a possible object of desire, but rather simply
is the meeting of human desires with the objects of those desires.
Indeed, elsewhere the YCRJZ describes the same interaction using not shou
受 but ai 愛, “craving,” the universal Chinese translation of tṛṣṇā: “the internal
emotions [āyatanas] crave the objects of desire” (nei qing ai yu 內情愛欲).67 Or,
in another passage, “the emotions [āyatanas] of a human being hunger and
thirst for the six objects of consumption” (ren qing ji ke yu liu fu 人情飢渴于六
服), with the sense objects themselves here likened to the food and drink.68
And in keeping with this understanding, the YCRJZ often describes the hoped
for fruit of Buddhist practice as the opposite of this, as a state where the six
emotions “no longer grasp the six desirable objects” (bu fu shou wai liu yu 不復
受外六欲).69 Or, as another passage has it, where “the six emotions no longer
seek out anything in the three spheres” (liu qing bu fu yu san jie zhong you suo
qiu suo ye 六情不復於三界中有所求索也).70
The difficulty that arises when trying to understand these explanations in
terms of normative Indian Buddhist understanding is that by glossing the word
āyatana as “emotion” (qing 情), the YCRZJ seems to attribute desire and craving
to the sense organs themselves. Indian Buddhist presentations of sense per-
ception do, of course, often take on a similar tone, and there are certainly
66 T.33.1694: 11a2–3; some versions here read “crave” ai 愛 in place of “grasp” shou 受.
67 T.33.1694: 15a3.
68 Liu fu 六服 for the six sense objects is not attested elsewhere. The image here is perhaps
drawn from Liu du ji jing 六度集經, T.3.152: 15c23–24.
69 T.33.1694: 19c15; here the context is a description of the fourth dhyāna.
70 T.33.1694: 18b17. Note that “to seek for” (qiu 求), here describing how the sense organs find
and make contact with their objects, is also one of An Shigao’s translations of upādāna
(see 9b in the table above).
268 Greene
71 We find this, for example, in the extended treatment of pratītyasamutpāda within the Pāli
Vibhaṅga, where both wholesome (kusala) and unwholesome (akusala) states of mind
(citta) are analyzed; the formula remains the same up until taṇhā, which in wholesome
states of mind is replaced by other terms (Vibh: 135–192).
72 In the YCRZJ, this same principle seems to apply to mental objects, and the root text’s fa
法 (dharmas, the object of the mental organ) is glossed with terms like “defiled thoughts”
(xie nian 邪念; T.33.1694: 9c9–10, c15–16; in the first passage it seems likely that xie rong
邪榮 is a mistake for xie nian 邪念). Here the YCRJZ seems to imply that objects of the
mind are not simply mental objects in the abstract, but specifically those defiled objects
that one must eliminate (AHKJ similarly uses the term nianyu 念欲, “thoughts concern-
ing desire,” rather than fa 法; T.25.1508: 54b3).
73 T.33.1694: 17a29.
Pratītyasamutpāda in the Translations of An Shigao 269
5 Conclusions
The material examined above has allowed us to move through the complete
cycle of translation and interpretation, from the original Indic terms of the
pratītyasamutpāda formula, to their Chinese translations in the works of An
Shigao, and finally to the interpretation of these translations by early Chinese
commentators. Tracing this process has allowed us to discern one key area
– notably the concept of sparśa, sensory “contact” – where An Shigao may
have struggled to find ready Chinese equivalents, and where, moreover, the
bewildering diversity of his translations may have encouraged interpreta-
tions potentially at odds with what I have styled the “normative” Indian
understanding.
However while something interesting was thus clearly going on here, in
both An Shigao’s translations and in Chinese commentaries such as the YCRJZ,
it is more difficult to determine precisely what accounts for these unusual
translations and interpretations. Do they represent some kind of Chinese mis-
understanding that was reflected even in An Shigao’s translations themselves
as his assistants struggled to find Chinese concepts to express the potentially
mysterious Indian Buddhist notion of sense “contact” (sparśa)? Or do they
reflect An Shigao’s own attempts to inflect his presentation of these ideas with
the understanding of certain Indian commentarial traditions? Indeed, we saw
some examples – such as the translation of āyatana as “destroyer” (shuai 衰) –
that might be understood as “exegetical” translations, translations that add a
layer of explanation or glossing within the translated words themselves.
Although further investigation of this must await another occasion, it might
be possible to explain An Shigao’s treatment of the term sparśa in a similar
fashion. We know that the precise nature of sparśa was the subject of consider-
able debate within the Abhidharma and early commentarial tradition,74 and
sense perception more generally was a topic of recurring controversy, between
Buddhist schools as well as between Buddhists and adherents of other Indian
philosophical systems.75 It is possible that An Shigao was well versed in some
version of these debates, and that his less-than-straightforward translations of
sparśa were attempts to somehow include the latest Indian scholastic under-
standings within the Chinese words he used. For example, in many Indian
Buddhist commentarial traditions the category of saṃskāra eventually comes
to include almost all otherwise unclassified mental or pseudo-mental fac-
tors, including sparśa. This could have something to do with the occasional
76 T.33.1694: 11a16–17. The Indian commentarial tradition does eventually feel the need to go
beyond the basic list of the pratītyasamutpāda formula and explain how sensory con-
sciousness and the material elements of perception (the organs and their objects) can
possibly come together at all. The need for such an explanation seems to arise from, and
reflect, the deep Buddhist intuition that consciousness cannot arise merely on the basis
of material causes (Griffiths, 1986). Thus in many commentarial traditions sparśa is said
to depend not only on the sense organs and their objects (as one might expect if sense
activity were an entirely “passive” affair), but also on what we might call “attention,” often
manasikāra or, in a term found as far as I know only in the Pāli commentaries, “adverting”
(āvajjana; for a discussion of these issues, see Sarachchandra, 1958: 26–27). However
despite this recognition that the causes and conditions of sense experience must also
include ontologically mental factors, those factors ultimately posited in such a role, such
as “adverting,” are never explained as karmically productive, and they are strictly distin-
guished from defilements such as craving.
77 This contrast between the “active” and “passive” members of the chain is even encoded
grammatically, as past participle versus agentive noun, within the Abhidharmasamuc-
caya’s categories of “projected members” (ākṣiptāṅga) versus “actualizing members”
(abhinirvartakāṅga; see above note 34)
Pratītyasamutpāda in the Translations of An Shigao 271
“know” (zhi 知) them, or “long for” (si 思) them.78 Human desire for sense
objects is, in this view, a matter of the sense organs themselves desiring their
objects – the “eyes are fond of beauty” (mu hao se 目好色)79 – not a matter of
the mind desiring something seen, heard, or otherwise sensed with the organ
in question (grammatically, Indian Buddhist texts invariably phrase sense per-
ception by placing the organ of perception in the instrumental case).
In pre-Buddhist Chinese understanding the sense organs themselves were
thus thought of as active, willful agents. As Mark Edward Lewis puts it:
The independence of the desires of the sense organs is an integral part of the
standard metaphor for the body in early China, that of a state in which the
sense organs are the “officials” (guan 官) and the mind is the ruler.81 And as
Lewis again observes, “the political image of the body suggests that certain
parts of the body would, like certain officials, do as they pleased and create
chaos if not rigorously controlled”.82
In other words, in early China the senses were not seen merely as windows
onto the world, but as localized embodiments of particular forms of desire, as
independent centers of willful activity. In keeping with this understanding, the
normal way of discussing the problem of desire was by arguing that one willful
agent, the ruler (the mind), must be strong enough to control the various other
willful agents that are dispersed throughout the body. This model is, I would
suggest, fundamentally different from the typical Indian Buddhist “solution” to
this problem, in which there is only ever a single willful agent (usually the mind)
78 Geaney, 2002: 51. For “to long for” (si 思), see Xiang Xiu’s 向秀 (211–300) Nan Ji Shuye yang
sheng lun 難嵇叔夜養生論: “the mouth longs for the five flavors, the eye for the five
colors” (kou si wu wei, mu si wu se 口思五味,目思五色; QSGW, 2:1876). That si 思 also
figures among An Shigao’s translations of sparśa strongly suggests that he was, at least
occasionally, borrowing the vocabulary used in typical Chinese discussions of the sense
organs.
79 Xunzi ji jie: 338.
80 Lewis, 2006: 39.
81 Sato, 2003: 154–161.
82 Lewis, 2006: 40.
272 Greene
who must restrain the desires that arise in response to sensory perception.
This is the model that is encoded in the pratītyasamutpāda formula, and it is a
model that is reflected in – that is in fact enabled by – the conceptual distinc-
tions afforded by vocabulary such as āyatana, sparśa, vedanā, and tṛṣṇā.83
The extent to which the pre-Buddhist Chinese model of the senses and their
relationship to desire influenced An Shigao’s choice of translation terminology
must remain a topic for further research. What we may say at this point is that,
intentionally or not, certain ambiguities An Shigao introduced into his transla-
tions of key members of the pratītyasamutpāda formula potentially brought
the Buddhist model closer to the Chinese one by associating sparśa with kar-
mically productive desires and defilements. And this seems to have encouraged
or enabled early Chinese readers of An Shigao’s translations to make sense of
these arguably fundamental Buddhist classification systems by drawing from
established Chinese models of human desire and its relationship to sensory
perception.
It is important to again reiterate the still primitive state of our knowledge of
An Shigao’s doctrinal background. Our broader grasp of post-canonical Indian
Buddhist commentarial traditions is, of course, considerably better. Here too,
however, we must acknowledge that the relevant surviving material, from
which we can derive what I have called the “normative” Indian understanding
of a topic such as pratītyasamutpāda, is at most only a fraction of what once
circulated. It is thus far too early to judge which elements of the peculiar inter-
pretations found in a text such as the YCRJZ are best seen as distinctly Chinese
and which elements might simply accurately reflect, albeit using Chinese
vocabulary, An Shigao’s own understanding of the relevant Buddhist concepts
(an understanding perhaps quite different from what we might have excepted
for an Indian Buddhist). Nevertheless, as I have hinted throughout this chapter,
the commentaries of An Shigao’s Chinese followers, and even in some cases An
Shigao’s translations themselves, can be profitably examined as stemming
from a distinctly Chinese environment. This material can be fully explained
only as the product of the complex and multifaceted, yet still very little under-
stood process of negotiation, insight, confusion, surprise, and misunderstanding
– as well as genuine cross-cultural communication – that accompanied the
very earliest (and ultimately, perhaps, all) translations of Indian Buddhist texts
into Chinese.
83 In a greatly modified form, this same intuition seems even to underlie the much later
theories of Dignāga and Dharmakīrti concerning the non-conceptual (kalpanāpoḍha)
nature of sense perception.
Pratītyasamutpāda in the Translations of An Shigao 273
Abbreviations
Bibliography
Barua, Arabinda (ed.) (1964). Peṭakopadesa. London: The Pāli Text Society.
Bo shu Laozi jiao zhu 帛書老子校注 (1996). Gao Ming 高明 (ed.). Beijing: Zhonghua
shuju.
Bodhi Bhikku (2000). The Connected Discourses of the Buddha. Somerville MA: Wisdom
Publications.
Bronkhorst, Johannes (1986). The Two Traditions of Meditation in Ancient India. Stuttgart:
Franz Steiner Verlag.
Bucknell, Roderick S. (1999). “Conditioned Arising Evolves: Variation and Change in
Textual Accounts of the Paticca-Samuppāda Doctrine”, Journal of the International
Association of Buddhist Studies 22/2, pp. 311–342.
274 Greene
Hartmann, Gerda (1940). “Symbols of the Nidānas in Tibetan Drawings of the Wheel of
Life”, Journal of the American Oriental Society 60, pp. 356–360.
Hu, Chirui 胡敕瑞 (2002). Lun Heng yu Dong Han Fodian ciyu bijiao yanjiu 《論衡》與
東漢佛典詞語比較研究. Chengdu: Bashu shushe.
Johansson, Rune E.A. (1979). The Dynamic Psychology of Early Buddhism. London:
Curzon Press.
Jong, Jan Willem de (1979). “The Daśottarasūtra”, In Buddhist Studies by J.W. De Jong.
Edited by Gregory Schopen. Berkeley CA: Asian Humanities Press, pp. 251–273.
Jurewicz, Joanna (2000). “Playing with Fire: The Pratītyasamutpāda from the Perspective
of Vedic Thought”, Journal of the Pāli Text Society 26, pp. 77–104.
Kritzer, Robert (1999). Rebirth and Causation in the Yogācāra Abhidharma. Wien:
Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien.
La Vallée Poussin, Louis de (1913). Bouddhisme, Études et Matériaux, Théorie des Douze
Causes. Londres: Luzac & Co.
Lewis, Mark Edward (2006). The Construction of Space in Early China. Albany NY: State
University of New York Press.
Lunheng jiao shi 論衡校釋. Huang, Hui 黄暉 (ed.) (1990). Beijing: Zhonghua shuju.
Luo Zhufeng 罗竹风 (ed.) (2000). Hanyu da cidian: suoyinben 汉语大词典: 缩印本.
Beijing: Zhonghua shuju.
McDermott, James P. (1980). “Karma and Rebirth in Early Buddhism”, In Karma and
Rebirth in Classical Indian Traditions. Edited by Wendy Doniger O’Flaherty. Berkeley
CA: University of California Press, pp. 165–192.
Morohashi, Tetsuji 諸橋轍次 (1955–1960). Dai Kan-Wa jiten 大漢和辞典. 13 vols. Tokyo:
Taishukan shoten.
Morris, Richard and Hardy, Edmund (ed.) ([1885–1900) 1999–2013). Aṅguttaranikāya. 5
vols. London: The Pāli Text Society.
Müller, Edward (ed.) (1979). Atthasālinī. Revised edition. London: The Pāli Text Society.
Nakamura, Hajime (1980). “The Theory of Dependent Origination in Its Incipient Stage”,
In Buddhist Studies in Honor of Walpola Rahula. Edited by André Bareau. London:
Gordon Fraser, pp. 165–172.
Ñāṇamoli Bhikkhu (1964). The Piṭaka-Disclosure (Peṭakopadesa). London: The Pāli Text
Society.
Nattier, Jan (2008). A Guide to the Earliest Chinese Translations: Texts from the Eastern
Han and Three Kingdoms Periods. Tokyo: The International Research Institute for
Advanced Buddhology.
Ng, Ong-cho (1998). “Is Emotion (Qing) the Source of a Confucian Antinomy?”, Journal
of Chinese Philosophy 25/2, pp. 169–190.
Ochiai Toshinori 落合俊典 (2004). Kongōji Issaikyō no Kisoteki Kenkyū to Shinshutsu
Butten no Kenkyū: Heisei 16-nen 3-gatsu 金剛寺一切経の基礎的研究と新出仏典の
276 Greene
Zacchetti, Stefano (2008). “The Nature of the Da Anban Shouyi Jing 大安般守意經 T602
Reconsidered”, Journal of the International Association of Buddhist Studies 31/1–2,
pp. 421–484.
Zacchetti, Stefano (2010a). “Defining An Shigao’s 安世高 Translation Corpus: The State
of the Art in Relevant Research”, In Xiyu lishi yuyan yanjiu jikan disan ji 西域歷史語
言研究集刊第三輯. Edited by Shen, Weirong 沈衛榮. Beijing: Kexue chubanshe,
pp. 249–270.
Zacchetti, Stefano (2010b). “Some Remarks on the Authorship and Chronology of the
Yin Chi Ru Jing Zhu 陰持入經註 T 1694 – the Second Phase in the Development of
Early Chinese Buddhist Exegetical Literature”, In Buddhist Asia 2. Edited by Silvio
Vita. Kyōto: Italian School of East Asian Studies, pp. 141–198.
Zürcher, Erik ([1959] 1972). The Buddhist Conquest of China. The Spread and Adaptation
of Buddhism in Early Medieval China. Reprint. Leiden: Brill.
Zürcher, Erik (1979). “Late Han Vernacular Elements in the Earliest Buddhist Translations”,
Journal of the Chinese Language Teachers Association 12/3, pp. 177–203.
Abhidharma in China 279
Chapter 8
1 See Zürcher, 2007: 22; Ch’en, [1972] 1973: 29–30; Nattier, 2008: 35. The accounts of the dream
of Emperor Ming derive from the “Preface to the Sutra in Forty-two Sections” (Sishi’er zhang
jing xu 四十二章經序) which was, in the sixth century, incorporated in Sengyou’s 僧祐 Chu
sanzang ji ji 出三藏記集 (T.55.2145: 42c15–19). As noted by Storch, 2014: 25, “The authorship
and the date of this sutra have been debated since the beginning of the twentieth century”.
For a first major study on this legend: See Maspéro, 1910. For more legends concerning the
introduction of Buddhism in China: see Ch’en, [1972] 1973: 27–29.
2 See Zürcher, 2007: 26–27.
3 See Loewe, 1986: 670–671; Zürcher, 2007: 26.
The variant of Confucianism that had been elevated to the status of state
orthodoxy in 136 BCE was an interpretation of the early doctrine that was
highly influenced by numerological and ‘Daoist’ concepts.4 This highly syncre-
tistic from of Confucianism was called Jin wen jia 今文家 (New Text School
Confucianism).5 The importance of this peculiar type of Confucianism as
orthodoxy for the development of Buddhism in China is twofold. A first ele-
ment of importance is that this doctrine developed to be a part of Chinese
self-identification. By the time of the Zhou 周 dynasty (1046–256/221 BCE), a
hierarchical feudal order headed by the ‘Son of Heaven’ (tianzi 天子) had
developed in the central plains (zhongyuan 中原) along the Yellow River. This
‘Son of Heaven’ was thought to rule by divine mandate (tianming 天命).6 When
Confucianism was established as state doctrine, this Zhou concept of rulership
was given a Confucian interpretation: the now Confucian rule was seen as a
divine undertaking, and ‘governing’ was interpreted as ‘maintaining the har-
monious order between Heaven, Man and Earth’.7 This is clear from the
explanation of the character for ‘wang’ 王, ‘ruler’ (generally translated as ‘king’)
given in Xu Shen’s 許慎 (ca. 58–ca. 147) Shuo wen jie zi 說文解字, the oldest
extant etymological dictionary of the Chinese language, dated 100 CE. Here we
read that, according to Dong Zhongshu who had been a major exponent of the
development of New Text School Confucianism, it is so that “when depictions
(wen 文) were created in olden [times], three strokes that were connected
through the middle were called wang. The three are Heaven, Earth and Man,
and the one who connects them is the ‘ruler’ (wang)”.8 This concept also
4 Dull, 1994: 3 remarks that the reign of Emperor Wu 武 (140–87 BCE) of the Han dynasty is the
period in which Confucianism for the first time was recognized as the ‘ism,’ “to the exclusion
of all others, that was to be acceptable to the state and was to become the object of study for
those who hoped for official careers”. Schwartz, 1985: 377 calls Emperor Wu of the Han the
‘Constantine of Confucianism,’ who followed the momentous advice of Dong Zhongshu 董
仲舒 (179?–104? BCE) and others to establish the “five classics” of Confucianism as the founda-
tion of all official education and to proscribe all “unorthodox” doctrines. The term ‘daojia’ 道
家, ‘Daoist school’ only started to be used in the second century BCE. See Cheng, 1997: 103.
5 On the element ‘jia,’ ‘school’: see further and p. 11 note 48.
6 According to Michèle Pirazzoli-t’Serstevens, [2008] 2009: 302, although the doctrine of the
Mandate of Heaven (tianming) had not been taken into account before Emperor Chengdi 成
帝 (r. 33–7 BCE) or as full-fledged theory before the time of the usurper Wang Mang 王莽 (r.
5–23 CE), the concept had been operative even before the first century BCE.
7 Yu, 2005: 34 remarks that the politico-religious narrative of Confucianism, in fact, builds on
the system that was developed already in the Shang 商 dynasty (trad. 1766–1122 BCE), when
ancestors were transformed from kin to symbols of divine power.
8 Xu, [1981] 1988: 7b. This also explains why Han Confucianism has been labeled ‘cosmological
Confucianism’ by Joseph Needham (1958: 281–282).
Abhidharma in China 281
implies that it is from Heaven that the ruler, the ‘Son of Heaven,’ obtains his
mandate to rule, and that it therefore also is Heaven that can, ultimately, with-
draw this mandate.9 Phrased differently, it is through divine approval of his
moral virtue – philosophically articulated in the Confucian concept ren 仁,
‘humaneness’ – that the ‘Son of Heaven’ maintains his legitimacy to rule.10 The
connection between the indigenous Confucian political ideology and the age-
old acceptance of a divine origin of the ‘Chinese’ people made it possible that
the inhabitants of the central plains imagined themselves and the territory
they inhabited (that is: Zhongguo 中國) as fundamentally different from the
people, creeds and customs of the regions surrounding the central plains.11
A second element of importance is that Confucian state orthodoxy in its
New Text School variant was, as mentioned, a syncretistic doctrine. When Han
imperial power was usurped by Wang Mang 王莽 in the first decade CE, a
period of gradual decline of central authority set in. This development must
have come as a disenchantment for the Confucian literati for whom the uni-
fied Confucian empire had gradually become perceived as normative. On a
political level, such new independent religious groups as the Wudoumi dao 五
斗米道 (Way of the Five Pecks of Grain) or the Taiping dao 太平道 (Way of
Great Peace) presented themselves as an alternative for the Confucian order.12
On an ideological level, these groups even managed to develop Daoism into a
unified state-sponsored religion, capable to compete with Buddhism. As the
Chinese intellectuals were, through the New Text School interpretation of
Confucianism, acquainted with Daoist concepts, it was possible for them to
reinterpret these concepts within a Confucian framework. This intellectual
Among the first translations of Buddhist scriptures into Chinese as the out-
come of a systematized translation project, we not only find popular sūtras on
the life of the Buddha or jātaka (birth) stories that were rendered in a simple
language,17 but also translations of the highly technical and intellectual
Abhidharma treatises. This hints to it that not only the common people had
become interested in Buddhism, but that also the former (New Text School)
Confucians had taken up an interest is this new doctrine and were in need of
Chinese versions of texts of this erstwhile ‘foreign’ creed in China.18
It is An Shigao 安世高 (fl. 148–168), a Parthian prince who renounced the
throne he was supposed to inherit from his father and who came to Luoyang 洛
陽 in the mid-second century CE, who is credited with the first Abhidharma
translations.19 According to Sengyou’s 僧祐 Chu sanzang ji ji 出三藏記集, it
more precisely is An Shigao’s translation of the Apitan wufa xing jing 阿琵曇五
法行經, done during the reign of Emperor Huan, that introduced the knowl-
edge of the Abhidharma in China.20
It would take a few centuries, however, before Abhidharma studies started
to gain popularity. Especially the period of Emperor Fu Jian 符堅 (357–384) of
the Former Qin 前秦 dynasty (351–394), one of the sixteen kingdoms that ruled
in Northern China after the Han dynasty had finally collapsed in 220 CE, was
important for the take-off of Abhidharma translation activities. One of the
translators who were active in the capital city Chang’an 長安, is the famous
Saṃghadeva. During the years of Emperor Fu Jian, he and Zhu Fonian 竺佛念
together had started the translation of the *Aṣṭagrantha (Apitan ba jiandu lun
阿毘曇八犍度論, T.26.1543), one of the major early Sarvāstivāda Abhidharma
works.21
After political turmoil in fourth century North China had urged the Chinese
intelligentsia to move to the South, Lushan 廬山 in Jiangxi 江西, South China,
gradually developed to be an important center of Abhidharma studies.22 It was
especially the arrival of Saṃghadeva on Mt. Lu that stimulated the study of
the Abhidharma.23 It is reported that Saṃghadeva’s lectures on Abhidharma
attracted huge audiences to Mt. Lu. Among these was his sponsor Wang Xun
王珣 (350–401), one of the most important dānapatis of the period.24 Wang
Xun belonged to the inner circles of Huan Wen 桓溫 of the Eastern Jin 晉
18 For reflections on the different styles of translation of Buddhist texts and their possible
relation to readership: See Nattier, 2008: 18.
19 See Zürcher, 1991: 282, 289. On the translations of An Shigao: see Mizuno, [1982] 1995: 45;
Tsukamoto, 1985: 78–112; Zacchetti, 2010. See the contribution in this volume by Eric M.
Greene.
20 T.55.2145: 6b4–5. Zürcher, 1991 does not list this text among what he considers to be genu-
ine translations by An Shigao.
21 See T.55.2154: 620 a23–24. On the *Aṣṭagrantha / Jn͂ ānaprasthāna: See Willemen et al.,
1998: 221–229.
22 See Tsukamoto, 1985: vol.1, pp. 401, 414–415, 444; vol.2, p. 760 ff.
23 See Dessein, 2010: 49–50.
24 See T.49.2034: 329a10–19, 361b23–27; T.50.2059: 329a15, 361b24; T.55.2145: 64a17. T.50.2059:
329a10–19, 361b23–27 also mentions Wang Min 王珉 (361–388), Wang Xun’s brother,
284 Dessein
dynasty (317–419), and of the Jin emperor Xiaowu 孝武 (r. 379–397).25 The
translation of Dharmaśreṣṭhin’s *Abhidharmahṛdaya (Apitan xin lun 阿毘曇
心論, T.28.1550) Saṃghadeva did together with Huiyuan 慧遠 (334–416) in 391
and his lecturing on this text, set off the flourishing of the Mt. Lu Abhidharma
School.26 Numerous monks studied the text and lectured on it,27 and the so-
called ‘group of Mt. Lu’ were all specialists in the *Abhidharmahṛdaya.28 The
translation of the *Abhidharmahṛdaya was followed by the final translation
of Dharmatrāta’s *Saṃyuktābhidharmahṛdaya (Za apitan xin lun 雜阿毘
曇心論, T.28.1552) by Saṃghavarman in 434.29 On this Chinese translation,
numerous commentaries were written.30 The Apitan xin lun and the Za api-
tan xin lun became the two most important texts of the Chinese Abhidharma
School into the Tang 唐 dynasty.31 The Abhidharma School was superseded
by the so-called Kośa School when Xuanzang 玄奘 translated Vasubandhu’s
Abhidharmakośa (Apidamo jushe lun 阿毘達摩俱舍論) in 653.32 From these
accounts of the Buddhist activities on Mt. Lu, we can thus deduce that also
the Chinese Abhidharma School that developed on Mt. Lu was shaped in the
pattern of traditional school formation in the Chinese cultural context: recit-
ing texts and accepting the teacher’s interpretation by the disciple.33 The third
among the audience on Mt. Lu. However, as Saṃghadeva arrived on Mt. Lu only in 397,
this Wang Min cannot be the brother of Wang Xun. See Mather, 1976, 124–125.
25 See Tsukamato, 1985, vol.1: 386. Huan Wen was a descendant from a low-level military
family that had risen to power after the Western Jin dynasty had moved to the South in 316
CE and had reestablished itself as Eastern Jin. In the South, they subdued the Daoist the-
ocracy that had been established there in the aftermath of the Wudoumi movement. See
Lewis, 2007: 203–204, and Lewis, 2009: 199–200. On Huan Wen: see Lewis, 2009: 64.
26 Saṃghadeva had already begun translating the*Abhidharmahṛdaya in 388, while still in
the North. See T.50.2059: 357c23–361b13.
27 See T.50.2059: 359b23–24, T.50.2060: 529c10, 599b23, 668a24–25, 675a23–24. See also: Des-
sein, 2010: 52–54.
28 See T.49.2035: 343a16–28.
29 See T.55.2145: 74c3–7. Also see Dessein, 1999,1: lxxvii-lxxxii.
30 See T.50.2059: 244a8–9, 368b28–29, 338a19, 369a23, 373a13, 373a22–23,373b18–20, 373c1,
375a4, 376a22, 376b28–29, 378b19–22, 381a25–26,382b23-c2, 401c3, 442b7, 460a26;
T.50.2060: 442a7–8, 461a24–25, 482c25–28, 483a11–12, 495c13–16, 501b8, 502a2, 504c25,
507c21, 508a12, 508c7–14, 509a6, 510b3–4, 524b27, 527b7, 529a6–7, 530a21, 532b27, 544b21–
22, 549c24, 572a19, 594c21. See also: Dessein, 2010: 5762.
31 See T.50.2059: 359b23–24.
32 See T.2149.311c12; T.2154: 557a22–23, 620c22–23, 695c10, 720b6. Also see Watanabe, Mizuno,
and Ōishi, 1932: 123.
33 It is also with Mt. Lu that Pure Land (Jingtu 淨土) Buddhism is connected. The fourth
century thus also is the period in which original Chinese ‘schools’ (jia) of Buddhism
started to develop.
Abhidharma in China 285
When the highly technical Abhidharma texts had to be translated into Chinese,
equivalents for the Buddhist terms and concepts had to be found. As stated by
Whalen Lai (1979: 238–239):
The early translators did not all follow the same translation technique. While
the already mentioned An Shigao preferred to transliterate proper names, he
sought the best matching equivalent in Chinese philosophical vocabulary for
Buddhist terms. The Yuezhi 月氏 Lokakṣema and his disciple Zhi Qian 支謙
(born in China in the third century CE) preferred to transliterate Indian techni-
cal terms instead of translating them.35 The Parthian An Xuan 安玄 (fl. 181)
attempted to not only translate the doctrinal terms, but also the proper names.36
While, on the one hand, opting for transliterations may point to it that the
early translators were very cautious to preserve the original meaning of the
Buddhist texts and tried to avoid that rendering a Buddhist term with a Chinese
term that did not fully express the Buddhist term’s original meaning would
lead to misunderstandings of the Buddhist doctrine, the tremendous variety of
Chinese translations we find for one and the same technical term may equally
reveal the struggle of the individual translator to give the most perfect Chinese
term. The latter can be illustrated by the eight different translations and tran-
scriptions Zhi Qian gives for the term ‘arhat’: aluohan 阿羅漢, aluohe 阿羅訶,
luohan 羅漢, wusuozhuo 無所著, wuzhuo 無著, yingyi 應儀, yingzhen 應真,
zhenren 真人 and zhizhen 至真.37
Given the intellectual climate of second century China, it does not come as
a surprise that it was especially in the Daoist doctrine that equivalents for
Buddhist concepts and terms were sought. A well-known example is the use of
the word dao 道 to render the Buddhist term dharma (doctrine). This same
word dao was, however, also used to translate bodhi (enlightenment) or yoga.
The just mentioned term zhenren 真人 which was one of the words used to
translate arhat was, actually, a word used in Neo-Daoist philosophy to refer to
a Daoist immortal.38 Also the Buddhist final extinction (nirvāṇa), that is, the
state where all karmic activity has come to a halt, was easily paired with the
Daoist wuwei 無為 (non-action).39 The same wuwei was, however, also a com-
mon translation for the Buddhist notion of being ‘unconditioned’ (asaṃskṛta),
i.e. existing free from karmic influence. Not all equivalents were sought in the
Daoist sphere, though. In the translations done by Kang Senghui 康僧會 (third
century CE), a Sogdian who came to Luoyang toward the end of the reign of
Emperor Huan,40 we also find terms borrowed from popular religion and
Confucianism. Kang Senghui in this way refers to Taishan 太山 as recluse after
death, he discusses the fate of the hunling 魂靈 spirit(s), and he uses the
Confucian term ren 仁 (humaneness).41
Difficulties to find corresponding terms might have been even bigger in the
case of the ‘numerical lists’ (mātṛkā) of which the Abhidharma texts abound.
36 See Nattier, 2008: 91. On the translations by An Shigao: see the contribution by Eric M.
Greene in this volume.
37 See Nattier, 2008: 119, note 23.
38 See Wright, 1959: 36.
39 See Liu and Shao, 1992: 39–40.
40 T.55.2145: 95c24. For further biographical information on Kang Senghui: see T.55.2145:
96a29–97a17; T.50.2059: 325a13–18 and 325b4–326b13.
41 See Nattier, 2008: 151 and 151, note 113.
Abhidharma in China 287
These lists form logical units, and, although also in the Chinese cultural con-
text numerical lists were common, these did not necessarily match the
Buddhist list with respect to content. In this way, the Buddhist series of ‘great
elements’ (mahābhūta: earth (pṛthivī), water (ap), fire (tejaḥ), wind (vāyu), and
matter derived (upādāyarūpa) from the four great elements) were matched
with the Chinese ‘five phases’ (wu xing 五行: water (shui 水), fire (huo 火),
wood (mu 木), metal (jin 金), and earth (tu 土)) known from ancient Chinese
numerology, and the Confucian five constituent virtues (wu chang 五常:
humaneness (ren 仁), righteousness (yi 義), ritual (li 禮), knowledge (zhi 智),
and faith (xin 信)) were paired with the five lay precepts (no harming living
beings, no stealing, no sexual misconduct, no lying, and no intoxication).
Liu Yiqing’s 劉義慶 (403–444) Shi shuo xin yu 世說新語, a collection of short
stories and conversations compiled in ca. 430 and the best known literary
reflection of the qingtan movement, alludes to the difficulty to understand
these often enigmatic lists in the following story:
The above described technique to render Buddhist terms with a term of the
Chinese – especially Daoist – philosophical tradition is generally referred to as
‘matching meanings’ (geyi 格義).43 We mentioned that difficulties in transla-
tion must have been particularly prominent when ‘numerical lists’ of
would “deviate from the principles [of Buddhism] (yu li duo wei 於理多違)”.51
The Chu sanzang ji ji informs us that Huirui 慧叡 (fourth century), a disciple of
Kumārajīva, reportedly said that the method of geyi “deluted and perverted the
doctrine”.52 Judging from Kumārajīva’s (344–413) biography, included in the
Chu sanzang ji ji, Sengyou (435–518) appears not to have understood the mean-
ing of the term any more.53
Controversy thus must have arisen within Buddhist circles on the suitability
of the ‘geyi’ method to be applied in other than purely Abhidharmic contexts.
An example of this application of the ‘geyi’ method that goes beyond the expla-
nation of Abhidharmic categories is the preface to Kumārajīva’s translation of
the *Dvādaśadvāraka, Shi’er men lun 十二門論 (T.30.1568) done by Kumārajīva’s
disciple Sengrui 僧叡 (born in 352 CE) and Jizang’s 吉藏 (549–623) commen-
tary on this text, titled Shi’er men lun shu 十二門論疏 (T.42.1852). It may not be
without importance that while Sengrui’s introduction only quotes from the
Zhuangzi 莊子,54 Jizang’s commentary expands the references with quotations
from the Laozi 老子, from Guo Xiang’s 郭象 (died 312) Neo-Daoist commen-
tary on the Zhuangzi, and even from the Confucian Lunyu 論語 and the already
mentioned Mao shi version of the Shijing.55
It is not unlikely that it was precisely at the moment that the use of ‘geyi’ as
an exegetical technique was expanded to non Abhidharma texts, that it became
criticized by learned monks such as Dao’an.56 The questionability of the tech-
nique may also explain why as good as none of the extant Chinese translations
of Abhidharma texts uses this ‘matching meanings’ technique.57
4 ‘Geyi’ Confucianism
The period between the fall of the Han dynasty in 220 CE and the reunification
of the country under the Sui 隋 dynasty in 589 CE has, in Chinese historiogra-
phy and in Western scholarship alike, generally been described as “an age of
religious escapism, metaphysical speculation, and uninhibited eccentricity”.58
Confucian scholars for whom the possibilities to pursue their official careers
had been halted are then thought to have indulged themselves in philosophical
discussions, a trend that is generally designated as ‘qingtan,’ ‘pure conversa-
tion’. Confronted with the advent of Daoism and Buddhism, also an altogether
different reaction by the Confucian intellectuals is therefore imaginable. It is
not unlikely that along with the ‘qingtan’ movement, a ‘conservative’ move-
ment that wanted to reformulate Confucianism developed. Some literati may,
indeed, have blamed a degenerate Confucianism for the fall of the Han, and
may have vowed to become Confucian models to emulate in their own right,
and therefore follow the principles of Confucianism more strictly than those
in power.59 It is therefore also possible that we should see the ‘geyi’ exegetical
method as a coping strategy deployed by those few ‘conservative’ Confucian
literati who wanted to redefine Chinese culture in a context of growing influ-
ence of Daoism and Buddhism.60
Having proven its unsuitability to explain/translate the numerical lists of
Buddhist texts, the ‘geyi’ technique must have disappeared as a Buddhist
method not much longer than one generation after it was first used.61 Hereafter,
in fourth and fifth century China, ‘geyi’ developed to be an instrument to rede-
fine Confucian orthodoxy. This may explain why it especially were “young
members of gentry families” who “were all well-versed in the secular canons
(shi dian 世典), but had not yet become conversant with the principles of
Buddhism”62 who adhered to Zhu Faya for his classical knowledge and his
Buddhist instruction, and why Dao’an opposed to ‘geyi’ because it would “devi-
ate from the principles [of Buddhism]”.63 Seen from this perspective, so-called
fourth and fifth century ‘geyi Buddhism’ should rather be recoined as ‘geyi
Confucianism’.
5 Conclusion
From the above, we can deduce that the technique of ‘geyi’ most likely started
as a type of exegesis applied to the numerical lists that were prevalent in the
Abhidharma texts. This technique, further, was practiced in circles of monks
who had enjoyed a traditional schooling and were well-versed in the Chinese
classics. We also know that this peculiar technique was used up to the fourth
century, i.e. the period in which also such intellectual trends of qingtan
flourished.
Given the denouncement of the ‘geyi’ technique in Buddhist circles itself, it
is therefore possible to formulate a hypothesis on the nature of ‘geyi’ Buddhism
and a hypothesis on why this technique attained more attention than its fac-
tual importance allows, that differs from the generally accepted one. Rather
than that, as commonly accepted, Buddhist – especially prajñā – studies were
carried under the aegis of Chinese xuanxue thinking, it might have been, on
the contrary, Buddhism that stimulated creativity in Chinese philosophy, thus
forming xuanxue thinking and many devotees of qingtan also engaging in
xuanxue.64 This supposition is corroborated both in terms of chronology and
content. The similarities in psychological and metaphysical vocabulary
between qingtan and xuanxue facilitated this movement, and, gradually, also
Buddhist monks and Confucians who were knowledgeable in Buddhism must
have participated in these debates and conversations.65When Buddhist
Ābhidharmikas retreated from the ‘geyi’ method, it was adopted by Confucian
literati to reformulate their doctrine in the complex philosophical and reli-
gious atmosphere that preceded the reunification of the empire in the Sui
dynasty. This must have helped to pave the way for the development of so-
called Neo-Confucianism.
62 T.50.2059: 347a19–20.
63 T.50.2059: 355a25.
64 See also Lewis, 2009: 205.
65 Some eminent monks like Zhi Dun 支遁 (fourth century) even became leading figures in
the qingtan movement of the Jin dynasty. See Lewis, 2009: 205.
292 Dessein
Abbreviation
Bibliography
Ch’en, Kenneth ([1972] 1973). Buddhism in China. A Historical Survey. Princeton NJ:
Princeton University Press.
Cheng, Anne (1997). Histoire de la pensée chinoise. Paris: Seuil.
Dessein, Bart (1997). “The Chinese Reception of Indian Buddhism: Shih-erh Men Lun
Hsü (T.Vol. 30, No. 1568): Seng-jui’s Preface to Nāgārjuna’s ‘Twelve Gates Treatise’ and
Shih-erh Men Lun Hsü Shu (T.Vol. 42, No. 1825): Chi-tsang’s commentary on this pref-
ace – An analysis)”, The Indian Journal of Buddhist Studies 9/1–2, pp. 38–68.
Dessein, Bart (1999). Saṃyuktābhidharmahṛdaya. Heart of Scholasticism with Miscella
neous Additions. 3 vols. (Buddhist Tradition Series 33–35). Delhi: Motilal Banarsidass.
Dessein, Bart (2010). “The Abhidharma School in China and the Chinese Version of
Upaśānta’s *Abhidharmahṛdayasūtra”, The Eastern Buddhist. New Series 41/2,
pp. 49–69.
Dessein, Bart (2014). “Faith and Politics: (New) Confucianism as Civil Religion”, Asian
Studies II (XVIII)/1, pp. 39–64.
Dull, Jack L. (1994). “Determining Orthodoxy: Imperial Roles”, In Imperial Rulership and
Cultural Change in Traditional China. Edited by Frederick P. Brandauer and Chün-
chieh Huang. Seattle: University of Washington Press, pp. 3–27.
Fan, Ye (1976). Hou Han shu 後漢書. Beijing: Zhonghua shuju, 1978.
Harrison, Paul (1993). “The Earliest Chinese Translations of Mahāyāna Sūtras: Some
Notes on the Works of Lokakṣema”, Buddhist Studies Review 10/2, pp. 135–177.
Jansen, Thomas ([2008] 2009). “New Tendencies, Religious and Philosophical, in the
Chinese World of the Third through Sixth Centuries”, In Conveiving the Empire. China
and Rome Compared. Edited by Fritz-Heiner Mutschler and Achim Mittag. Oxford:
Oxford University Press, pp. 397–419.
Abhidharma in China 293
Lai, Whalen (1979). “Limits and Failure of Ko-I (Concept-Matching) Buddhism”, History
of Religions 18/3, pp. 238–257.
Lewis, Mark E. (2007). The Early Chinese Empires. Qin and Han. Cambridge MA and
London: Harvard University Press.
Lewis, Mark E. (2009). China Between Empires. The Northern and Southern Dynasties.
Cambridge MA and London: Harvard University Press.
Link, Arthur E. (1958). “Biography of Shih Tao-an”, T’oung pao, Second Series, Vol. 46, 1/2,
pp. 1–48.
Liu, Jiahe and Shao, Dongfong (1992). “Early Buddhism and Taoism in China (ad 65–
420)”, Buddhist-Christian Studies 12, pp. 35–41.
Liu, Yiqing 劉義慶, Shi shuo xin yu 世說新語. <http://ctext.org/shi-shuo-xin-yu/zh>.
Loewe, Michael (1986). “The religious and intellectual background”, In The Cambridge
History of China. Volume 1: The Ch’in and Han Empires. 221 bc–ad 220, Edited by Denis
Twitchett and Michael Loewe. Cambridge: Cambridge University Press,
pp. 649–725.
Lüshi chunqiu 呂氏春秋 ([1986] 1988). In Ershi’er zi 二十二子. Shanghai: Changhai guji
chubanshe.
Mair, Victor (2012). “What is Geyi, After All?”. China Report, pp. 29–59.
Maspéro, Henri (1910). “Le songe et l’ambassade de l’empereur Ming: Étude critique des
sources”, Bulletin de l’École Française d’Extrême-Orient 10/1, pp. 95–130.
Mather, Richard B. (1976). Shih-shuo Hsin-yü. A New Account of Tales of the World. By Liu
I-ch’ing, with commentary by Liu Chün. Minneapolis MN: University of Minnesota
Press.
Mittag, Achim ([2008] 2009). “Forging Legacy: The Pact between Empire and
Historiography in Ancient China”, In Conceiving the Empire. China and Rome
Compared. Edited by Fritz-Heiner Mutschler and Achim Mittag. Oxford: Oxford
University Press, pp. 143–165.
Mizuno, Kōgen ([1982] 1995). Buddhist Sutras. Origin, Development, Transmission. Tokyo:
Kōsei Publishing Co.
Nattier, Jan (2004). “Beyond Translation and Transliteration: A New Look at Chinese
Buddhist Terms”. Unpublished paper delivered at the annual meeting of the Western
Branch of the American Oriental Society. Portland, 16 October 2004.
Nattier, Jan (2008). A Guide to the Earliest Chinese Buddhist Translations. Texts from the
Eastern Han 東漢 and the Three Kingdoms 三國 Periods. (Bibliotheca Philologica et
Philosophica Buddhica X). Tokyo: The International Research Institute for Advanced
Buddhology. Soka University.
Needham, Joseph (1958). Science and Civilisation in China. Vol. 2: History of Scientific
Thought. Cambridge: Cambridge University Press.
294 Dessein
Pirazzoli-t’Serstevens, Michèle ([2008] 2009). “Imperial Aura and the Image of the Other
in Han Art”, In Conceiving the Empire. China and Rome Compared. Edited by Fritz-
Heiner Mutschler and Achim Mittag. Oxford: Oxford University Press, pp. 299–317.
Schwartz, Benjamin I. (1985). The World of Thought in Ancient China. Cambridge MA and
London: Harvard University Press.
Shils, Edward (1971). “The Intellectuals in the Political Development of New States”, In
Political Development and Social Change. Edited by Jason L. Finkle and Richard W.
Gable. New York NY: John Wiley & Sons, pp. 249–276.
Storch, Tanya (2014). The History of Chinese Buddhist Bibliography. Censorship and
Transformation of the Tripitaka. Amherst: Cambria Press.
Takakusu, Junjirō 高楠順次郎, Watanabe, Kaigyoku 渡邊海旭, and Ono Gemmyō 小野
玄妙 (ed.) (1924–1934). Taishō shinshū daizōkyō 大正新脩大藏經, Tokyo: Taishō
issaikyō kankōkai.
- T.26.1543: Kātyāyanīputra, *Aṣṭagrantha, Apitan ba jiandu lun 阿毘曇八犍度論,
Saṃghadeva, Zhu Fonian 竺佛念,.
- T.28.1550: Dharmaśreṣṭhin, *Abhidharmahṛdaya, Apitan xin lun 阿毘曇心論,
Saṃghadeva and Huiyuan 慧遠.
- T.28.1552: Dharmatrāta, *Saṃyuktābhidharmahṛdaya, Za apitan xin lun 雜阿毘曇
心論, Saṃghavarman.
- T.30.1568: *Nāgārjuna, *Dvādaśadvāraka, Shi’er men lun 十二門論, Kumārajīva.
- T.42.1825: Jizang 吉藏, Shi’er men lun shu 十二門論疏.
- T.49.2034: Fei Changfang 費長房, Lidai sanbao ji 歷代三寶記.
- T.49.2035: Zhipan 志磐, Fo zu tong ji 佛祖統記.
- T.50.2059: Huijiao 慧皎, Gaoseng zhuan 高僧傳.
- T.55.2145: Sengyou 僧祐, Chu sanzang ji ji 出三藏記集.
- T.50.2149: Daoxuan 道宣, Da Tang neidian lu 大唐內典錄.
- T.55.2154: Zhisheng 智昇, Kaiyuan shi jiao lu 開元釋教錄.
Tang Yongtong 湯用彤 (1987). Han Wei Liang Jin Nanbeichao Fojiao shi 漢魏兩晉南北朝
佛教史. 2 vols. Shanghai, 1938. Reprint: Taipei: Luotuo Chubanshe.
Tsukamoto, Zenryū (1985). A History of Early Chinese Buddhism. From its Introduction
to the Death of Hui-yüan. 2 vols. Tokyo, New York, San Francisco: Kodansha
International. (English translation of the original Japanese work by Leon Hurvitz.
Original publication: Chūgoku Bukkyō tsūshi. Tokyo: Shunjūsha, 1979).
Watanabe, Baiyū; Mizuno, Kōgen; and Ōishi, H. (1932). Abidonshinron. (Kokuyaku
Issaikyō. Bidonbu vol.21). Tokyo: Daitō Shuppansha.
Willemen, Charles; Dessein, Bart & Cox, Collett (1998). Sarvāstivāda Buddhist
Scholasticism. (Handbook of Oriental Studies 2/11). Leiden etc.: Brill.
Wright, Arthur F. (1959). Buddhism in Chinese History. Stanford CA: Stanford University
Press.
Abhidharma in China 295
Xu, Shen 許慎 ([1981] 1988). Shuo wen jie zi zhu 說文解字注. Shanghai: Shanghai guji
chubanshe.
Yu, Anthony C. (2005). State and Religion in China. Historical and Textual Perspectives.
Chicago IL and La Salle IL: Open Court.
Zacchetti, Stefano (2010). “Defining An Shigao’s 安世高 Translation Corpus: The State
of the Art in Relevant Research”, In Historical and Philological Studies of China’s
Western Regions 3. Edited by Weirong Shen. Institute for Historical and Philological
Studies of China’s Western Regions, Renmin University of China, pp. 249–270.
Zürcher, Erik (1977). “Late Han Vernacular Elements in the Earliest Buddhist Translations”,
Journal of the Chinese Language Teachers Association 12/3, pp. 177–203.
Zürcher, Erik (1991). “A New Look at the Earliest Chinese Buddhist Texts”, In From Benares
to Beijing. Essays on Buddhism and Chinese Religion. Edited by Koichi Shinohara and
Gregory Schopen. Oakville CA: Mosaic Press, pp. 277–304.
Zürcher, Erik (2007). The Buddhist Conquest of China. The Spread and Adaptation of
Buddhism in Ealry Medieval China. Third Edition with a Foreword by Stephen F. Teiser.
Leiden: Brill.
296 Teng
Chapter 9
Elsewhere I have tried to show that Kuiji 窺基 (632–682) who was recognized
historically as the successor of Xuanzang 玄奘 especially in the latter’s legacy
of the Chinese Yogācāra school, the Faxiang school (Faxiang zong 法相宗),
intended to set Chinese understanding of Buddhism aright by reorienting it
toward Indian Buddhism. I tried to demonstrate in that work that what consti-
tuted such a reorientation was not a mere promotion of a Yogācāra form of
Buddhism as has been generally understood by scholars of Chinese Buddhism.
I argued that Kuiji’s more significant potential contribution to such a reorien-
tation lay in his construction of a methodological foundation for Chinese
Buddhist intellectual practice, which can be interpreted as a response to the
much debated issue of the ‘sinification of Buddhism’. The three methodologi-
cal dimensions of Kuiji’s endeavors I emphasized are: recontextualization,
exegesis, and logic. The present study explores the ‘recontextualization’ dimen-
sion of Kuiji’s methodology.1
In the prolonged course of the spreading of Buddhism to China, Buddhist
texts and ideas were studied and interpreted divorced from their home cul-
tural, historical, and intellectual contexts. On the other hand, they were
received, digested and made use of the new contexts informed by the local
cultural and socio-political concerns. Echoing Yūki Reimon, Robert Gimello
has tried to show that in response to the socio-political context of the late sixth
and early seventh centuries, most of the leading Buddhists in that period were
concerned to relieve Buddhism of the onus of its foreignness and approached
Buddhist texts and teachings in such a context.2
3 Gimello, 1976: 95. Peter Gregory has a similar suggestion. (Gregory, 1991: 110–112). A compre-
hensive historical survey in English of the Chinese Buddhist ‘doctrinal classification’ system
(panjiao 判教 can be found in Mun’s The History of Doctrinal Classification in Chinese
Buddhism. However, Mun did not address the hermeneutical issue or exegetical nature of the
panjiao system.
4 Gimello, 1976: 111.
5 Gimello, 1976: 113. More Chinese expressions of the ‘new Buddhism’ can be found in the same
reference.
6 The necessity of a “contextual theology” has been argued for initially by Shoki Coe (Coe, 1993),
and more recently by Stephen Bevan in his Models of Contextual Theology: The Struggle for
Cultural Relevance, in which he puts forward five models of “contextual theory,” namely (1)
translation; (2) the anthropological; (3) praxis; (4) the synthetic; and (5) the transcendental
issues related to the localization of Chinese theology (Bevans, 2002). These five models pro-
vide a helpful analytical framework for thinking about various forms of encounter between
‘Buddhisms’ and ‘Chinese cultures’.
298 Teng
When I examined the teachings (of the teachers), each of them professed
exclusive orthodoxy [of his teachings]. When I tried to test them against
the [available] textual testimony, the meanings were either implicitly or
explicitly inconsistent. I am left with no one to follow,7
and in Kuiji’s:
teachings and concepts. In what follows, let me introduce briefly such func-
tions relevant to this study.
10 The use of the pejorative term ‘Hīnayāna’ here is to highlight the Chinese Buddhist sectar-
ian conception of Abhidharma Buddhism as opposed to the Mahāyāna school, a point to
be discussed in this chapter.
11 According to the Theravāda tradition, all of the seven Abhidhamma texts are delivered by
the Buddha and transmitted through Śāriputra. Sammāsambuddhassa Abhidhamma
desanā pariyosānañ ca tesaṃ bhikkhūnaṃ sattappakaraṇauggahanañ ca ekappahāren’
eva ahosi. Abhidhammo cācanāmaggo nāma Sāriputtattherappabhavo. The tradition rec-
ognized that Buddha is first an Ābhidharmika, Sammāsambuddho va paṭhamataraṃ
Ābhidhammiko (Dhammasaṅgaṇi-atthakathā, p. 17).
12 The Sarvāstivāda school acknowledges Kātyāyanīputra and others’ authorship of the
Abhidharma texts, but believes that the Buddha was the original author. “na hi
vinābhidharmopadeśena śiṣyaḥ śakto dharmān pravicetum iti | sa tu prakīrṇa ukto
bhagavatā | bhadanta Kātyāyanīputraprabhṛtibhiḥ piṇḍīkṛtya sthāpito bhadanta Dharma
trātodānavargīyakaraṇava” (Pradhan, 1967: 3). See also Dhammajoti’s Sarvāstivāda
Abhidharma on a variety of scripture genres or features that lead to the development of
the Abhidharma exposition: abhidharma kathā, vaidalya, vibhaṅga, mātṛkā, and upadeśa
(Dhammajoti, 2007: 3–9).
13 For more illustrations of the functions of the Abhidharma exposition: see Dhammajoti,
2007: 9–13.
14 Potter et al., 1996: 74. See also the introduction to this volume.
300 Teng
The Abhidharma surpasses the teachings in the Sūtra, because the vari-
ous classifications of the elements of existence are listed haphazardly in
the sūtras, while the Abhidharma gives them in their definitive catechetic
forms. Thus, the sūtras are preached from the standpoint of conventional
truth (saṃvṛtisatya) according to specific worldly circumstances, but the
Abhidharma deals with absolute truth, and is concerned with the analy-
sis of mind and matter (nāmarūpapariccheda).16
The wise should skillfully distinguish the meanings in the sūtras and not
simply explain in the manner that the words are uttered [therein]. If one
simply explains in the manner that the words are uttered, one will create
contradictions among the noble teachings and also give rise to topsy-
turvy attachment in one’s own mind.18
Before the introduction to China of the Abhidharma literature proper, near the
end of the fourth century, a number of Abhidharmic taxonomies and numeri-
cal categories [of dharmas] in meditation (chanshu 禪數) were made known to
the Chinese Buddhists through a few meditation texts translated by the
Parthian monk An Shigao 安世高 who arrived in Luoyang 洛陽 in 148 CE. The
Abhidharmic taxonomies were mentioned in An Shigao’s translation of the
Discourse on the Aggregates, Elements, and Sense-spheres (Yin chi ru jing 陰持入
經), in which were introduced basic dharma-taxonomies such as the five aggre-
gates (skandha), six sense-spheres (āyatana), eighteen elements (dhātu), and
The actual translations of these texts were probably made by Zhu Fonian 竺佛
念, who was a Chinese polyglot with little knowledge of Buddhist doctrine,
under the sponsorship and management of Dao’an.26 Dao’an himself studied
Abhidharma thought from An Shigao’s translations and was eager to under-
stand more about it. He remarked that the earlier commentaries made by Kang
Senghui 康僧會 were obscure, and took the initiative of composing another
commentary.27
Though a Prajñāpāramitā scholar, Dao’an was totally open to the newly
arrived Hīnayāna Sarvāstivādin teaching. He was keen to introduce it to the
world of Chinese Buddhism, which had been carried away by Prajñāpāramitā
scholarship.28 Thanks to Dao’an’s sponsorship, Abhidharma studies thrived for
a short period of time. There was probably some Abhidharma learning fever
even among non-Buddhist intellectuals that urged the use of a pedagogical
strategy known as “geyi 格義” by Zhu Faya 竺法雅. In Chapter Four, “Literature,”
of the New Account of Tales of the World 世說新語,29 we find an anecdote in
which “a middle-troop general by the name of Yin Hao 殷浩 took great interest
in studying Buddhism after he was ousted from his office. In general, Yin Hao
had a very good grasp of the meanings of Buddhist texts, but when he came
across shishu 事數, that is the numerical categories of dharmas, he was not
able to understand them.30
Geyi was a method of adopting numerical categories from non-Buddhist
texts as an analogy to the Buddhist numerical categories, for the purpose of
lecturing non-Buddhist intellectuals on Buddhist teachings.31 Dao’an found
Although during the time of Dao’an, Chinese Buddhists began to take an inter-
est in Abhidharma, there had been no serious study and in-depth understanding
of the Abhidharma system. Despite of the fact that a good number of
Abhidharmic texts were continuously introduced to China and translated into
Chinese, there was no sign of employing the Abhidharma method by the indig-
enous Buddhist exegetes.
Apparently Xuanzang recognized the weakness of Abhidharma studies in
Chinese Buddhism. Even just from the sheer size and variety of the Abhidharma
texts translated by Xuanzang we could not miss the importance that Xuanzang
attached to the Abhidharma. In what follows I will show how Kuiji defended
the importance of the Abhidharma and how he recontexualized Buddhist doc-
trines and concepts in the Abhidharma system.
The upadeśa36 refers to the mātṛkā that repeatedly discerns the sūtras.
All ultimate (nītārtha) teachings are called mātṛkā. This (upadeśa) is
where the World-honored One thoroughly distinguishes the characteris-
tics of the dharmas, and this is where the Buddha’s noble disciples, who
have penetrated into reality, distinguish accurately the characteristics of
the dharmas by their own realization of the truth. What is named the
mātṛka is also called Abhidharma. On account of this, after the parinirvāṇa
of the Buddha, only with respect to upadeśa of the twelve divisions of the
teaching is there no difference between its being taught by the Buddha
and by the noble disciples. This is so with the Abhidharma of the three
baskets.37
36 According to the Sarvāstivādin (e.g. the Saṃyukāgama and the *Mahāvibhāṣā, Da pipo-
sha lun 大毘婆沙論) and several Mahāyana sūtras and śāstras (e.g. Pañcaviṃśatisāhas
rikāprajñāpāramitāsūtra, Yogācārabhīmiśāstra), upadeśa is one of the twelve divisions
of Buddhist scripture (dvādaśāṅga): 1. sūtra, 2. geya, 3 vyākaraṇa, 4. gāthā, 5. udāna,
6. nidāna, 7. avadāna, 8. itivṛttaka, 9. jātaka, 10. vaipulya, 11. adbhūtadharma, 12. upadeśa.
See Hirakawa, 1963: 61–62.
37 T.45.1861: 274 c1–12: “論議經者,謂諸經典循環研覈磨呾理迦,一切了義經皆名磨
呾理迦。謂於是處,世尊自廣分別諸法體相,又於是處,諸聖弟子已見諦
306 Teng
From this passage we see that Kuiji defended the Abhidharma by highlighting
its uniqueness and superiority to other teachings. Since the Abhidharma
teaches the ultimate teaching, it could be relied upon even by the Mahāyāna
followers.
Furthermore, as there is in fact an Abhidharma text in the Mahāyāna teach-
ing, Kuiji could defend the importance of Abhidharma from the Mahāyāna
point of view. Kuiji quoted Asaṅga’s Abhidharmasamuccaya and its Vyākhyā by
Sthiramati and, in his subcommentary, entitled Zajilun shuji 雜集論述記 to
Sthiramati’s Abhidharmasamuccayavyākhyā, Kuiji pointed out the indispens-
ability of Abhidharma for the discernment of the true nature of the dharmas:38
迹,依自所證無倒分別諸法體相,此亦名為摩呾理迦,亦名阿毘達磨。由此
文故傳受釋云許佛滅後十二分中論議一分,三藏之內對法一藏通弟子說,不
許餘之十一分教,并餘二藏通弟子說。所以者何?唯此對法及論議教,弟子
所造義理與佛意無有異”.
38 This passage is found in Kuiji’s sub-commentary to Sthiramati’s Abhidharmasamucca
yavyākhyā: X48.796: 1c9–2a2: “教起所因者,瑜伽釋言:諸有情等無始時來於法實
相無知僻執,造二種業受三苦。果如來隨機為說實相,令知諸法如是非空,
諸法復由如是非有,既了諸法非有非空遠離無知疑惑,僻執隨種姓別起處中
行永滅諸障、證寂滅樂。佛涅槃後魔事糺紛,部執競興,多生有見,龍猛菩
薩採集大乘,遣相空教造中論等,究暢真宗除彼有執。聖提婆等諸大論師造
百論等,弘闡大義。由是眾生復起空見,然有菩薩懼廣文海未起聞思,預生
怯退。大聖無著具廣慧悲,集阿毗達磨經所有宗要,括瑜伽師地論一切法
門,令懼文海者初依略教易可受持、終耐多聞、能達大義、證菩提果、利樂
有情,亦兼為彼二乘種姓各依自乘,證自利果”. (Punctuation mine).
Kuiji’s Abhidharmic Recontextualization of Chinese Buddhism 307
the doctrine of emptiness in order to reject the [real existence of] phe-
nomena. He composed the Mūlamadhyamakakārikā and other treatises
advocating his thought and rejecting attachment to the view of real exis-
tence. Subsequently, Āryadeva and other treatise masters composed
treatises such as the Catuḥśataka (Bai lun 百論) and so on, and propa-
gated extensively the meaning [of their teachings]. As a result, people
came to hold fast the view of emptiness. Then bodhisattva Asaṅga who
attained the first [bodhisattva] stage and obtained the concentration of
dharma-luminosity (faguang ding 法光定) and great supernatural pow-
ers, invited the Noble Maitreya to teach the Treatise of the Yogācārabhūmi
to reveal the teaching of the “middle path” that eradicates attachments to
the views of emptiness and of real existence.
However, there are bodhisattvas who were intimidated by the vastness
of this text and withdrew from learning it. Endowed with great wisdom
and compassion, the Noble Asaṅga composed a compendium on the
“Abhidharmasūtra” that includes all teachings in the Treatise of the
Yogācārabhūmi.
In this passage Kuiji pointed out that the teaching of the Abhidharma was
originally intended by the Buddha for discerning the true nature of the dhar-
mas. And it is only by a clear discernment of the true nature of the dharmas
that the eventual elimination of suffering is possible. Kuiji also pointed out the
potential danger of attachment to the view of emptiness.
The more important contextualization that Kuiji effected, in my view, is
doctrinal contextualization in the Abhidharmic systematization. Abhidharma
is the foundational analytic tool, with a complex system of categorization of
sight into the external world we experience, as well as into ourselves as the
experiencing subject. It is granted that different Abhidharmic systems would
likely present the external world and internal self differently, but at least an
effective and grounded dharma discussion and argument can be possible.
related to each other and mutually subsumed. The following is Kuiji’s mapping
of the five groups of the one hundred dharmas (wu wei bai fa 五位百法) taxon-
omy against the taxonomies of the five aggregates (skandha), six sense-spheres
(āyatana), and eighteen elements (dhātu):39
於色蘊中攝十一種色,受蘊攝遍行中受數,想蘊攝遍行中想數,行蘊
攝相應心所四十九法,及攝不相應行二十四法,合七十三法,識蘊唯
攝心法八種,總攝九十四法,唯除六無為。故俱舍云:蘊不攝無為義
不相應故。
39 T.45.1861: 334b25-c7.
40 In the Theravāda Abhidhamma system, there are 51 thought-concomitants, but two
thought-concomitants, sensation and perception, are excluded in this context, because
they belong to vedanā- and saṃjñāskandha respectively, instead of to saṃskāraskandha.
41 In the Sarvāstivāda Abhidharma system, there are eleven types of matter listed in the
‘five-group-taxonomy,’ of which matter associated with thought objects (dharmāyatanikāni
rūpāni) is not subsumed under the rūpāyatana of the ‘twelve-āyatana-taxonomy’.
Kuiji’s Abhidharmic Recontextualization of Chinese Buddhism 309
十二處攝者,內五處外五處攝十種色,意處攝八識,法處攝四類法。
所謂色法法處所攝色有五種。是相應法有五十一,不相應法二十四,
無為法有六。合有八十六法。
十八界攝者,唯開意處立六識界,餘竝與處同
五位分別者,極樂者心所收。世界者即色法攝。三科收攝者,極樂者
即行蘊收。世界者即色蘊攝。十二處中極樂法處收。世界即色處攝。
十八界中法界及六境中攝。
42 Kuiji’s equation of mana-āyatana with the six thought-dhātus is a mistake, because the
mana-āyatana should include manodhātu in addition to the six thought-dhātus.
43 T.37.1758: 338a6–10.
310 Teng
Question: the Maitreya Inquiry Sūtra says: “Chanting the name of the
Buddha by an ordinary person whose mind is still fettered and defiled
will not lead to the rebirth in the Pure Land. If this is so, how do people
who are still at the stage of practicing and are yet to attain the noble
stage, ever strive to be reborn in the ‘Pure Land,’ since their mind is yet to
be purified?”44
Kuiji replied: “At the moment of chanting the Buddha’s name, the mind
that is operative is the sixth consciousness. It has this characteristic that
when it is taking the chanting of Buddha’s name as its mental object, it is
only associated (saṃprayukta) with the five universal thought-concomi-
tants bianxing wu suo 遍行五所 (sarvatraga-caitta) and the eleven
wholesome thought-concomitants shan shiyi suo 善十一所 (kuśala-
caitta). This thought of chanting Buddha’s name will not activate the
defiled thought-concomitants because there is no other mental object to
serve as the condition for their arising.”45
It should be clear now how such profound puzzles could be resolved system-
atically with the Abhidharma methodology. First of all, Kuiji applied the
method of categorizing (saṃgraha) to identify the locus of the thought of
chanting Buddha’s name. Then he applied the theory of association (sam-
prayoga) to determine what types of thought-concomitants are associated
with the thought of chanting Buddha’s name.
6 Conclusion
Abbreviation
Bibliography
Bevans, Stephen B (2002). Models of Contextual Theology. Revised and Expanded Edition.
Maryknoll NY: Orbis Books.
Coe, Shoki (1993). Recollections and Reflections. New York: Formosan Christians for
Self-Determination.
Cox, Collett (1995). Disputed dharmas. Early Buddhist theories on existence: an annotated
translation of the section of factors dissociated from thought from Saṅghabhadra’s
Nyāyānusāra. (Studia philologica Buddhica, Monograph Series 11). Tokyo: The
International Institute for Buddhist Studies.
Demiéville, Paul (1951). “La Yogācārabhūmi de Saṅgharakṣa”, Bulletin de l’École française
d’Extrême-Orient 44, pp. 339–436.
Dhammajoti Bhikkhu KL (2007). Sarvāstivāda Abhidharma. Hong Kong: University of
Hong Kong.
Enomoto, Fumio (1996). “A Sanskrit Fragment from the Vibhāṣā Discovered in Eastern
Turkestan”, In Sanskrit-Texte aus dem buddhistischen Kanon: Neuentdeckungen und
Neueditionen III. Edited by Fumio Enomoto, Jens-Uwe Hartmann, and Hishashi
Matsumura. Göttingen: Vandenhoeck & Ruprecht, pp. 135–143.
312 Teng
Gimello, Robert (1976). “Chih-Yen (智儼 602–668) and the Foundation of Hua-Yen
Buddhism”. Unpublished PhD dissertation, Columbia University.
Gregory, Peter N (1991). Tsung-mi and the sinification of Buddhism. Princeton NJ:
Princeton University Press.
Hayashima, Osamu, et al. (2003). Abhidharmasamuccaya-bhāṣya & vyākhyā. Sanskrit,
Tibetan, and Chinese texts Compared. <http://www.shiga-med.ac.jp/public/yugagyo/
AS/AS_ETEXT_V3_ALL.pdf> (accessed 08/27/2014).
Hirakawa, Akira (1963). The Rise of Mahāyāna Buddhism and its Realtionship to the
Worship of Stūpas. (Memoirs of the Research Department of the Toyo Bunko 22).
Tokyo: The Toyo Bunko.
La Vallée Poussin, Louis de (1930). “Documents d’Abhidharma”, Mélanges chinois et boud-
dhiques I, pp. 65–121.
Lopez, Donald S. (1988). Buddhist Hermeneutics. (Studies in East Asian Buddhism 6).
Honolulu HI: University of Hawaii Press.
Lusthaus, Dan (2002). Buddhist phenomenology: a philosophical investigation of Yogacara
Buddhism and the Ch’eng Wei-shih lun. (Curzon critical studies in Buddhism series).
London, New York NY: RoutledgeCurzon.
Müller, Edward (ed.) (1897). Dhammasaṅgaṇi-atthakathā (Atthasālilī). London: Pāli Text
Society.
Potter, Karl H.; Buswell, Robert E.; and Jaini, Padmanabh S. (1996). “The Development
of Abhidharma Philosophy”, In Encyclopedia of Indian Philosophie. Vol. 7: Abhidharma
Buddhism to 150 ad. Delhi: Motilal Banarsidass, pp. 73–119.
Pradhan, Prahallad (ed.) (1975). Abhidharmakośabhāṣyam of Vasubandhu. (Tibetan
Sanskrit Works Series 8). Patna: K.P. Jayaswal Research Institute (rev. 2nd ed.).
Takakusu, Junjirō 高楠順次郎, Watanabe, Kaigyoku 渡邊海旭, and Ono Gemmyō 小野
玄妙 (ed.) (1924–1934). Taishō shinshū daizōkyō 大正新脩大藏經, Tokyo: Taishō
issaikyō kankōkai.
- T.27.1545: 500 arhats, [Abhidharma]mahāvibhāṣā[śāstra], Apidamo da piposha lun
阿毘達磨大毘婆沙論, Xuanzang 玄奘.
- T.37.1758: Kuiji 窺基, Amituo jing tongzan shu 阿彌陀經通贊疏.
- T.44.1851: Huiyuan 慧遠, Dasheng yizhang 大乘義章.
- T.45.1861: Kuiji 窺基, Dasheng fayuan yilin zhang 大乘法苑義林章.
- T.47.1964: Kuiji 窺基, Xifang yaojue shiyi tonggui 西方要決釋疑通規.
- T.48.2023: Rujin 如卺, Zimeng jingxun 緇門警訓.
- T.50.2053: Huili 慧立 and Yancong 彥琮, Da Tang Daci’en si sanzang fashi zhuan 大
唐大慈恩寺三藏法師傳.
- T.52.2102: Sengyou 僧祐, Hongming ji 弘明集.
- T.55.2145: Sengyou 僧祐, Chu sanzang ji ji 出三藏記集.
Kuiji’s Abhidharmic Recontextualization of Chinese Buddhism 313
Teng, Weijen (2011). Recontextualization, Exegesis, and Logic: Kuiji’s (632–682) Metho
dological Restructuring of Chinese Buddhism. Unpublished PhD dissertation, Harvard
University.
Tsukamoto, Zenryū (1985). A History of Early Chinese Buddhism. From its Introduction
to the Death of Hui-yüan. 2 vols. Tokyo, New York, San Francisco: Kodansha
International. (English translation of the original Japanese work by Leon Hurvitz.
Original publication: Chūgoku Bukkyō tsūshi. Tokyo: Shunjūsha, 1979).
Willemen, Charles (1975). The Essence of Methaphysics. Abhidharmahṛdaya. Translated
and annotated. (Série Etudes et Textes 4). Bruxelles: Institut Belge des Hautes Études
Bouddhiques.
Zürcher, Erik (2007). The Buddhist Conquest of China. The Spread and Adaptation of
Buddhism in Ealry Medieval China. Third Edition with a Foreword by Stephen F. Teiser.
Leiden: Brill.
314 Chapter 10
Chapter 10
* I would like to thank the late Lance S. Cousins and Rupert Gethin for their comments which
helped me to improve my paper in various ways.
1 Cousins, 1973: 115.
2 On gNubs-chen Sangs-rgyas ye-shes’ dates and the time of composition of the bSam-gtan
mig-sgron, see Esler, 2014: 5–27.
3 gNubs-chen Sangs-rgyas ye-shes, rNal-’byor mig-gi bsam-gtan or bSam-gtan mig-sgron: A trea-
tise on bhāvanā and dhyāna and the relationships between the various approaches to Buddhist
contemplative practice, Reproduced from a manuscript made presumably from an Eastern
Tibetan print by ’Khor-gdong gter-sprul ’Chi-med rig-’dzin, Smanrtsis shesrig spendzod, vol.74
(Leh: Tashigangpa 1974). Henceforth abbreviated as C.
4 de La Vallée Poussin, 1971, vol.1, ch.1: 4.
5 Monier-Williams, 2001: 61.
6 Conze, 1973: 63.
fact that they held an accurate knowledge of the phenomena comprising real-
ity to be the necessary condition to effectively remove the contaminations
(kleśa).25
Turning to the second error, the tendency in the West to reduce metaphysics
to a mere branch of philosophy has led to the widespread opinion that it is
little more than mental activity concerned with abstractions. More often than
not, the rejection of metaphysics betrays a lack of understanding concerning
both its nature and scope.26 Traditionally speaking, metaphysics is the science
which deals with the nature of reality, and hence which enables man to discern
the real from the illusory. It is, moreover, integrally wed to methods enabling
the spiritual realization of this knowledge and, as scientia sacra, lies at the
heart of all the other traditional sciences.27 In the Buddhist case, the close con-
nection between metaphysics and meditative praxis is illustrated by the fact
that the taxonomic enterprise of the Abhidharma probably had its origin in
certain early Buddhist techniques of meditative introspection, where the prac-
titioner deconstructed through analysis his everyday experience, and labeled
and identified its individual components.28
While the ontological difference, i.e. the distinction between the ontic (the
concern with beings and with existence) on the one hand, and the ontological
(the concern with Being) on the other, articulated by Martin Heidegger29 is
crucial philosophically, it fails metaphysically to take into account that which
lies beyond Being and transcends manifestation.30 This failure means that it
does not go beyond the domain of philosophy. What distinguishes philosophy
from metaphysics is precisely its systematic character, which makes all philo-
sophical systems, whatever their orientation, of little value metaphysically,
because they are closed and seek to analyze a particular aspect of reality, all
the while excluding those aspects that do not conform to their limited point of
view. While it may be legitimate to investigate a particular possibility (which is
what amounts to a science), it is certainly illegitimate to affirm that this is the
whole of reality and to deny all that goes beyond the narrow confines of this
limited understanding.31
waves to a blind person who has never seen the sea.37 Having laid these meta-
physical premises, we may now proceed to examine the individual traces of
Abhidharma in the bSam-gtan mig-sgron, which we have classified under the
rubrics of cosmological, phenomenological and meditational categories.
2 Cosmological Categories
41 This catalogue of dreams, itself derived from the Ratnakūṭasūtra, has been studied in:
Esler, 2012a: 317–328.
42 C 102.1–2; see rMi-lam bstan-pa, in dKon-mchog brtsegs-pa’i mdo, in KD, vol.39, 412.1,
449.2–4.
43 Csoma de Koros, 1980: 45.
44 C 63.
45 de La Vallée Poussin, 1971, vol.2, ch.3: 141–142.
Traces of Abhidharma in the bSam-gtan mig-sgron 321
hypothetical situation to that which exists for the humans of the southern con-
tinent of Jambudvīpa, “who are unable to show the path to the ghastly people of
the northern continent of Uttarakuru”.46 The presumable reason the people of
Uttarakuru are here termed ‘ghastly’ is found in the Abhidharmakośa, where it
is explained that these beings cannot detach themselves from the desire realm
(kāmadhātu) and so are unable to enter contemplation. Furthermore, they are
also incapable of taking up the vows of individual release (pratimokṣa).47
46 C 209.6–210.1: […] sangs rgyas sems can tha dad na ’brel ba med pas snang du mi btub ste/
lho’i ’dzam bu gling gi mi sgam pos/ byang gi sgra mi snyan gyi mi ma legs pa’i lam bstan du
mi btub la/ […].
47 de La Vallée Poussin, 1971, vol.2, ch.3: 183; vol.3, ch.4: 104.
48 For instance, Brahmā’s repeated pleas addressed to the Buddha in the weeks following his
enlightenment finally convince him to teach his doctrine. See Lamotte, 1947: 52, 54.
49 Note for example Śakra’s role as interlocutor of the Buddha in these sūtras. See Conze,
1994: ch.3: 106; ch.31: 283, 296.
50 C 93.5–94.3; see rMi-lam bstan-pa, in dKon-mchog brtsegs-pa’i mdo, in KD, vol.39, 410.5–6,
428.5–429.4.
51 C 104.4–6; see rMi-lam bstan-pa, in dKon-mchog brtsegs-pa’i mdo, in KD, vol.39, 412.3,
454.6–455.3.
322 Esler
approach (§2.2.4.1.4.2.1).52 These defects cause one to fall into the first to fourth
levels of the gods of the desire realm. We will now turn to the Abhidharmakośa
to see how it classifies these different categories of divine beings.
There are six gods of the desire realm: the Cāturmahārājakāyikas, the
Trāyastriṃśas, the Yāmas, the Tuṣitas, the Nirmānaratis and the Paranirmita
vaśavartins. Like their human counterparts, the gods of the first four levels
enjoy the desirable sense-qualities that are at their disposal. It is only the gods
of the fifth and sixth levels of the desire realm who can emanate the objects
of their desire (kāma).53 The head of the Paranirmitavaśavartins is Māra, who
rules the desire realm and constantly seeks to oppose the Buddha.54
The four great kings (cāturmahārājakāyikas) are the rulers of the first and
lowest of the six classes of gods of the desire realm; they are Ḍhṛtarāṣṭra, king
of the east and lord of the scent-eaters (gandharva); Virūḍhaka, king of the
south and lord of the kumbhaṇḍa demons, who are pictured with huge stom-
achs and genitals; Virupākṣa, king of the west and lord of the serpent-spirits
(nāga); and Vaiśaravaṇa, king of the north and lord of the goblins (yakṣa). As
protectors of the world (lokapāla), their function is to examine the religious
conduct of human beings and report back to the Trāyastriṃśa gods.55
The Trāyastriṃśa heavens represent the second among the six classes of gods
of the desire realm and are located in the upper half of Mount Meru.56 They
are in a sense intimately associated with the Abhidharma, since it is believed
that the Buddha first taught the Abhidharma to his deceased mother who had
taken birth in the Trāyastriṃśa heavens. It is only later that Śāriputra received
these teachings and passed them on to his disciples.57 Śakra is an epithet of
Indra,58 who, in Buddhist cosmology, is the lord of the gods of the Trāyastriṃśa
heavens and of the lower levels of the desire realm.59
Tuṣita is the fourth among the six classes of gods of the desire realm and
is located in the space above Mount Meru; it is here that a bodhisattva of the
tenth stage dwells while waiting to take his final birth as a Buddha.60
To gain an impression of the decreasing materiality of these different gods, we
may recall that the Abhidharmakośa informs us that the Cāturmahārājakāyikas,
52 C 233.2–5.
53 de La Vallée Poussin, 1971, vol.2, ch.3: 164, 166.
54 Lamotte, 1958: 761.
55 Lamotte, 1958: 759–760.
56 de La Vallée Poussin, 1971, vol.2, ch.3: 161.
57 Lamotte, 1958: 200.
58 Cf. Amarasiṃha, 2008, ch.1, svargavarga, v.42: 23.
59 Lamotte, 1958: 760.
60 Lamotte, 1958: 761.
Traces of Abhidharma in the bSam-gtan mig-sgron 323
along with the next category, the gods of the Trāyastriṃśa heavens, abide
on the earth and engage in sexual union just like their human counterparts.
However, they appease their desire by emitting wind, since they do not
carry a seminal substance. The Yāmas appease the fire of desire by embrac-
ing, the Tuṣitas by holding hands, the Nirmāṇaratis by laughing, and the
Paranirmitavaśavartins by simply looking at each other. Furthermore, the gods
above the Trāyastriṃśa heavens (viz. the Yāmas, the Tuṣitas, the Nirmāṇaratis,
the Paranirmitavaśavartins, and the gods of the form realm) dwell in seraphic
mansions (vimāna; Tib. gzhal-yas-khang).61
The Brahmās inhabit the Brahmakāyika heavens, the first among the three
abodes of the first contemplation (bsam-gtan dang-po’i gnas-gsum) of the
form realm (rūpāvacara), which will be examined below. The form realm
(rūpadhātu), located above the desire realm, consists of four worlds, corre-
sponding to the first contemplation (two or three levels), the second and third
contemplations (three levels each) and the fourth contemplation (eight lev-
els). The different levels of the form realm are superposed as terraces.62 The
beings in the Brahmakāyika heavens have cleaned away their sins and dwell
on the four abodes of Brahmā (brahmavihāra); these are synonyms of the four
immeasurables when taught within the context of the Hīnayāna. The mea-
sure of the merit of the Brahmakāyika gods is that they abide blissfully in this
heaven for an aeon (kalpa).63
In these heavens is found a high abode inhabited by a single sovereign,
Mahābrahmā, which corresponds to the third abode of the first contempla-
tion. The gods inhabiting the Brahmakāyika heavens are characterized by the
diversity of their bodies on the one hand, and by the uniformity of their per-
ceptions on the other. For example, although there may be differences in their
height, etc., if Brahmā thinks that the other gods are created by him, his aco-
lytes too will think that they are created by Brahmā.64
The ethereal classes (gtsang-rigs), also called ethereal levels (śuddhāvāsa
kāyika; Tib. gtsang-ris) or ethereal abodes (śuddhāvāsa; Tib. gnas gtsang-ma),
which are five in number, are located just under Akaniṣṭha, which is at the
peak of the form realm (rūpadhātu; Tib. gzugs-khams) and corresponds to
the fourth contemplation. Only those having perfected the path of seeing
(darśanamārga; Tib. mthong-lam) can be born there. The gods of the ethereal
classes are unable of going to the formless realm (ārūpyadhātu; Tib. gzugs-
med-kyi khams) or of falling to the lower realms.65
68 For the six causes, see de La Vallée Poussin, 1971, vol.1, ch.2: 245–271.
69 For the four conditions, see de La Vallée Poussin, 1971, vol.1, ch.2: 299–309; and Kalupah-
ana, 1994: 148–149.
326 Esler
focus on their relations, failing which they would have been left with lists of
disconnected phenomena.70 Incidentally, it might be noted that Nāgārjuna
himself, despite his deconstruction of Abhidharma phenomenological catego-
ries, had no objection to this formulation of conditional relations, as long as
the links in the causal chain were not taken to have an intrinsic nature
(svabhāva).71
the time when no more denizens of hell are born up to the destruction of the
universe; (3) the aeon of formation (vivartakalpa; Tib. chags-pa’i bskal-pa),
which lasts from the primordial wind (prāgvāyu) up to the time when are born
denizens of hell; and (4) the great aeon (mahākalpa; Tib. bskal-pa chen-po),
defined as the period of one universe, lasting from its inception to its
destruction.77
Several allusions to aeons are found in the bSam-gtan mig-sgron, again in
the context of defining the time it takes to accomplish the fruition. We find a
reference to the so-called incalculable aeon (asaṃkhyeyakalpa) in the same
passage in Chapter VI just mentioned, which gives the various possible dura-
tions required to attain the fruition on the Hīnayāna.78 In the same chapter,
a passing allusion to the aeon is also made in the section describing the
Mahāyoga’s relative superiority to Chan: the latter approach is here depreci-
ated for taking aeons in order to reach the eleventh bodhisattva stage, the stage
of universal light (§5.1.11).79 We have yet another reference in a quotation from
the Saṃdhinirmocanasūtra cited in Chapter V (§4):80
Mind is the essence of the religious path. The body is the vessel for the
religious path. The virtuous friend originates due to conditions. If the
mind is comfortable, objects too are comfortable. If one is without accep-
tance and rejection towards the two limitations, one views the eighteen
sense-elements.
The idea expressed here is that the faculty of understanding (prajñā) permits
one to pierce through the illusory nature of relative truth (“acceptance and
rejection”) to reveal that what we habitually take to be concrete entities are in
fact but the assemblage of the psycho-physical sense-elements. Given the
Chan orientation of this quotation, we can take the eighteen sense-elements
mentioned in this passage as shorthand for ‘emptiness’ (śūnyatā), the ultimate
nature of phenomena.
In Chapter VI, we find several references to the aggregates, sense-elements
and sense-spheres taken together to signify that they are in fact the mind’s
luminous clarity arising as gods and goddesses. We find such statements in the
sections concerning the tantric views of the inseparability of means and
understanding (§1.3)87 and of non-duality (§1.5),88 as well as in the section on
the tantric approach’s fruition (§4.1).89 Later in Chapter VII, this notion is criti-
cized from the perspective of Atiyoga for still being contrived (§2.3.4.11).90
Let us here take a look at the Abhidharma’s classification of the eighteen
sense-elements and twelve sense-spheres. The eighteen sense-elements (dhātu;
Tib. khams) are divided into three groups of six. (I) The first group consists of
the faculties: (1) the sense-element of the visual faculty (cakṣurindriyadhātu),
(2) the sense-element of the auditory faculty (śrotrendriyadhātu), (3) the
sense-element of the olfactory faculty (ghrāṇendriyadhātu), (4) the sense-
element of the gustatory faculty (jihvendriyadhātu), (5) the sense-element of
the tactile faculty (kāyendriyadhātu) and (6) the sense-element of the psychic
faculty (manendriyadhātu). (II) The second group consists of the objects: (7)
the sense-element of form (rūpadhātu; Tib. gzugs-khams), (8) the sense-ele-
ment of sound (śabdadhātu), (9) the sense-element of odor (gandhadhātu),
(10) the sense-element of taste (rasadhātu), (11) the sense-element of tangi-
bles (spraṣṭavyadhātu) and (12) the sense-element of [psychic] phenomena
(dharmadhātu; Tib. chos-khams). (III) The third group consists of the conscious-
nesses: (13) the sense-element of visual consciousness (cakṣurvijñānadhātu),
(14) the sense-element of auditory consciousness (śrotravijñānadhātu), (15)
the sense-element of olfactory consciousness (ghrāṇavijñānadhātu), (16)
the sense-element of gustatory consciousness (jihvāvijñānadhātu), (17) the
sense-element of tactile consciousness (kāyavijñānadhātu) and (18) the
sense-element of psychic consciousness (manovijñānadhātu). According to
Vasubandhu’s Abhidharmakośa, groups I (i.e. sense-elements 1–6) and III (i.e.
87 C 199.
88 C 205.
89 C 268.
90 C 435.
330 Esler
4 Meditational Categories
94 C 77.2–6.
95 Miyazaki, 2004: 899.
96 Kamalaśīla (Gyaltsen Namdrol, 1997: 156–158 (Tibetan text), 259 (Sanskrit text)).
97 Anacker, 2008: 247 (English translation), 447 (Sanskrit text).
98 Guenther, 1989: 76–77.
99 Anacker, 2008: 247.
332 Esler
When one attains the qualities of having completed calm abiding, [medi-
tation] is given the name ‘contemplation’. By attaining the other special
qualities of imaging, [meditation] is given the names ‘contemplation,’
‘[formless absorption of the] form[less] realm,’ and ‘release in the form-
less realm’.
It is apparent from this last sentence that, as mentioned above, each of the four
contemplations can be either an absorption (samāpatti) or a state of existence
(upapatti). According to Vasubandhu, the former is contemplation-as-cause
(kāraṇadhyāna), the latter contemplation-as-effect (kāryadhyāna).101 As is
explicitly stated in the passage below, it is only when one is free from the thirst
of the desire realm (kāmadhātu) that one can enter these four contemplations,
which pertain to the form realm (rūpadhātu). gNubs-chen Sangs-rgyas ye-shes’
description proceeds as follows:102
100 C 77.6–78.2: zhi gnas rdzogs pa’i yon tan thob pas bsam gtan gyi ming thob pas ni/ de la
dmigs pa’i yon tan khyad par can gzhan thob pas/ bsam gtan dang/ gzugs [med pa’i snyoms
par ’jug pa] dang/ gzugs med pa’i rnam par thar pa’i ming thob bo/. The words in square
brackets have been added according to the corresponding passage in Kamalaśīla (Gyalt-
sen Namdrol, 1997: 45 (Tibetan text), 214 (Sanskrit text)).
101 de La Vallée Poussin, 1971, vol.5, ch.8: 128.
102 C 78.2–6: de ltar gang gi tshe tshor ba btang snyoms dang ldan zhing rtog dpyod bcas par
ldan tsam na/ mi lcogs pa med pa’i brtan pa dang sbyor ba’i sems so/ /yang ’dod pa’i sred pa
dang bral [bral em : dbral C]/ dga’ ba bde [bde em. : de C] dang ldan [ldan em. : tsan C] par
gyur tsam na/ bsam gtan dang po zhes bya’o/ /bsam gtan dang po de la rtog pa tsam med
pa ni/ bsam gtan khyad par can no/ /yang rtog dpyod dang bral bas bsam gtan dang po sred
[sred em. : brjod C] pa dang bral ba de/ dga’ ba bde nang [nang em. : na C] yongs su dang
ba dang ldan pa de bsam gtan gnyis pa’o/ /de’i sred pa dang bral ba btang snyoms shes bzhin
dang ldan na bsam gtan gsum pa’o/ /yang de’i sred pa dang bral te sdug bsngal yang ma
yin/ bde ba yang ma yin pa’i btang snyoms dang dran pa de bsam gtan bzhi pa’o/.
Traces of Abhidharma in the bSam-gtan mig-sgron 333
These two passages follow quite closely the wording found in Kamalaśīla’s first
Bhāvanākrama, so much so that they appear to be a summary or paraphrase
thereof. Here is a translation of the section in the Bhāvanākrama:103
103 Kamalaśīla (Gyaltsen Namdrol, 1997: 44–46 (Tibetan text), 214–215 (Sanskrit text)); Van
den Broeck, 1977: 27–28 (French translation). Here is the Tibetan with the parallel words
in Roman type: /sems rtse gcig pa de phyi phyir las su rung ba nyid dang ldan zhing dmigs
pa la sogs pa’i yon tan gyi khyad par thob pas bsam gtan dang/ gzugs med pa’i snyoms par
’jug pa dang/ rnam par thar pa la sogs pa’i ming thob bo/ /’di ltar gang gi tshe btang snyoms
kyi tshor ba dang ldan zhing rtog pa dang bcas pa dang/ dpyod pa dang bcas par gyur pa
de’i tshe mi lcogs pa med pa zhes bya ste/ bsam gtan dang po’i sbyor ba’i sems so/ /gang gi
tshe ’dod pa’i sred pa dang/ sdig pa’i chos rnams dang bral zhing rtog pa dang/ dpyod pa
dang/ dga’ ba dang/ bde ba dang ldan par gyur pa de’i tshe bsam gtan dang po zhes bya’o/
/bsam gtan dang po de nyid las rtog pa tsam med pa de ni bsam gtan khyad par can zhes
bya’o/ /gang gi tshe rtog pa dang dpyod pa dang bral te bsam gtan dang po’i sa’i sred pa
dang bral bar gyur nas dga’ ba dang bde ba dang/ nang yongs su dang ba dag dang ldan
par gyur pa de’i tshe bsam gtan gnyis pa zhes bya’o/ /gang gi tshe bsam gtan gnyis pa’i
sa’i sred pa dang bral bar gyur te bde ba dang/ btang snyoms dang/ dran pa dang/ shes
bzhin dag dang ldan par gyur pa de’i tshe bsam gtan gsum pa zhes bya’o/ /gang gi tshe
bsam gtan gsum pa’i sa’i sred pa dang bral te sdug bsngal yang ma yin bde ba yang ma yin
zhing/ btang snyoms dang/ dran pa dag dang ldan par gyur pa de’i tshe bsam gtan bzhi pa
zhes bya’o/. The Sanskrit reads: eṣā ca cittaikāgratā uttarottarakarmaṇyatāsampra-yogād
ālambanādiguṇaviśeṣayogācca dhyānārūpisamāpattiḥ vimokṣādivyapadeśaṃ labhate/
tathā hi yadopekṣāvedanāsamprayuktā savitarkasavicārā sā bhavati, tadā’nāgamyā
ucyate [prathamadhyāna-prayogacittatvāt]/ yadā ca kāmatṛṣṇayā [pāpadharmaiśca]
viviktā bhavati, [vitarkavicāra] prītisukhādhyātma-samprasādaiḥ samprayuktā
bhavati, tadā prathamaṃ dhyānam ucyate/ ata eva prathamadhyānaṃ vitarkamātra-
rahitaṃ dhyānāntaramucyate/ yadā vitarkavicārarahitā prathamadhyānabhūmitṛṣṇayā
viviktā ca bhavati, prītisukhādhyātmasamprasādaiḥ samprayuktā bhavati, tadā dvitīyaṃ
dhyānamucyate/ yadā tu dvitīyadhyānabhūmitṛṣṇayā viviktā bhavati, sukhopekṣā-
smṛtisamprajanyasamprayuktā bhavati, tadā tṛtīyaṃ dhyānam ucyate/ yadā
334 Esler
105 de La Vallée Poussin, 1971, vol.5, ch.8: 147–149, 147, note 1; the information given in [square
brackets] is supplied by de La Vallée Poussin in his note to this passage.
106 Rahula, 1971: 111.
107 de La Vallée Poussin, 1971, vol.1, ch.2: 173.
108 Kalupahana, 1994: 35.
109 Cousins, 1973: 122.
110 de La Vallée Poussin, 1971, vol.1, ch.2: 175, note 2.
111 Horner, 1969: 86.
112 Lamotte, vol.3, 1970: 1488.
336 Esler
Visuddhimagga’s careful distinction between five types of joy: (1) minor (P.
khuddikā) joy makes the body’s hairs stand on end; (2) momentary (P. khaṇikā)
joy is comparable to sudden flashes of lightning that occur from time to time;
(3) descending (P. okkantikā) joy descends upon the body like waves hitting the
seashore; (4) buoyant joy (P. ubbegā) is sufficiently powerful to cause the body
to jump in the sky; and (5) pervading (P. pharaṇā) joy pervades the entire body
just as a torrent of water would inundate a grotto.113 With regard to transport-
ing joy, Lance S. Cousins notes that this form of levitation is particularly linked
to strong forms of religious devotion.114 The emotional feeling-tone of the term
is also evoked by the fact that in common non-technical Indian usage, prīti
signifies “the joyful and blissful aspects of family and marital love”.115
As far as bliss (sukha) is concerned, it is differentiated from the relative
excitement of joy in that it is a calm and steady state of well-being. Furthermore,
the bliss pertaining to the first two contemplations must be distinguished from
the bliss of the third contemplation, in that in the former case it is the bliss of
veritable refinement (praśrabdhisukha), whereas in the latter case it is a feel-
ing of bliss (sukhā vedanā). This feeling of bliss is, in the case of the third
contemplation, purely mental in that it does not rely on any of the other sense-
faculties. Such a purely mental bliss is beyond the desire realm.116
As for recollection (smṛti), in the context of the contemplations it is defined
as the ability not to forget the sign (nimitta, here meaning motive or reason) of
equanimity.117
As has already been pointed out, equanimity is of two kinds: on the one
hand, equanimity of conditionings (saṃskāropekṣā) is characteristic of the
third contemplation, where joy is still present. On the other hand, equanimity
of feelings (vedanopekṣā), which is neither pleasure nor suffering, is peculiar to
the fourth contemplation.118
To summarize in a less technical language, what we have then is a pro-
cess of gradual simplification as the adept progresses along the various
contemplations: first, examination and scrutiny are abandoned in the second
contemplation; then, joy is abandoned in the third contemplation; and finally,
145 C 118.1–2.
146 de La Vallée Poussin, 1971, vol.1, ch.2: 279.
147 Griffiths, 1999: 22–23.
148 C 211, quoting rDo-rje bkod-pa, in NGM, vol.15/ba, 358.3–5: dbus mtha’ mi dmigs mnyam pa’i
lam/ /gcig tu [tu C: pu NGM] mnyam par chud nas ni/ /yul med tshol ba’i sems bral te/ /
blang dor gnyis kyi mtha’ [mtha’ NGM: mnya’ C] spangs nas/ /gnas pa med pa’i tshul gyi [gyi
C: gyis NGM] gnas/ /chos nyid dbyings su ro gcig phyir/ /shes dang shes bya gnyis su med
[gnyis su med C: gnyis med do NGM]/ /ye nas shes shing [shing C: nyid NGM] chos shes na/
/chos nyid gzhan nas bcos mi dgos [bcos mi dgos C: btsal ci dgos NGM]/.
342 Esler
149 Lamotte, 1958: 681–682; Cornu, 2006: 709; and Negi, 2006: 11.
Traces of Abhidharma in the bSam-gtan mig-sgron 343
(darśanamārga) per se. The sixteenth moment, cognizing the principle in its
subsequent realization with regard to the path (mārge ’nvayadharmajñāna),
belongs as such to the path of cultivation (bhāvanāmārga), since here there is
no longer anything unseen to be seen, but merely meditation on the truth as it
has been seen.150
4.6 Transcendence
Although the bSam-gtan mig-sgron is written from a tantric, not to say rDzogs-
chen perspective, there are several references to the classical Abhidharma
formulations of the path’s fruition. In Chapter VI, which is devoted to the tant-
ric approach of Mahāyoga, in the passage cited above concerning the time
required to accomplish the fruition in the tantric vehicle, a comparison is
drawn with the followers of the Hīnayāna, who hold that its maturation can be
felt in this life (dṛṣṭadharmavedanīya), in the next (upapadyavedanīya), after
any number of rebirths (aparaparyāyavedanīya), or be of uncertain ripening
(aniyatavipāka) (§4.4).151
It will be noted that according to the Abhidharmakośa, a deed (karma) may
be either of uncertain ripening (aniyatavipāka) – i.e. whose ripening may
or may not be felt – or of certain ripening (niyatavipāka). In the latter case,
there are three possibilities: (1) it may be felt as a phenomenon seen [in this
life] (dṛṣṭadharmavedanīya); (2) felt in the next life (upapadyavedanīya); or
(3) after any number of rebirths (aparaparyāyavedanīya).152 Furthermore,
in the Abhidharmakośa, the transcendence obtained visibly in this life
(dṛṣtadharmanirvāṇaprāpta) is defined as the transcendence with a remain-
der of the aggregates (sopadhiśeṣanirvāṇa; Tib. phung-po dang-bcas-pa’i
mya-ngan-las ’das-pa); this is the transcendence which an arhat obtains in
this life, while still living out the karmas which make up his physical body.
It is contrasted with the transcendence without remainder of the aggregates
(nirupadhiśeṣanirvāṇa; Tib. phung-po lhag-ma med-pa’i mya-ngan-las ’das-pa),
which is obtained after death.153 While the transcendence with a remainder of
the aggregates is not explicitly mentioned in our text, gNubs-chen makes sev-
eral allusions to the transcendence without remainder of the aggregates and
There are two manners of transcendence that are found in the oral
injunctions. In [the case of] the transcendence without remainder of the
aggregates, it is doubtless that the yogin [will attain it] as soon as he is
free from the knotted net of the body.
5 Concluding Remarks
154 C 278.4–5: mya ngan las ’das pa’i tshul gnyis ni bka’ nyid las bzhugs pas/ phung po lhag ma
med pa mya ngan las ’das na’ang/ […].
155 There are four types of awareness-holder counted in Mahāyoga, among which this is the
second. The four are the awareness-holder of maturation (rnam-smin rig-’dzin), the
awareness-holder who dominates life (tshe-dbang rig-’dzin), the awareness-holder of the
great seal (phyag-rgya chen-po’i rig-’dzin) and the awareness-holder of spontaneous pres-
ence (lhun-grub rig-’dzin). See Cornu, 2006: 694. It might be recalled that gNubs-chen
Sangs-rgyas ye-shes himself is credited with having attained the accomplishment of an
awareness-holder dominating life. See Guru bKra-shis, 1990: 246.
156 C 276–277.
157 See Meinert, 2002: 301.
158 On this and the other views of rDzogs-chen, see Esler, 2012b: 88–91.
159 C 320.3: […] lhun gyis grub pa’i chos nyid ngang chen po’i rang bzhin du lhag ma med par
sangs rgyas so/.
160 C 322.
Traces of Abhidharma in the bSam-gtan mig-sgron 345
161 Nyang-ral nyi-ma ’od-zer, 1988: 472: chos mngon pa mdzod ni/ sngon ka cog zhang gsum nas
brgyud pa (bar) skabs su stongs pa la/ dus phyis smri tis ’gyur gsar mdzad de nas thog ma’i
bshad pa rnams mchims rgyud kyis dar bar mdzad/. This may be translated as follows: “As
for the Abhidharmakośa, it was previously transmitted through the triad Ka-ba dpal-
brtsegs, Cog-ro klu’i rgyal-mtshan and sNa-nam zhang Ye-shes-sde, yet in the meanwhile
[its transmission] became extinct. Later, it was renewed by Smṛti, after whom the former
explanations were propagated through the lineage of mChims”.
162 Martin, 2002: 337.
163 On some of Smṛtijñānakīrti’s literary accomplishments, see Tanzin, 2013: 25–26.
346 Esler
Abbreviations
Bibliography
Meinert, Carmen (2002). “Chinese Chan and Tibetan rDzogs chen: Preliminary Remarks
on Two Tibetan Dunhuang Manuscripts”, In Religion and Secular Culture in Tibet:
Tibetan Studies II. Edited by Henk Blezer. PIATS 2000, vol.2. Leiden: Brill,
pp. 289–307.
rMi-lam bstan-pa (Svapnanirdeśa), In dKon-mchog brtsegs-pa’i mdo (Ratnakūṭa-sūtra).
In KD, vol.39, 406–473.
Miyazaki, Izumi (2004). “The Gradualist Chapter of the bSam gtan mig sgron and the
Teaching of Kamalaśīla”, Journal of Indian and Buddhist Studies 52/2, pp. 899–902.
Monier-Williams, M. (2001). A Sanskrit-English Dictionary. New Delhi: Asian Educational
Services.
Nasr, Seyyed Hossein (1989). Knowledge and the Sacred. Albany NY: State University of
New York Press.
Negi, J.S. (2006). Dharmasaṅgraha-Kośaḥ. Sarnath: CIHTS.
gNubs-chen Sangs-rgyas ye-shes (1974). rNal-’byor mig-gi bsam-gtan or bSam-gtan mig-
sgron: A treatise on bhāvanā and dhyāna and the relationships between the various
approaches to Buddhist contemplative practice. Reproduced from a manuscript made
presumably from an Eastern Tibetan print by ’Khor-gdong gter-sprul ’Chi-med rig-
’dzin, Smanrtsis shesrig spendzod, vol.74. Leh: Tashigangpa.
Nyang-ral nyi-ma ’od-zer (1988). Chos-’byung me-tog snying-po sbrang-rtsi’i bcud. Lhasa:
Bod-ljongs mi-dmangs dpe-skrun-khang.
Partridge, Eric (1982). Origins: A Short Etymological Dictionary of Modern English.
London: Routledge & Kegan Paul.
Pelliot tibétain 116.
Piatigorsky, Alexander (1984). The Buddhist Philosophy of Thought: Essays in Interpretation.
London: Curzon Press.
Rahula, Walpola (1971). Le Compendium de la Super-Doctrine (Philosophie)
(Abhidharmasamuccaya) d’Asaṅga. Paris: École Française d’Extrême-Orient.
Schuon, Frithjof (1993). Treasures of Buddhism. Bloomington IN: World Wisdom Books.
Schuon, Frithjof (1995). L’Œil du Coeur. Lausanne: L’Âge d’Homme.
Stcherbatsky, Theodore (1970). The Central Conception of Buddhism and the Meaning of
the Word “Dharma”. Delhi: Motilal Banarsidass.
Tanzin, Lopon P. Ogyan (2013). “Assessing the Greatness of Tibet’s Early Translations
according to Rong-zom Mahāpaṇḍita”, Translated by Dylan Esler. Temenos Academy
Review 16, pp. 21–42.
Van den Broeck, José (1977). La Progression dans la Méditation (Bhāvanākrama de
Kamalaśīla), Traduction du sanscrit et du tibétain. Bruxelles: Institut Belge des Hautes
Études Bouddhiques.
350 Teng
Kuiji’s Abhidharmic Recontextualization of Chinese Buddhism 351
Part 3
Philosophical Studies
∵
352 Teng
Madhyamaka in Abhidharma Śāstras 353
Chapter 11
1 Introduction
There are only a few very brief references to the Madhyamaka philosophi-
cal standpoint in the preserved Abhidharma śāstras. In the ninth chapter of
the Abhidharmakośabhāṣya (AKBh), Vasubandhu accuses those Buddhists
who propound the existence of pudgala (Vātsīputrīyas) and those who con-
tend that nothing exists (sarvanāstitāgrāhaka) of introducing heresy into
the Buddhadharma. These doctrines are wrong and do not lead to libera-
tion (mokṣābhāvadoṣa) (Pradhan, 1975: 472, lines 13–14). In his commentary
(Vyākhyā) on the AKBh (Wogihara, 1932–1936: 710, 31f.), Yaśomitra explicitly
identifies those who contend that nothing exists with the Madhyamaka school.
Vaibhāṣika master Saṃghabhadra, speaking in his *Nyāyānusāra about those
Buddhists who are not worthy of being called members of Sarvāstivāda, men-
tions those who “under the empire of distorted mirror hold that all dharmas
are devoid of own nature and that even present existence of dharmas is illu-
sory and wrong” (de La Vallée Poussin, 1937: 89). Another Vaibhāṣika śāstra, the
Abhidharmadīpa (kārikā 299 and vṛtti), criticizes those propounders of empti-
ness (śūnyatāvādinaḥ) who contend that nothing exists (sarvaṃ nāsti) in any
of the three time periods, calling them “upholders of destruction” (vaināśika)
and “propounders of the emptiness of the connection (between cause and
effect)” (ayogaśūnyatāvādins).1
This situation is slightly different in the case of the *Tattvasiddhi2 (hence
forth: TS) of Harivarman (third to fourth century CE), who, according to Shōryū
Katsura, was a member of the Bahuśrutīya school.3 If so, this Śāstra is the only
surviving representative of Mahāsāṃghika Abhidharma literature. The Śāstra
also discloses a strong Sautrāntika affiliation on many doctrinal issues. The TS
was translated into Chinese in the early fifth century, and had a huge influence
on Chinese Mādhyamikas at some point, as it was wrongly understood as a
proper Mādhyamika work. Nevertheless, although composed as a typical
Abhidharma śāstra centered around the four noble truths, some of its key
philosophical views have an undeniable Madhyamaka tinge.
The relevant portion of the Śāstra I will focus on in this chapter, is situated
in the fourth chapter called “Nirodhasatya,” sections 142–151. I will focus on
three issues:
Our part of the text begins with the enumeration of the four wrong doctrines
that presumably concern the existential status of objects of everyday experi-
ence, namely the doctrine of unity or identity (ekatva), of diversity or difference
(nānātva), of unspeakability (anirvacanīyatva) and of non-existence (abhāva)4
(142). According to the first doctrine, parts of, e.g. a pot (color, smell etc.) are
identical to the pot itself; according to the second, they are different from it;
according to the third, it cannot be said whether they are identical to or differ-
ent from the pot; according to the fourth, the pot does not exist (at all).
3 Cf. his argumentation on this point as well as on the TS’s doctrinal affiliations and disagree-
ments with other traditional Buddhist schools in Katsura, 1974: 29–49.
4 Katsura, 1974: 181, is of the opinion that Harivarman here presents four arguments concerning
“the relationship between concepts and dharmas”; Priestley, 1970: 31, thinks so also (concepts
and real phenomena). This would hold for the first three positions (the relationship between
parts and the whole), but not for the fourth position (“the pot does not exist”). I think that
Harivarman here more generally provides four possible explanations of the existential status
of everyday objects.
Madhyamaka in Abhidharma Śāstras 355
A brief note on the first three doctrines: the doctrine of identity is wrong
because the idea (buddhi) of parts (color etc.) and the idea of the pot are differ-
ent. Only the parts exist in reality, while the pot exists only by name since its
existence is “borrowed,” being merely (conceptually) based on its (real) parts
(143); the doctrine of diversity or difference is wrong because there is no pot
apart from its color etc., since no idea of it arises in their absence, and it is
therefore illogical for the pot to exist apart from its color etc. (144); the doctrine
of unspeakability is wrong because no substantial dharma (dravyadharma) is
unspeakable in terms of unity and diversity – each phenomenon (dharma) has
particular characteristics (svalakṣaṇa), which implies its identity with itself
and its difference with regard to other (dharmas), and so it is speakable: this is
rūpa, etc. (145)
For Harivarman, all of these doctrines, including the fourth (the pot does
not exist) to which we will now turn, are wrong because the pot etc. is nomi-
nally or conceptually existent (prajñaptisat).5
Harivarman dedicates the most space by far to the presentation and criti-
cism of the fourth doctrine, that of non-existence (abhāva). He firstly provides
a general criticism of this doctrine, or better said, outlines some unwanted
consequences of this doctrine that are very similar in nature to the objections
of Nāgārjuna’s anonymous opponent from the beginning of the Madhyama
kaśāstra (MMK) 24 (1–6), as well as elsewhere, e.g. in the Vigrahavyāvartanī
(VV). A free translation of some of these objections follows (TS, 146, 366–367):
If all dharmas are abhāva, then there is neither virtue nor its opposition
nor deliverance. If it is thought that what exists does not exist, nor will
that position (‘apprehension,’ grāha) exist, because in that case neither
speaker nor listener exist. Doctrines about existence and non-existence
are expressed through faith either in perception or in inference or in
scriptures. That there is nothing (kiñcana nāstīti) does not come under
any one of the above categories … You think that the idea of a dharma
arises as a result of a delusion. But if everything is non-existent
(abhāvatva), this illusion will also be non-existent. How will one act? If
you think that all dharmas are non-existent, in dependence on what
basis does this knowledge arise? No knowledge arises in dependence on
the non-existent. That which knows dharmas is knowledge. If dharmas
were in the nature of absolute non-existence (atyantabhāvā), then
5 Harivarman defines prajñapti as a “mental construct” (vikalpa) based on the five aggregates
(pañcaskandhān upādāya) (141). For other definitions of prajñapti: see Katsura, 1974:
179–181.
356 Kardaš
people would do anything they like. But wise people are interested in
charity, etc. … The pot, etc. are cognized through the perceptive faculty,
and yet you think that every perceived dharma is a non-entity. None will
have faith in the sūtras if they teach non-existent dharmas (abhāvadharma
katvācca na sūtre śraddhādhīta).
Then Harivarman provides the abhāvavādin (or nāstika, “nihilist”) the oppor-
tunity to philosophically elaborate on his position. Although not sharply
distinguished, it seems that the advocates of abhāva put forward three argu-
ments, epistemological and ontological in nature, for the non-existence of
dharmas. Here is a brief account of some of these arguments (cf. also Katsura,
1974: 183–184 and Priestley, 1970: 31–32):
b) Refutation of perception
Most arguments against the possibility of perception utilize the Ābhidharmika
concept of momentariness, turning it against their upholders:
b1) (Question:) Visual consciousness first grasps color, and then non-sen-
sory consciousness remembers it. (Answer:) … Visual consciousness, after
seeing color, disappears, and then non-sensory consciousness arises.6 This
6 This Sautrāntika position was criticized by Saṃghabhadra in his defence of the Vaibhāṣika
position according to which visual objects, the visual faculty, and visual consciousness
necessarily arise simultaneously because otherwise the visual faculty and visual objects
could not serve as conditions for visual consciousness; cf. Dhammajoti, 2007: 137. The
Sautrāntika solution to this problem is that contact between visual objects and the cor-
responding sense-faculty in the first moment produces an “image” (ākārā) of the object,
Madhyamaka in Abhidharma Śāstras 357
consciousness does not see color. If it does not see it, how can it recollect it? In
that case, the blind would also recollect it [without seeing it].7
b2) [You may say that] non-sensory consciousness arises from visual con-
sciousness, and hence recollects it. This cannot be. Why? Because in that case,
all consciousness of the last [moment] that originated from visual conscious-
ness would recollect it and never forget it.8
b3) The eye cannot see rūpa regardless of whether it reaches (prāpya) it or
not (aprāpya). In the first case, as has already been shown, visual conscious-
ness disappears after (possibly) reaching rūpa, having the characteristic of the
past (atītalakṣaṇam), and so cannot “deliver” its content to non-sensory con-
sciousness. In the second case, the absurd consequence would be that visual
consciousness would grasp any rūpa.9
b4) If the eye consisted of four great elements and would see color, the ear
etc. would also see, because it similarly consists of four great elements.10
b5) Because of the momentariness (kṣaṇika) of consciousness (citta) and
sound (śabda), none can actually hear the unity (ekatva) of e.g. the word
“puruṣa,” but only the momentary arising and disappearance of syllables one
after another (‘pu,’ ‘ru,’ ‘ṣa’).11
b6) Non-sensory consciousness does not grasp dharmas; for it does not
grasp color, taste, odor and touch in the present moment…12
which is an exact mental representation of the object and serves as a condition for the
origination of corresponding visual consciousness in the second moment.
7 (pṛ) cakṣur vijñānena rūpe gṛhīte tato manovijñānam anusmarati … (u) cakṣur vijñānaṃ
rūpaṃ dṛṣṭvā niruddham eva | tata ūrdhvaṃ manovijñānam utpadyate | manovijñānam
idaṃ na rūpaṃ paśyati | adṛṣṭvā katham anusmaret … andho ’pi rūpam anusmaret, TS, ibid.
8 (pṛ) cakṣur vijñānān manovijñānam utpadyate | ato ’nusmarati | (u) maivam | kasmāt |
sarvāṇi caram acittāni cakṣur vijñānam upādāya samutpannāni anusmareyuḥ | na eva
vismareyuḥ, TS, ibid.
9 yadi vadasi cakṣuḥ paśyati iti | kiṃ rūpaṃ prāpya paśyati kiṃ vāprāpya paśyati | yadi
prāpya [paśyati iti] tadā na paśyati | cakṣur nātītalakṣaṇam iti idaṃ pūrvam eva
pratipāditam | yady aprāpya paśyati iti | tadā sarvasthaṃ rūpaṃ paśyet, TS, ibid., 369.
10 kiñca cakṣuś caturmahābhūta[mayam] | yadi cakṣuḥ paśyati | śrotrādīni api paśyeyuḥ |
caturmahābhūtasāmyāt, TS, ibid.
11 cittaṃ hi kṣaṇikam | śabdo ’pi kṣaṇikaḥ | yathā vadanti puruṣa iti | aya [mekatva]vādo na
śrāvyaḥ | kasmāt | «pu» śravaṇam anu vijñānaṃ na «ruṃ» śṛṇoti | «ruṃ» śrutvā na «ṣaṃ»
śṛṇoti, TS, 148, 370. This was essentially Vasubandhu drawing an absurd consequence,
obviously taken over from some earlier source (probably the Mahāvibhāṣāśāstra), of the
Vaibhāṣika’s understanding of syllables (along with names and phrases) as real existing
entities (dravyadharmas); cf. Bhāṣya to AK 47ab and also Cox,1995: 168.
12 Manovijñānam api dharmān na gṛhṇāti | kasmāt | manovijñānaṃ hi na pratyutpannān
rūparasagandhasparśān gṛhṇāti | [yat] atītamanāgataṃ tan na asti iti …, TS, 150, 371.
358 Kardaš
13 There is, however, a hint of this argument in Vaidalyaprakaraṇa (32–39) where Nāgārjuna
(if he is the author of this work at all) refutes the Naiyāyika’s five-membered (avayava)
syllogism since they are not subject to a whole (avayavin); and even if they were one with
a whole, they would be identical with it, etc. (cf. Lindtner, 1990: 89–90).
14 For example, he uses the argumentative structure “x cannot make contact with y either by
reaching (prāpya) it or not reaching (aprāpya) it” (cf. b3 above) in the context of light and
darkness (MMK 7.10–11; also VV 38–39).
Madhyamaka in Abhidharma Śāstras 359
15 The commentaries (e.g. Candrakīrti), of course, do: tatra yadi pūrvaṃ kāraṇaṃ paścāt
kāryaṃ syāt, akāryakaṃ kāraṇaṃ nirhetukaṃ syāt | atha pūrvaṃ kāryaṃ paścāt kāraṇam,
evam api kāraṇāt pūrvaṃ kāryaṃ nirhetukam eva syāt | atha yugapat kāryakāraṇe syātām,
evam ubhayam apyahetukaṃ syāt. PP 225.6–8 (vṛtti to MMK 11.7–8); cf. also vṛtti to MMK
9.12 (the same type of investigation between seeing and perceiving a subject, ātman).
16 Cf. PP 376.10–15 (vṛtti to MMK 18.10): naiva yad eva bījaṃ sa eva aṅkuraḥ, janyajanakayor
ekatvaprasaṅgāt … ananyatvāc ca aṅkurāvasthāyām aṅkuravadbījagrahaṇam api syāt,
bījavac ca aṅkurasya api grahaṇaṃ syāt … na ca anyad api tattasmāt | nāpi bījādaṅkura
syānyatvam, bījamantareṇa api aṅkurodayaprasaṅgāt; cf. also vṛtti to MMK 20.19–20.
17 Ācāryabuddhapālitas tv āha – na svata utpadyante bhāvāḥ, tadutpādavaiyarthyāt,
atiprasaṅgadoṣāc ca | na hi svātmanā vidyamānānāṃ padārthānāṃ punarutpāde prayoja-
nam asti | atha sannapi jāyeta, na kadācin na jāyeta iti … na parata utpadyante bhāvāḥ,
sarvataḥ sarvasaṃbhavaprasaṅgāt … dvābhyām api nopajāyante bhāvāḥ, ubhayapakṣā
360 Kardaš
455) “Were the eye to apprehend [the object] without going, then it should
perceive this entire world!” (CŚŚ XIII 15; cf. Lang, 1983: 456) “If the eye goes far
and far, it sees by and by”. (ŚŚ, cf. Tucci, 1929: 51)
b6 argument: “Deprived of the sense faculties, what will the mind (citta) do,
even after it has gone [to the object]?” (CŚŚ XIII 21; cf. Lang, 1983: 460)
c1 argument: “If the fruit exists, then, because the fruit is pre-existing in the
cause, the cause would be without effect on account of the non-existence of
the fruit”. (ŚŚ; cf. Tucci, 1929: 61; cf. also Tucci, 1929: 66, 70) “An existent dharma
does not arise from an existent dharma … a non-existent dharma does not
arise from an existent dharma”. (CŚŚ XV 14; cf. Lang, 1983: 511; cf. also CŚŚ XV 15,
Lang, ibid. and CŚŚ XI 12d and 15c)
In the light of these similarities in reasoning, and especially in the textual
topical ordering of false doctrines (ŚŚ chapters 3–8 and TS 143–151), my conten-
tion (which, of course, cannot be completely conclusive) is that the works of
Āryadeva, not those of Nāgārjuna, were the main source for Harivarman’s pre-
sentation (and knowledge) of Madhyamaka philosophical views and reasoning,
especially bearing in mind the more highly developed style of argumentation
presented in Āryadeva’s works compared to Nāgārjuna’s.20 This advancement,
as has been shown above, is also reflected in Harivarman’s treatise.21 Moreover,
there is one explicit (and affirmative) citation in TS (105) of CŚŚ (II 8), albeit in
a context that is not relevant to our presentation, but that nevertheless explic-
itly suggests that Harivarman was at least familiar with Āryadeva’s works.
The next issue I would like to discuss is the nature of Harivarman’s criticism of
an apparently Madhyamaka position which he, like the other Abhidharma
sources mentioned, terms abhāvavāda, calling its followers abhāvavādins (or
nāstikas). I believe that his criticism streams from his total ignorance of the
actual (Nāgārjuna’s) Madhyamaka position regarding the existential status of
objects of everyday experience and the conventional (vyavahāra, saṃvṛti) in
20 Frauwallner (1956: 218) has noted that although Āryadeva agrees with Nāgārjuna in all
important views, he exceeds him in his “style of presentation” (Art der Darstellung).
Unlike Nāgārjuna, Āryadeva treats the philosophical views of his opponents more thor-
oughly and in greater detail.
21 We cannot identify any major figure in the development of Madhyamaka philosophy
between the time of Nāgārjuna (ca. second century CE) and the classical commentarial
tradition (Buddhapālita, Bhāvaviveka and Candrakīrti, from the fifth century CE onward)
other than Āryadeva.
362 Kardaš
general, and that Āryadeva is to blame for this.22 However, before providing
some textual evidence for this thesis, it would not be out of place here to briefly
outline Harivarman’s ontological position, which is partly based on the stan-
dard Abhidharma ontology and is obviously crucially influenced in part by the
Mādhyamika position itself.
In certain parts of the text, Harivarman posits the standard Abhidharma
theory of the two truths (saṃvṛti- and paramārthasatya), whose “originator” is
the Buddha himself (TS 3, 12) but whose actual preaching was based on praj-
ñapti (“concepts”) (TS 4, 16), i.e. on saṃvṛtisatya. TS 141, 354–355 gives the
standard Abhidharma account of the content of the two truths:
22 Although, unlike his commentators, Nāgārjuna never discusses the actual “content” of the
conventional, he is more than explicit in that his emptiness-based philosophical analysis
should be understood within the realm of the conventional, cf. e.g. his auto-commentary
to VV 28: “We, however, do not say ‘all dharmas are empty’ without relying on the conven-
tional truth, rejecting the conventional truth. For it is not possible to teach the Dharma
without having recourse to the conventional truth” (api ca na vayaṃ vyavahārasatyam
anabhyupagamya vyavahārasatyaṃ pratyākhyāya kathayāmaḥ śūnyāḥ sarvabhāvā iti / na
hi vyavahārasatyam anāgamya śakyā dharmadeśanā kartuṃ). On the other hand, this
emphasis on the conventional grounding of the teaching of emptiness is completely
absent in Āryadeva, and that could give his opponents an impression that for the Mādhya
mika “does not exist ultimately” essentially means “does not exist at all” (abhāvavāda).
23 “paramārthasatyaṃ yaduta rūpādayo dharmā nirvāṇañ ca | saṃvṛtisatyaṃ yat prajñapti
mātraṃ niḥsvabhāvam | yathā rūpādipratyayo ghaṭaḥ sidhyati | tathā pañcaskan
dhapratyayaḥ puruṣaḥ sidhyati”.
24 That does not mean, according to Harivarman, that what exists conceptually does not
exist at all. It simply means that it does not exist substantially (na ca prajñaptisannāsti iti
vastumātraṃ na bhavati, TS, 35, 93) and that it is devoid of own nature (niḥsvabhāva; TS
142, 360).
Madhyamaka in Abhidharma Śāstras 363
The arising of knowledge is not in prajñapti, but rather in rūpa, etc. that
is present before man’s eyes. After that, [a man conceptually] construes
with [his] non-sensory consciousness: “I see the jar” … In prajñapti, doubt
arises as to whether it is a pillar or a man. But in [the case of] rūpa, etc.,
no doubt arises as to whether it is rūpa or śabda.25
(Answer): It is said [so] for the good of men [because] there are some
[men] who harbor the notion of paramārtha towards the pañcaskandhas.26
It is said that conditioned by birth [there is] old age-and death. This is
called the middle way. This should be understood [here as] that there is
no old age-and death from the paramārtha level. That conditioned by
birth [there is] old age-and death is said from the saṃvṛti level.27
It thus follows that teachings based on the investigation of dharmas are noth-
ing but an “expedient means” (upāya) by which to reach nirvāṇa devised for
those who “harbor the notion of paramārtha towards the pañcaskandhas”:
“Meditation on skandhas, dhātus, āyatanas, etc. are [only] different doors and
expedient means for entering nirvāṇa”.28
This “shift” in the meaning of the two truths places the Ābhidharmika’s
paramārtha analysis on the level of saṃvṛti analysis29 together with the “origi-
nal” saṃvṛtisat (or prajñaptisat), leaving the emptiness or non-existence of all
dharmas as the sole “content” of paramārthasatya. Harivarman obviously con-
siders the Ābhidharmika’s paramārtha “context-bound,” i.e. relative to the
26 (pṛ) yadi pañca skandhāḥ saṃvṛtisatyataḥ santi | kasmād ucyante rūpādayo dharmāḥ
paramārthasatyā iti | (u) sattvānāṃ kṛta ucyante | santi kecit pañcaskandheṣu samutpanna
paramārthasaṃjñāḥ (TS 153, 376).
27 kiñcāha- jātipratyayaṃ jarāmaraṇaṃ madhyamā pratipadityucyate | nāsti jarāmaraṇaṃ
paramārthata ity uktamiti jñātavyam | saṃvṛtita ucyate jātipratyayaṃ jarāmaraṇam iti (TS
153, 377). In this example, Harivarman seems to use the same argument as the
Mādhyamikas do in their pointing to the emptiness of dharmas – the fact of the condi-
tionality of all dharmas. There are even more explicit examples that corroborate this ‘bor-
rowing’ to which we will turn soon.
28 skandhadhātvāyatanādīnāṃ bhāvanā nānādvārāṇyupāyā nirvāṇe ’vataraṇasya (TS 201,
537).
29 Cf. TS 153, 376: “Then the pañcaskandhas do not exist substantially? They do not. They
exist from the saṃvṛti level” ((pṛ) … kimidānīṃ na santi vastutaḥ pañcaskandhāḥ | (u) na
santi vastutaḥ | santi tu saṃvṛtitaḥ).
Madhyamaka in Abhidharma Śāstras 365
Ābhidharmika’s saṃvṛti level, while the last and final paramārtha is ‘context-
free,’ i.e. it is absolute.30
Following this analysis, Harivarman develops a theory of three basic types
of consciousness with three corresponding types of content: prajñapticitta
(makes e.g. a pot its content), dharmacitta (makes rūpa, etc. its content, thus
suppressing the idea of the pot), and śūnyatācitta (makes the non-existence or
emptiness of the rūpa, etc. its content, thus suppressing the idea of dharma
lakṣaṇa). The last citta is suppressed by entering into nirodhasamāpatti (cf. TS
141, 354). Harivarman seems to clearly distinguish here between śūnyatā as the
(still conceptual) content of śūnyatācitta and the experiential “state” of
śūnyatā, which is beyond any relational designation.
There is no doubt that Harivarman adopted the Madhyamaka position
regarding the final nature of all dharmas “as they are” (yathābhūta).31 Many of
his characterizations of the real nature of dharmas appear to have been copied
from a Madhyamaka śāstra. Here are some examples:
30 Priestley, 1970: 35, believes that Harivarman here “imperfectly formulated [the] theory of
three truths: conventional truth, which is truth based on concepts; conventional ultimate
truth, which is based on phenomena; and ultimate truth, which is the truth of Nirvāṇa”.
However, I believe he is incorrect, as his “conventional ultimate truth” (corresponding to
dharmacitta) is nothing but the “Buddhist conventional truth” especially “designed” for
those, as Harivarman puts it, who “harbor” the notion of paramārtha towards the pañca
skandhas” (see above). “Conventional truth” (corresponding to prajñapticitta) is, on the
other hand, conventional truth for all others, i.e. ordinary men (pṛthagjanas). What is
crucial in this scheme, is that the respective content of all levels of consciousness appears
as paramārtha, which can be downgraded to the level of saṃvṛti only by a subsequent
and, so to speak, more investigative type of consciousness. Consequently, and opposed to
Priestley, 1970: 36, I believe that Harivarman does not analyze pots, rūpas, etc. ontologi-
cally, but rather epistemologically as the content of consciousness.
31 Cf, TS 16, 50: śūnyataiva yathābhūtam and TS 192, 506: nirodhaḥ sa niyataṃ yathābhūta ity
ataḥ paramārthasan iti.
32 TS 20, 69: tathā cet tathāgatādayo ’saṃskṛtadharmā api santaḥ syuḥ … vastutastu asan-
taste | ato jñāyate skandhadhātvāyatanasaṅgṛhītā dharmā na sallakṣaṇā iti.
33 TS, 130, 317: skandhavṛttilakṣaṇavṛttim anusarato nātmam atiratyantaṃ prahīyate |
hetupratyayānām anirodhāt.
366 Kardaš
There are even some passages that strongly resemble the classical Mādhyamika
argument for the śūnyatā of all dharmas – their dependent origination which
prevents the generation of mental and linguistic proliferations about them:
In the light of the cited examples and their close affinity with the Mādhyamika
position, it is difficult to decide what exactly the point of contention between
Harivarman and the Madhyamaka is as the former sees it. My contention is, as
has already been stated, that he exhibits a total ignorance of the actual
Mādhyamika position regarding the existential status of objects of everyday
experience (including dharmas, rūpas, etc.).
Firstly, let us recall Harivarman’s initial objections (see above) regarding the
nāstika thesis that everything is non-existent. These objections are much the
same as those that appear in MMK and VV, and apparently belong to a realist
standpoint that does not necessarily solely include Nyāya realism, but also the
realism of e.g. Sarvāstivāda. However, in one passage of TS, Harivarman once
again seems to take the side of the Mādhyamika in answering a question that
smacks of his abhāvavādin or nāstika position:
(Question:) A dharma that does not exist ultimately does not exist at all.
Why is it then said that it exists according to saṃvṛti [‘empirically’]?
(Answer:) It is in agreement with all mundane [practices to say] “it exists;”
that is to say, action and retribution, or bondage and deliverance. [But]
all this is born out of illusion. Why? These pañcaskandhas are empty, illu-
sory, and because they continuously arise, they are like a flame. [But] for
the sake of the crossing over of the worldlings, [they] are spoken of as if
they would exist. If it were said that [they] do not exist [at all], then the
worldlings, being confused [frightened], would fall into nihilism…42
What is slowly being revealed here is that the nāstikas, as Harivarman pictures
them, deny the existence of dharmas straightforwardly, even on the saṃvṛti
level. Dharmas for the nāstikas simply do not exist in any way or any sense (or
any level) whatsoever. They hence do not (and do not need to) make a distinc-
tion between saṃvṛtisatya and paramārthasatya. This is unacceptable for
Harivarman, as in this case all our activities would be paralyzed, and the grad-
ual revelation of the (original) emptiness or non-existence of dharmas would
be based on nothing. Emptiness can be reached only through the analytical
investigation of the actual content of the corresponding consciousness that
should be considered to be existent at that level.43 It cannot be reached by
simple denial of that content:44
Just as a man who is ignorant of painting, etc. says that these diverse
dharmas do not exist, in the same way, you, having not proven anything,
say “these dharmas do not exist”. They exist for he who knows [them],
they do not exist for he who does not know [them]. It is just as when one
who is blind from birth would say that black and white do not exist
because he does not see them. It is not proper [to say] that colors do not
exist just because they are not seen.45
anuvartate | yadi nāsti iti vadet | tadā pṛthagjanā vyāmuhya yadi vā ucchedavāde
pateyuḥ….
43 Cf. TS 152, 374: “Only tathāgatas are capable of analytical knowledge of dharmas …
tathāgatas are, on the other hand, [able to] dispose of the non-substantiality of all dhar-
mas in their nature, all of their aspects as well as their characteristics, particular and uni-
versal … the knowledge of emptiness is easy to obtain, but the analytical knowledge of
dharmas is difficult to develop” (tathāgatāḥ kevalaṃ dharmavivecanajñānasamarthāḥ …
tathāgatāḥ paraṃ sarvadharmāṇāṃ sarvākāraṃ prakṛtito naissvābhāvyaṃ viśeṣasāmāny
alakṣaṇāni sarvāṇi pratividhyanti … evaṃ śūnyatājñānaṃ sulabham dharmāṇāṃ
pravicayajñānaṃ durutpādam).
44 Cf. TS 130, 313: “One who does not analyze does not enter emptiness” (yo na vivecayati | ko
’vakrāmati śūnya[tāyā]m).
45 TS 152, 374: yathā kaścit citrāṅkanādidharmavikalpamajñātvā vadati tāni na santi iti | tathā
bhavānapi yat kim apy asādhayitvā vadati nāstīdaṃ- vastu iti | jñātus tu asti | ajñātuḥ
punar nāsti | yathā jātyandho vadati nāsti kṛṣṇamavadātaṃ vā, mayādṛṣṭatvāt | na
cādṛṣṭatvād rūpāṇi na santīti sambhavati).
Madhyamaka in Abhidharma Śāstras 369
How then is one to resolve the obvious puzzle of the fact that Harivarman first
borrows a fundamental Madhyamaka ontological position nearly verbatim –
including a good deal of its standard argumentation, while simultaneously
criticizing the nāstikas or abhāvavādins who promote the same Madhyamaka
prasaṅga type arguments for the fundamental non-existence or emptiness of
dharmas as we know from their works? There are, I believe, at least two solu-
tions to this puzzle. One is hypothetical, and I cannot present any reliable
evidence for it as of yet. According to this hypothetical solution, there was a
brief tendency within early Mahāyāna philosophical circles centered around
the use of Prajñāpāramitā sūtras to interpret their straightforward negation of
all dharmas literally without reliance on any interpretative device on which
such a negation might be based. That is to say, they did not employ such inter-
pretative devices as distinguishing the two truths or pointing to the emptiness
of all dharmas through a proper understanding of pratītyasamutpāda, as the
Mādhyamikas (Nāgārjuna) did. With the latter, they could share only the prac-
tice of drawing unacceptable consequences (prasaṅgavākya) from all theses
that tried to establish the existence or validity of anything, as is evident from
46 “If one says without reaching the intuition of emptiness that a being does not exist, it is a
wrong view” (yadapratilabdhaśūnyatāprajño vadati nāsti sattva iti | iyaṃ mithyādṛṣṭiḥ, TS
141, 356, and also TS 141, 357: “If one who has not reached the reality of emptiness unfortu-
nately harbors the view that nothing exists, he falls into a wrong view, namely nihilism”
(apratilabdhaśūnyatāt attvo nāsti kiñcanetidṛṣṭyā durgatau patati yadutocchedadṛṣṭi[rūpa]
mithyādṛṣṭiḥ).
47 This was hinted at by Harivarman when he stated that “there is no mundane emptiness
(laukikī śūnyatā)” (nāsti tu laukikī śūnyatā, TS 191, 505. I hence believe Priestley to be
incorrect when he states that Harivarman distinguishes between “conceptual emptiness
and phenomenal emptiness” (1970: 34) for there is not, so to speak, any “mark” of empti-
ness on that level of consciousness.
370 Kardaš
Harivarman’s account. But this practice, if not backed by the conceptual struc-
ture of the two truths (MMK 18.8) and by the equation of pratītyasamutpāda
with śūnyatā (MMK 24.18), inevitably leads to nihilism.48 It was Nāgārjuna then
who performed this task, and so Harivarman was actually criticizing this philo-
sophical tendency within Mahāyāna philosophy and not the Mādhyamikas.
The second solution is, on the other hand, more factually based. In the first
part of this chapter, I attempted to provide some evidence of the fact that
Āryadeva’s works, and particularly ŚŚ, were probably the main source for
Harivarman’s presentation of the nāstikas’ or abhāvavādins’ position. If we
carefully read Āryadeva’s work, we will find that it strangely lacks the most
fundamental Mādhyamika argument and conceptual (or hermeneutical)
framework, those being the equation of pratītyasamutpāda with śūnyatā and
the theory of the two truths. These two are, as already mentioned, the only
“positive tools” preventing the degradation of the Mādhyamika position into
pure nihilism. CŚŚ is wholly and solely dedicated to the utter deconstruction of
all concepts and dharmas in terms of their referent or existence, leaving no
room for an affirmation of their provisional or conventional status, much in
the same spirit of Harivarman’s nāstikas. On the other hand, Nāgārjuna and
Candrakīrti especially, were careful enough to qualify their statements about
the non-existence of dharmas with qualifications such as svabhāvatas (“in
terms of their own being” or “substantially”).49
I conclude, therefore, that Āryadeva’s highly negativistic approach in
defending the Mādhyamika position significantly or even crucially contrib-
uted to the subsequent misunderstanding and improper appreciation of the
Mādhyamika standpoint regarding the non-ultimate, i.e. saṃvṛti nature of
dharmas reflected both in Harivarman’s work and in later Abhidharma śāstras
such as AKBh, *Nyāyānusāra and Abhidharmadīpa, which unanimously qual-
ify the apparently Mādhyamika position as nihilistic.
48 Arguing against the equation of utter non-existence (abhāva) with emptiness in the con-
text of the Mādhyamika position, Candrakīrti rhetorically asks: “And who criticizes us in
the manner just described? It is he who does not know the infallible distinction between
the two truths as taught in the words of the Buddha” (kaścāsmākaṃ yathoktam
upālambhaṃ karoti? yo bhagavatpravacanopadiṣṭāviparītasatyadvayavibhāgaṃ na jānāti,
kevalaṃ granthamātrādhyayanapara eva iti (PP 492.28–29).
49 It, of course, goes without saying that Āryadeva was familiar with his Master’s theory of
the two truths, as well as with the “diamond argument” for emptiness, that being the
conditional existence of dharmas (pratītyasamutpāda). However, even in his major work,
CŚŚ, these notions rarely occur and are never mentioned straightforwardly (cf. CŚŚ VIII,
19–20; IX, 2–3, 25; XIV, 23).
Madhyamaka in Abhidharma Śāstras 371
5 Conclusion
We are still far from any promising clue as to who Harivarman’s nāstikas actu-
ally are. As presented by him, they are definitely Mādhyamika-like Buddhists
who abundantly employ prasaṅga argumentation for the non-existence of
dharmas, and this type of argumentation backed by nothing seems to be dis-
carded by Harivarman. If Harivarman’s attack on his nāstikas is in fact an
attack on the Mādhyamikas whose fundamental position he misunderstood
due to Āryadeva, then how is one to explain that, in other parts of the text, he
exhibits his knowledge of the conceptual structure of the two truths much in
the same line as the Mādhyamikas, including the understanding of pratītyasam
utpāda as a door to the śūnyatā-insight? Shōryū Katsura (1974: 184) believes
that “the fact that nāstikas’ attack is mainly directed at the Vaiśeṣikas and
Sāṃkhyas … seems to suggest that Harivarman is here intentionally copying
some arguments set by Nāgārjuna or his followers”. However, the nāstikas’
attack is in fact mainly directed at Buddhist (Abhidharma) doctrines, and only
in a few cases at these non-Buddhist schools (refutation of satkārya- and
asatkāryavāda, and, to some extent, refutation of parts and wholes). Our con-
fusion is further increased by the fact that Buddhapālita’s and Candrakīrti’s
nāstikas are probably not Buddhists, but are more likely Cārvākas/Lokāyatas or
Ajñānikas (agnostics), since these philosophical schools or trends (and only
they) are generally known in later Buddhist works as nāstikas.
Nevertheless, as stated in the first solution to our puzzle, could there have
been quasi-Mādhyamika Buddhist philosophers prior to or contemporaneous
with the early Mādhyamikas proper who propounded the non-existence of
dharmas and delighted only in reductio ad absurdum arguments to this end?
Abbrevations
Bibliography
Tucci, Giuseppe (ed.) (1929). Pre-Diṅnaga Buddhist Texts on Logic from Chinese Sources
(Part I: Texts. pp. 1–89. English translation of ŚŚ together with Vasu’s Commentary).
Baroda: Oriental Institute.
Vaidya, Paraśurāma Lakṣmaṇa (ed.) (1923). Catuḥśatakaśāstra: Études sur Āryadeva et
son Catuḥśataka. chapitres VII–XVI. Paris: Librairie Orientaliste Paul Geuthner.
Vaidya, Paraśurāma Lakṣmaṇa (ed.) (1960). Madhyamakaśāstra of Nāgārjuna with the
Commentary: Prasannapadā by Candrakīrti. Darbhanga: The Mithila Institute of post-
graduate studies and research in Sanskrit learning.
Wogihara, Unrai (ed.) (1932–1936). Sphutārthā Abhidharmakośavyākhyā by Yaśomitra.
2 vols. Tokyo: Publishing Association of Abhidharmakośavyākhyā.
Svalakṣaṇa (Particular) and Sāmānyalakṣaṇa (Universal) 375
Chapter 12
1 Introduction
I am sitting in the study room and experiencing the surrounding things, such
as table, pen, music, and bell sound from the elementary school not far from
the apartment where I live.1 There are also many thoughts bubbling up even
without notice. But I am still able to say that “I see a pen,” “I am reading Martin
Heidegger’s Being and Time,” “I have heard the bell ringing,” and so on. The
experience I just described contains both cognitive and psychological tones.
Sometimes I am in good mood when I am sensing and thinking. Sometimes
I am blue. If I go further to analyze my experience of, say, a pot of tulip on the
table, I see the color of cup shaped pedals, even though I am not able to pre-
cisely identify a correct term for that shade of color. But I do see something. In
Buddhist terms, that “something” is called “particular” (svalakṣaṇa). If I go on
to say, “The color of this tulip is purple,” then since the words “color” and “pur-
ple” in the statement are applicable to the qualities that are exemplified in
many individual entities, they are called “universal” (sāmānyalakṣaṇa).2
Imagine another scenario. If the Buddha, an enlightened being, sits in the
same room, does he have the same cognitive and psychological experience as
mine just described above? As claimed in many Yogācāra texts, the enlight-
ened cognition is pure perception only. As for the Buddha himself, he does not
need to engage in reasoning at the moment of enlightenment. All he experi-
ences is nothing but perception in the momentary flux. The difference of the
enlightened experience from the ordinary experience is that he does not con-
ceptualize the momentary flux of perception within the dichotomy of ‘”subject”
and “object,” “I” and “mine”.
However, the puzzle has not been completely solved. Is it true that the
enlightened one does not need to employ concepts and reasoning for commu-
nication after enlightenment? Or, does he only experience reality in perception
without any aid of conception and reasoning? These puzzles were discussed in
Kuiji’s 窺基 (632–682) Cheng weishi lun shuji 成唯識論述記 (Commentary on
the Treatise on the Establishment of Consciousness-Only), a foundational trea-
tise in the Chinese Yogācāra school in the seventh century.3 Concerning this
question, I will explore in this paper the different theories of cognition with
regard to svalakṣaṇa (particular) and sāmānyalakṣaṇa (universal) in the
Abhidharma and Yogācāra contexts. The main question is, what is the correct
cognition, the cognition of universal or the cognition of particular, which can
be employed to eliminate the psycho-cognitive hindrances for attaining final
liberation? This is the basic question raised by both the Abhidharma and
Yogācāra Buddhists.
As the Abhidharma literature shows, the above question is raised in the con-
text of mental and cognitive cultivation. The practitioner is guided to realize
the truth of impermanence, suffering, and selflessness, which will in turn help
free one from ignorance and contamination, through following the various
stages of meditation. “Impermanence,” “suffering” and “no-self” are considered
to be universals for the reason that they are universal truth of phenomena,
while the individual objects in the categories of five aggregates, twelve sense-
spheres, and eighteen elements are identified as particulars. When the question
is raised again in the same context, asking which kind of cognition, particular
or universal, leads to the final elimination of contaminations for attaining lib-
eration, the Abhidharma answers that it is the cognition of universal which
leads to liberation. This conviction is firmly stated in the Abhidharmama
hāvibhāṣa: “Only in the path which takes universal as the object, one is able to
eliminate the contaminations”.4 However, there are disagreements regarding
this issue among Yogācāra commentators.
Another important issue is about the ontology of universals. In both Western
and Indian (including Buddhist) philosophy, generally speaking, there are
three positions with regard to the ontological status of universals: (1) meta-
physical realism, which holds the view that universal is real in the sense that it
exists independently of mind; (2) nominalism, claiming that universal is not
real, while particular is real; (3) conceptualism, which argues that while uni-
3 The same passages in Kuiji’s Cheng weishi lun shuji is also found in his Dasheng fayuan yilin
zhang 大乘法苑義林章 and Huizhao’s Dasheng fayuan yilin zhang buque 大乘法苑義林
章補闕, which will be examined in the following of this article.
4 T.27.1545: 820a1–2: “唯共相境道能斷煩惱”.
Svalakṣaṇa (Particular) and Sāmānyalakṣaṇa (Universal) 377
versal is not ultimately real, it does exist in the mind in the form of concept.
Regarding the ontological status of universal, most of the Buddhists stand for
nominalism and reject metaphysical realism. However, when we consider the
various ontological stances taken by Sarvāstivāda, Sautrāntika, Madhyamaka,
and Yogācāra, the label of “nominalism” cannot be satisfactorily applied to all
Buddhist schools without further clarification. Could conceptualism be an
option for Yogācāra? In Yogācāra commentaries, we find some discussion rele-
vant to this issue. An important note with regard to the ontology of universals
is that those theories are not resulted from the metaphysical speculation only.
On the contrary, the debates on the ontological status of universals are mainly
concerned with the soteriological consequences. For underlying the various
theories of universal there is a fundamental belief that only truth will set free
those who are entrapped in ignorance and contamination: “Only mental atten-
tion to the reality (tattvamanaskāra) is capable of cutting off the afflictions”.5
In response to the above questions, firstly, we come to Kuiji who cited three
theories from the Buddhabhūmyupadeśa:6 (1) The first theory holds that all
characteristics of objects, including particular and universal, are cognized by
direct perception (pratyakṣa) in the concentrated state of mind, whereas
according to Buddhist logic (hetuvidyā) (i.e. Nyāyamukha and Nyāyapraveśa)
– particular and universal are experienced in the non-concentrated state of
mind by direct perception and inference respectively.7 (2) The second theory
holds that only particular is experienced in the concentrated state of mind.
8 T.43.1830: 584b12–18: “(2) Some hold that in the concentrated state of mind only the par-
ticular characteristic is cognized. However, it is introduced by means (upāya) of the uni-
versal characteristic. What is cognized is the truth/principle manifested in the various
characteristics of universal. It is provisionally called ‘cognition of universal characteristic’.
It is called ‘cognition of particular characteristic’ when it is not so. According to this rea-
son, some name ‘suchness’ (tathatā), which is also named ‘emptiness’ (śūnyatā) and ‘self-
lessness’ (nairātmya), as the universal characteristic of existents. Some say that ‘suchness,’
which is manifestation of two kinds of emptiness, is not a universal characteristic.
According to this theory, it is a provisional theory held by the Abhidharma to claim that
the cognition of universal characteristics is capable of eliminating the afflictions. As a
matter of fact, it is the discerning of particular characteristics that is capable of elimina-
tion”. (“第二說,有義,定心唯緣自相,然由共相方便所引,緣諸共相所顯理
者,就方便說名知共相,不如是者名知自相。由此道理,或說真如名空無我
諸法共相,或說真如二空所顯非是共相。由此義故,對法等說緣共相智能斷
惑者依方便說,實自相觀方能斷之”).
9 T.43.1830: 584b18–29: “(3) “As regards the true meaning [of dharmas] explained here, the
treatises on logic (hetuvidyā) offer a different account of svalakṣaṇa and sāmānyalakṣaṇa.
They claim that the real objects (tattvārtha) of all dharmas are named svalakṣaṇa for the
reason that the svalakṣaṇa and sāmānyalakṣaṇa of dharmas exist separately by them-
selves and hence are not shared with each other. When the conceptual mind establishes
the classes (jāti) of signifier and signified which is applied to dharmas in the similar way
that the thread is used to string flowers together, they [class, etc.,] are called “sāmānyala
kṣaṇa”. They are conceptual construct in the non-concentrated state. They are the object
of inference. All minds in the concentrated state, which are free from conceptualization,
are called “pratyakṣa”. Even if [in the concentrated state,] suffering, impermanence, and
so forth, of existents are cognized, they exist in each individual existent. Hence they are
called “svalakṣaṇa”. Although suchness (tathatā) is disclosed through sāmānyalakṣaṇa, it
Svalakṣaṇa (Particular) and Sāmānyalakṣaṇa (Universal) 379
to the Abhidharma teaching; the latter maintains that only the cognition of
universal is capable of eliminating the contaminations before realizing the
final enlightenment.
After having cited three theories of particular and universal, Kuiji asserts his
own view which disagrees with the Abhidharma theory. He says:
According to the above analysis, some Abhidharma [texts] hold the view
that the cognition which takes universal as the object is capable of
destroying the contaminations.10 The universal is thus named for the rea-
son that the conceptual mind uses class (jāti) to string the various
existents on [the level of] truth/thusness (tathatā). Some hold that uni-
versal is named for the meaning (artha) of existents which are grounded
in tathatā. The cognition of universal is thus named for that which takes
this [i.e. universal] as the object. As a matter of fact, the existence of
tathatā is real by nature, so it is not universal. In reality, the cognition of
particular [i.e. svalakṣaṇa] is capable of destroying the contaminations.
As explained above, [it is because] universal is conceptually designated.11
is not sāmānyalakṣaṇa, because it is the real nature of existents, and because it possesses
its own characteristic (lakṣaṇa). It cannot be named “sāmānyalakṣaṇa” simply because it
is neither identical nor different from all existents. It is because svalakṣaṇa is also neither
identical nor different from all sāmānyalakṣaṇas”. As a result, the treatises [on logic] hold
a view distinct from this (i.e. Buddhabhūmyupadeśa – T.26.1530.318b4–14)”. (“第三說,
如實義者,彼因明論立自共相,與此少異。彼說一切法上實義皆名自相。以
諸法上自相共相。各附已體 不共他故。若分別心立一種類 能詮所詮 通在諸法
如縷貫華 名為共相。此要散心分別假立 是比量境。一切定心 離此分別 皆名
現量。雖緣諸法苦.無常等。亦一一法各別有 故名為自相。真如雖是共相所
顯。以是諸法自實性故。自有相故。亦非共相。不可以其與一切法不一不異
即名共相。自相亦與一切共相不一異故。是故彼論與此不同”).
10 In Kuiji’s Cheng weishi lun biechao 成唯識論別抄, “Abhidharma” is identified as referring
to Asaṅga’s Abhidharmasamuccaya and Sthiramati’s Abhidharmasamuccayabhāṣyā. Cf.,
Sthiramati’s Abhidharmasamuccayabhāṣyā: T.31.1606: 727a10–12: “Question: By what kind
of intentional attention can one eliminate [the contaminations]? Answer: One can elimi-
nate the contaminations through the intentional attention to sāmānya by which all exis-
tents are realized to be the lack of self-nature. By ‘intentional attention to sāmānya,’ it
means that one applies the intentional attention to the universals of all existents”. (“問:
何等作意能斷耶?答:總緣作意觀一切法皆無我性,能斷煩惱。總緣作意
者,謂令緣一切法共相行作意”).
11 T.43.1830: 584b29-c5: “由此義故,對法等說緣共相智能斷惑者,依分別心 於一種
類,真如之上,通在諸法,說名共相。或真如體諸法皆有義,名共相。緣此
之智,名共相智。論實,真如法實性故,非是共相。據實而言,即別相智能
斷惑也。共相假立,已如前辨”.
380 Chen-kuo Lin
Kuiji distinguishes two types of universal: (1) concepts, such as class and spe-
cies, which are superimposed upon the real existents by the conceptual mind;
(2) the meaning/property of tathatā, which also describes sixteen modes
(ākāras) of the four noble truths. Both types of universal are the cognitive
object of inference. In the case of tathatā, tathatā in itself is viewed as sub-
stance, while “suffering,” “impermanence,” and so on, are viewed as property/
meaning (artha). For Kuiji, tathatā itself cannot be conceptually cognized.
Hence it cannot be taken as universal. In short, as regards the issue whether the
cognition of universal or particular leads to the elimination of afflictions, Kuiji
clearly disagrees with the Abhidharma view. The main reason for his disagree-
ment is that the Abhidharma confuses the two aspects of tathatā: tathatā in
itself (ti 體, *svarūpa) and tathatā as property (yi 義, artha). Tathatā in itself is
the ultimate reality which can be experienced in direct perception. However,
one has to cognize “tathatā as property” before one directly perceives tathatā
in itself which leads to the final enlightenment.
In the following I will examine how the Chinese Yogācāra Buddhists during
the seventh and the eighth centuries are indebted to the Abhidharma theory of
svalakṣaṇa and sāmānyalakṣaṇa, even though their ontological positions are
different.12 I will also point out that the continuity between Abhidharma and
Yogācāra is grounded in the same matrix of cultivation and theorization. For
both Abhidharma and Yogācāra, the ontology of particular and universal can-
not obtain its full meaning unless it is understood within the context of
meditative practice. In the following section, I will first examine the crucial
passages in Kuiji’s Cheng weishi lun shuji to see how critically the Chinese
Yogācāra scholars adapted the Abhidharma doctrine to a more complex
framework.
It should be noted that the above three theories cited from Kuiji’s Cheng weishi
lun shuji appear in the commentary on the path of cultivation. Adopting the
Abhidharma model, which is believed to have been accepted and practiced in
the entire community of Buddhists, the path of cultivation in the Cheng weishi
12 Part of materials in this paper has been presented in another paper, “How to attain
enlightenment through cognition of particulars and universals – Huizhao on svalakṣaṇa
and sāmānyalakṣaṇa,” at the International Workshop on “The Ontology of Asian Philoso-
phy – Perspectives from Buddhist Studies and Analytic Philosophy”, 13–14 April, 2013, Ryu-
koku University and Kyoto University, Kyoto.
Svalakṣaṇa (Particular) and Sāmānyalakṣaṇa (Universal) 381
lun is divided into five stages: (1) the stage of equipment (sambhāra, ziliang wei
資糧位), (2) the stage of application (prayoga, jiaxing wei 加行位), (3) the stage
of penetration (nirvedha, tongda 通達), which is also called “the stage of seeing
(darśana, jiandao 見道)” in the Abhidharma literature, (4) the stage of cultiva-
tion (bhāvanā, xiuxi wei 修習位), and (5) the stage of completion (niṣṭhā,
jiujing wei 究竟位).13 In each stage, several sub-stages of cultivation are also
listed. Roughly speaking, the whole system of cultivation is designed to achieve
the final liberation by eliminating psychological and cognitive hindrances,
including the contamination-hindrances (kleśāvaraṇa, fannao zhang 煩惱障)
and the hindrances of cognitive objects (jñeyāvaraṇa, suo zhi zhang 所知障).14
According to the Yogācāra doctrine, the hindrance which should be gradually
eliminated includes those in forms of mental phenomena (samudācāra, xian
xing 現行 (lit., “presentation” – “present manner/feature”), seeds (bīja, zhongzi
種子), and habitual residues (vāsanā, xiqi 習氣). The mental phenomena are
the form that should be tamed first. However, a more crucial step should be
further taken to eliminate the seeds and habitual forces which are contained in
the store-consciousness. For the Yogācāra, the elimination of the contamina-
tion-hindrances leads to the realization of nirvāṇa, while the elimination of
cognitive hindrances leads to the realization of enlightenment (bodhi).
Although a long course of cultivation is required to eliminate the contamina-
tion-hindrances, the elimination of cognitive hindrance is rather considered
more foundational. It is precisely in the context of eliminating the cognitive
hindrances that the cognition of particular and universal is brought into the
agenda of practice, for both particular and universal are the aspects of objects
which should be correctly cognized.
In the Cheng weishi lun, the first four stages of practice are further divided
into the mundane path and the sagely path. The mundane path consists of
the first two stages, the stage of equipment and the stage of application, while
the sagely path consists of the latter two stages, the stage of penetration and
the stage of cultivation. Since the entire system of cultivation is rather com-
plex, it is not practical to lay it out in excruciating details. At this juncture we
merely need to point out that, as for the discernment of particular and univer-
sal, the practice at the stage of application is most crucial before one attains
the insight of emptiness at the stage of penetration. The stage of application is
further divided into four sub-stages: “warmth” (ūṣmagata, meaning lit., “gone
13 For a brief account of the Abhidharma path, see Dhammajoti, 2007: 564–612.
14 Cf., Muller, 2004: 207–235. A brief version of interpretation can be viewed at the entry
“two hindrances” by the same author at Digital Dictionary of Buddhism (<http://www.
buddhism-dict.net>).
382 Chen-kuo Lin
Commenting on the path of seeing, where two hindrances are said to be elimi-
nated through discerning the sixteen modes or ākāras of the four noble truths,
Kuiji followed the Abhidharma path in pursuing the issue of the cognition of
particular and universal. The questions are raised again: In the path of cultiva-
tion, when one meditates on four noble truths as the object of cognition
(ālambana), should he cognize four noble truths as a whole? Or, should he
cognize four noble truths individually in each object? Which way of discerning
is correct practice? If the four noble truths are cognized as a whole, then the
object of cognition should be none other than the universal. Here the question
arises again: Is the cognition of universal capable of eliminating the cognitive
hindrance? Since the cognition of universal relies on inference (anumāṇa),
which is conceptual activity (vikalpa), the same question can be posed as such:
Can one attain enlightenment through inference? The question can be also
stated conversely: Does one realize enlightenment through direct perception
(pratyakṣa)?20 These questions are considered by Chinese Yogācāra scholars to
be essential in their epistemological and soteriological project.
20 T.43.1830: 584a29-b4: “Question: As to the term ‘elimination [of affliction],’ is the cogni-
tion (jñāna) which takes the whole [of five aggregates, four noble truths, etc.] as a sup-
port-object (samastālambana) capable of eliminating [affliction]? Or, is the cognition
which takes the individual [item of five aggregates, four noble truths, etc.] as a support-
object (vyavacchinnālambana) capable of eliminating [affliction]? What is wrong with
either [theory]? [Answer:] The cognition that takes the whole as a support-object is not
the cognition of particular characteristics (svalakṣaṇa). How could the cognition of infer-
ence, which takes sāmānyalakṣaṇa [as the support-object], be capable of eliminating
affliction? If one holds that the cognition of the individual support-object is capable of
elimination, this will contradict the statements in fascicle fifty-nine and the theory in
fascicle seven of the Abhidharma [treatise]”. (“問:此言斷者,為總緣智能斷?為別
緣智能斷?此有何失?總緣之智非自相智,如何共相比量之智能斷惑耶?若
別相智能斷,即違對法五十九等文,對法七等說”).
Svalakṣaṇa (Particular) and Sāmānyalakṣaṇa (Universal) 385
[Question:] How should one know [the difference between] these two
types of understanding? Answer: When the various entities are placed
close to the blue gem, their color looks like that [gem] without manifest-
ing their own particular property. Thus should be known is the
understanding of discerning the universal. When the various entities are
placed in distance from the blue gem, the colors of blue, yellow, and so
on, are manifest individually. Thus should be known is the understanding
of discerning the particular. Furthermore, one should know that the
understanding of discerning the universal is like the shining everywhere
of the rising sun: all shadows are dispelled immediately. One should also
know that the understanding which discerns the particular is like the sun
that has arisen, gradually shining on the various things: the gaps between
walls and the valleys between mountains are all manifest. Furthermore,
one should know that the understanding of discerning the universal is
like someone who enters a dark room with a lamp, the darkness of all
corners is broken in an instant. After entering the room, [the light] gradu-
ally shines towards jars, clothes, utensils, baskets, etc. Thus should be
known is the understanding which cognizes the particular. Furthermore,
the understanding of discerning the universal is like a mirror that does
not make the particular property visible from afar. When the mirror is
placed close to the object, the individual property is revealed. Thus
should be known is the understanding of the particular. Furthermore,
one should know that the understanding of discerning the universal is
like someone who watches a mountain and forest from afar. One should
also know that the understanding of discerning the particular is like
someone who closely watches a mountain and forest.24
also adopted the Buddhist logician’s theory of cognition. In the Cheng weishi
lun shuji, Kuiji defined “particular” by referring it to “the existent itself which
is only known by direct cognition, but not designated by verbal expression”.
He also defined “universal” by referring it to “the nature of existent which can
be both designated by the verbal expression and cognized by the concep-
tual mind”.26 This definition is adopted from Dignāga’s epistemology, taking
particular and universal as the objects known through direct perception and
inference respectively.27 Following Dignāga’s ontology, Kuiji and Huizhao 慧
沼 also adopted his theory of apoha. In this regard, we have reason to assume
that their knowledge of apoha must come from Xuanzang’s 玄奘 oral teach-
ing, because Xuanzang did not translate Dignāga’s Pramāṇasamuccaya into
Chinese.
Now, let us examine some other textual sources. In the Cheng weishi lun
shuji, Kuiji continued to examine the ontological status of universal:
火等。此義即通一切火上。故言共相得其義也。非苦空等之共相理”.
29 Cf. Katsura, 2014: 99–118.
30 For the Yogācāra, the meaning (artha) of verbal designations is the conceptual construc-
tion (parikalpita) superimposed upon the mental phenomena (paratantra); the latter
arises depending on the cause and conditions. Ontologically speaking, while parikalpita
is not real at all, paratantra is real in the sense of phenomena.
31 Huizhao, “Erliang zhang” 二量章 in Dasheng fayuan yilinzhang buque 大乘法苑義林章
補闕, X55, No. 0882. For the study of this text, cf. Lin, 2014 (forthcoming): “How to attain
enlightenment through cognition of particulars and universals? – Huizhao on svalakṣaṇa
and sāmānyalakṣaṇa”.
390 Chen-kuo Lin
image serves as the content of cognition, which is also termed “the direct
object of cognition” (qin suoyuanyuan 親所緣緣), while the noema qua basic
stuff is named “indirect object of cognition” (shu suoyuanyuan 疏所緣緣),
which cannot be directly known by the seeing part of mind (noesis). As
Huizhao explains in The Lamp for Illuminating the Definite Meaning of Cheng
Weishi lun (Cheng weishi lun liaoyi deng 成唯識論了義燈)32, universal refers to
the conceptual construct which is created and superimposed by the concep-
tual mind upon the “noema qua original stuff”. This type of conceptual
construct is not real in itself, for it is parikalpita (conceptually-constructed) in
view of the theory of threefold nature (trisvabhāva). Direct perception in the
structure of three or four parts (*bhāga), including the seeing part (noesis) and
the image part (noema), however, is real in the sense of paratantra (phenom-
ena as causally arising), while universal is not real in the sense of parikalpita.
Hence, universal as prajñapti or parikalpita is said to be lack of substance,
therefore not real.
In the Treatise on Two Pramāṇas, Huizhao’s explanation is somehow differ-
ent.33 He says that universal can still be qualified as something real, if it is
viewed in terms of the “noema qua image” as the direct object of cognition.
Nevertheless, it is void of substance, if it is viewed in terms of “original stuff”
(bimba). This explanation seems to be used for explaining the cognition of
impermanence, suffering, etc., in the concentrated state of meditation, for
impermanence and suffering are vividly and intuitively perceived as the direct
object of experience, which therefore cannot be said to be not real at all. If
impermanence, suffering, etc., can be taken as universal, as the Abhidharma
holds, then they cannot be something entirely unreal. Again, the context of
meditation is taken into account with regard to the ontological issue of
universal.
To sum up, if Huizhao exclusively takes a nominalist position, he will be
forced to answer another difficult question: How is it possible for the universal
to be cognized when it is not real? If he takes a conceptualist position, on the
contrary, he will also have difficulty to explain Dignāga’s theory of anyāpoha,
which he learned in the Xuanzang School.34 According to the theory of apoha,
the universals can be known only through the exclusion of the rest of the
members in the same class, e.g. “blue” means the exclusion of the other colors
that are not blue. Nevertheless, due to the ontological commitment to the
6 Concluding Summary
In the following, some tentative remarks are concluded from the above
expositions:
Abbreviation
Bibliography
Cox, Collett (2004). “From Category to Ontology: The Changing Role of Dharma in
Sarvāstivāda Abhidharma”, Journal of Indian Philosophy 32, pp. 543–597.
Cox, Collett (1992). “Attainment through Abandonment: The Sarvāstivādin Path of
Removing Defilements”, In Paths to Liberation: The Mārga and its Transformations in
394 Chen-kuo Lin
Buddhist Thought. Edited by Robert E. Buswell Jr. and Robert M. Gimello. Honolulu
HI: University of Hawaii Press, pp. 63–105.
Dhammajoti Bhikkhu, KL (2007). Sarvāstivāda Abhidharma. Hong Kong: Centre of
Buddhist Studies, The University of Hong Kong.
Huizhao 慧沼, Dasheng fayuan yilin zhang buque 大乘法苑義林章補闕. X55.882.
Katsura, Shōryū (2011). “From Abhidharma to Dharmakīrti – With a special reference
to the concept of svabhāva”, In Religion and Logic in Buddhist Philosophical Analysis.
Proceedings of the Fourth International Dharmakīrti Conference (Vienna, August
23–27, 2005) (Denkschriften der phil.-hist. Klasse 424. Beiträge zur Kultur- und
Geistesgeschichte Asiens 69). Edited by Helmut Krasser, Horst Lasic, Eli Franco and
Birgit Kellner. Wien: Verlag der Östereichischen Akademie der Wissenschaften,
pp. 271–279.
Katsura, Shōryū (2014). “The theory of apoha in Kuiji’s Cheng weishi lun shuji”, In
A Distant Mirror: Articulating Indic Ideas in Sixth and Seventh Century China. Edited
by Chen-kuo Lin and Michael. Radich. Hamburg: University of Hamburg Press,
pp. 99–118.
Kuiji 窺基, Cheng weishi lun biechao 成唯識論別抄. X48.808.
Lin, Chen-kuo (forthcoming). “How to attain enlightenment through cognition of par-
ticulars and universals? – Huizhao on svalakṣaṇa and sāmānyalakṣaṇa”, In Dharma
and Dao: Chinese Buddhist Philosophy. Edited by Sandra Wawrytko and Wang Youru.
Springer.
Muller, A. Charles (2004). “The Yogācāra two Hindrances and their Reinterpretations in
East Asia”, Journal of the International Association of Buddhist Studies 27,
pp. 207–235.
Saito, Akira (ed.). Bauddha Kośa, <http://www.l.u-tokyo.ac.jp/~b_kosha/html/index_75
dharma.html#>.
Sokolowski, Robert (2000). Introduction to Phenomenology. Cambridge: Cambridge
University Press.
Sponberg, Alan (1986). “Meditation in Fa-hsiang Buddhism”, In Traditions of Meditation
in Chinese Buddhism. Edited by Peter Gregory. Honolulu HI: University of Hawaii
Press, pp. 15–43.
Takakusu, Junjirō 高楠順次郎, Watanabe, Kaigyoku 渡邊海旭, and Ono Gemmyō 小野
玄妙 (ed.) (1924–1934). Taishō shinshū daizōkyō 大正新脩大藏經, Tokyo: Taishō
issaikyō kankōkai.
- T.26.1530, Bandhuprabha et al., Buddhabhūmyupadeśa, Fodijing lun 佛地經論,
Xuanzang 玄奘.
- T.27.1545: 500 arhats, [Abhidharma]mahāvibhāṣā[śāstra], Apidamo da piposha lun
阿毘達磨大毘婆沙論, Xuanzang 玄奘.
- T.31.1585: Dharmapāla, et al., Vijn͂ aptimātratāsiddhi[śāstra], Cheng weishi lun 成唯
識論, Xuanzang 玄奘.
Svalakṣaṇa (Particular) and Sāmānyalakṣaṇa (Universal) 395
Chapter 13
1 Introduction
1 Hakamaya discussed the Indic origin of this term and suggested “dharmoddāna” rather than
the often assumed one, “dharmamudrā” (Hakamaya, 1979: 60–81).
2 For the rendering of the term anātman / anattā as “not-self,” see Chapter 2 “Impermanence,
Not-self and Suffering” in David Burton, 2004: 11–30; Chapter 6 “An Interpretation of the Not-
Self Doctrine” in Gowans, 2003: 63–75; Chapter 1 “Setting the Scene” in Hamilton, 2000: 18–32;
Tsai, 2006: 145–147; Tsai, 2013: 121–122.
various interpretations and thoughts. For example, the Indian Buddhist phi-
losopher Vasubandhu (ca. fourth/fifth century CE), offers just such an
interpretation in the ninth chapter, entitled “Refutation of the Doctrine of the
Self” (Ātmavāda-pratiṣedha) or “An Examination of the Person (or Individual)”
(Pudgala-viniścāya), which he added to his masterwork of Abhidharma stud-
ies: the Commentary on the Treasury of the Abhidharma (Abhidharmakośabhāṣ
ya). In the chapter “Refutation of the Doctrine of the Self” (henceforth abbrevi-
ated as “Refutation”), Vasubandhu briefly presents his perspectives on the
person and the self, raises objections to the viewpoint of the Pudgalavādins (or
Vātsīputrīyas, i.e. followers of Vātsīputra), replies to the Pudgalavādins’ objec-
tions to his own view, and then replies to the objections put forward by the
Tīrthikas (Forders), i.e. non-Buddhist thinkers of India, specifically the Nyāya-
Vaiśeṣika tradition.
Vasubandhu’s “Refutation,” together with his other works, has been a subject
of veneration and disputation across the Buddhist traditions, and has attracted
much attention in contemporary philosophical circles. Numerous scholarly
works have been written on the “Refutation,” especially James Duerlinger’s
Indian Buddhist Theories of Persons, a translation mainly based on the Sanskrit
text of the Abhidharmakośabhāṣya, which provides a comprehensive intro-
duction and analytical commentary.3 Yet, even with such an abundance of
commentaries, there are still several critical points that bear scrutiny to clarify
both what is intended by the concept that all factors are not-self, and what
“Refutation” leads to.
In what follows, I will demonstrate that the doctrine of not-self is about
metaphysical identification, i.e. identification in terms of the nature of reality,
rather than existential differentiation, i.e. differentiation between existence
and non-existence. To that end, I elucidate the person-concept (individual-
concept) and the self-concept, two key concepts that have played important
roles in Buddhist traditions, and adjust perspectives on not-person and not-
self in a way that is conducive to gaining insight into the sentient world.
Moreover, I offer three critiques of Abhidharmic oriented perspectives formu-
lated from an understanding of the “not-dharmic” oriented perspectives, and
conclude by proposing a shift from “Abhidharmic” to “not-dharmic” perspec-
tives when engaging in philosophical inquiry into Buddhist teachings.
4 See the Saṃyuktāgama nos. 23–24 at T.2.99: 5a11-b 27; nos. 30–34 at T.2.99: 6a24–8a4; no. 43
at T.2.99: 10c19–11a12; nos. 62–64 at T.2.99: 16a19–17a22; no. 76 at T.2.99: 19c12–24; nos. 82–87 at
T.2.99: 21b14–22b13; nos. 104–105 at T.2.99: 30c12–32c1; nos. 109–110 at T.2.99: 34a24–37c7; nos.
120–121 at T.2.99: 39b25–40a3; no. 125 at T.2.99: 40c6–27; Bhikkhu Bodhi, 2000: 616, 866, 888,
902, 927, 948, 969, 979, 992, 1003, 1135, 1166, 1389, …; Bhikkhu Ñāṇamoli and Bodhi, 2005: 123,
229–232, 279–282, 327–330, 527–529, 890, 1089–1091, 1122, etc.
5 See also Bhikkhu Anālayo, 2010: 49–50; Bodhesako, 2008: 79; Choong, 2000: 58–59; Wynne,
2007: 35, 140.
6 Example one: “rūpaṃ bhikṣavo ’nityaṃ. yad anityaṃ, tad duḥkhaṃ. yad duḥkhaṃ, tad anātmā.
yad anātmā, tan naitan mama, naiṣo ’ham asmi, naiṣa me ātmêty. evam etad yathābhūtaṃ
samyakprajñayā draṣṭavyam. vedanā saṃjñā saṃskārā vijñānam anityaṃ. …” (Chung, 2008:
Perspectives on the Person and the Self 399
The procedure for examining the issue of self/not-self can thus be analyzed
and differentiated into the following four steps.
First, one is required to observe the five aggregates in the flow of life. Without
the practice of observing the five aggregates, one is likely to be hindered from
attaining a firsthand understanding of the issue of self/not-self, and will, there-
fore, be confined within the sphere of beliefs, opinions, and ideologies.
However, to observe the five aggregates does not presuppose that there exists a
person or a self which is composed of the five aggregates. More to the point,
what is taken for granted as constituting the person or the self is to be rigor-
ously examined, or critically identified, but not as a given presupposition. The
five aggregates are brought on stage because they encompass existence in the
sentient world and are available to perception. What makes Buddhist observa-
tion special is that it does not fixate on the five aggregates but conforms to
their constant flux.
Second, actually observing the five aggregates in the flow of life affords a
vantage point from which the issue of the person or the self can be better
examined. Although the five aggregates can be perceived and observed, the
person or the self cannot be perceived or observed because the latter is located
at a level slightly remote from that of the perceptible and observable. Generally,
people assume the person or the self without pertinent observation or reason-
ing. In other words, the person or the self is unquestionably assumed or
projected in conventional usage. However, Buddhism offers a strategy for criti-
cally identifying the person or the self before any serious measures can be
taken beyond the level of conventional usage.
Third, after observing the five aggregates in the flow of life, it becomes clear
that none of the five aggregates can be identified as the person or the self.
Different philosophers or thinkers may have different ways in which they
define the person or the self. Various definitions come from different perspec-
tives and experiences, express dissimilar opinions, and then invite argument
and controversy. However, the task of identification, especially metaphysical
311–312, 323–326, 331–334). Example two: tathā hi rūpaṃ nâtmā, rūpavān nâpi câtmā, rūpe
nâtmā, nâtmani rūpam. evaṃ yāvat vijñānaṃ nâtmā, vijñānavān nâtmā, vijñāne nâtmā, nât-
mani vijñānam iti. tathā a[n]ātmānaḥ sarva-dharmā iti. (Yoshimizu, 2007: 20). (<http://www.
tulips.tsukuba.ac.jp/limedio/dlam/B26/B2674017/1.pdf>). Example three: see Harivarman’s
Satyasiddhiśāstra translated by Kumārajīva: T.32.1646: 314b16–18: “如長老差摩伽說。不說
色是我。不說受想行識是我。但五陰中有我慢、我欲、我使。未斷未盡是名我
慢”. Cf. yathā sthaviraḥ kṣemaka āha: ‘na khalv āyuṣman rūpam asmîti vadāmi, na vedanā, na
saṃjñā, na saṃskārā, na vijñānam, [nâpy anyatra vijñānād asmîti vadāmi]. api ca ma āyuṣman
pañcasûpādāna-skandheṣu anusahagato ’smîti mānaḥ, asmîti cchandaḥ, asmîty anuśayo,
’samuddhataḥ. [ity ādi]. ayam asmi-māna ity ucyate. (Sastri, 1975: 307; 1978: 285–286).
400 Yao-ming Tsai
present), this comes to be; from the arising of that, this arises. When that does
not exist (or is not present), this does not come to be; from the cessation of
that, this ceases”.10 From such a perspective, synchronic differentiation
between existence and non-existence can be avoided by observing the dia-
chronic sequence of both dependent origination (pratītyasamutpāda) and
cessation (nirodha).11
In sum, before engaging in discourse on the self, one should examine
whether any observable factors can be identified as the self. Since none of the
observable factors are qualified to be identified as the self, the label of not-self
is used to indicate the non-identifiability (or non-identity) of the self. This
insightful understanding of not-self does not oppose the inclusion of the self-
concept in verbal communication or teaching as long as the self-concept
operates by and within linguistic convention. Furthermore, views that posit or
presuppose the existence and non-existence of the self are equally problem-
atic. The postulation of the existence (astitva) of the self is labeled as the view
of eternalism (śāśvatadṛṣṭi), implying that the self is regarded as eternal and
unchanging. The postulation of the non-existence (nāstitva) of the self is
labeled as the view of annihilationism (ucchedadṛṣṭi), which asserts that the
self is cut off and utterly destroyed at death. Both of these two extremes not
only misrepresent the reality of the dynamic continuity of mental process but
also shut the door to liberation, and therefore are to be discarded.12
10 asmin sati, idaṃ bhavaty; asyôtpādād, idam utpadyate. asminn asati, idaṃ na bhavaty;
asya nirodhād, idaṃ nirudhyate. imasmiṃ sati, idaṃ hoti; imass’ uppādā, idaṃ uppajjati.
imasmiṃ asati, idaṃ na hoti; imassa nirodhā, idaṃ nirujjhati. “此有故彼有,此生故彼
生;此無故彼無,此滅故彼滅”. See Vaidya, 1958: 221.
11 See the Saṃyuktāgama no. 262 at T.2.99: 66b6–67a21; Bodhi, 2009: 946–7.
12 See the Saṃyuktāgama no. 961 at T.2.99: 245b9–25; Bodhi, 2009: 1393–4. See also Vasu-
bandhu’s Apidamo jushe shilun 阿毘達摩俱舍釋論 translated by Paramārtha, T.29.1559:
307b; Vasubandhu’s Apidamo jushe lun 阿毘達摩俱舍論 translated by Xuanzang,
T.29.1558: 156a; Duerlinger, 2003: 90–91.
402 Yao-ming Tsai
However, different schools and authors have held divergent views about
these concepts, in large part because of an incomplete knowledge of the perti-
nent concepts and different perspectives on the sentient world. Consequently,
the objective of this section is to present a philosophical elucidation of the
person-concept (individual-concept) and the self-concept, rather than engag-
ing in a history of thoughts. Therefore, the focus of this section will be on
clarifying the connotations of these two concepts in explaining how the sen-
tient world works.
The person-concept (individual-concept) considered in this chapter is a
partial equivalent of the Sanskrit word pudgala (or puggala in Pāli). As the
term pudgala is difficult to translate into English, an exposition of its three
connotations follows below.
First, in its connotation as individual, the term “pudgala is traditionally
said to be derived from puṃ- (joining) plus -gala (breaking)”.13 Hence, etymo-
logically, the term pudgala means the individual or discrete individual that
undergoes modifications by combination and dissection. As for the rendering
as separate individual, the term pudgala may well be related to pṛthak (dif-
ferent; separate; individual), and could therefore be related to pṛthagjana
(ordinary beings; still vulnerable to being subverted by constantly-varying
affections and aversions).14 In short, etymologically speaking, the constant
process of integration and disintegration is characteristic of individuals in the
world.
Second, pudgala is rendered as a person. If the focus is on human beings
rather than on physical objects or other manifestations of sentient beings,
then the literal meaning of the term pudgala can shift from the individual to
the person or a single human. In Sanskrit, several other terms such as puruṣa,
manuṣya, nara, or jana can be used to mean human being in general and per-
son in particular. However, the term pudgala is probably more salient in
conveying the connotation of a person separate from other persons by provid-
ing an idea of being different or separate. Such a connotation can further
elevate the term pudgala’s individual subjectivity and capacity to be situated
as an objectified entity.
Third, pudgala is rendered as a continuous stream.15 If the focus is on the
process of integration and disintegration rather than on individual or person,
then the term pudgala is called upon to act as continuous stream, i.e. the bun-
dle of tendencies that keeps reincarnating as an individual or a person in a
examining perspectives on the person and the self in the ninth chapter of the
Abhidharmakośa.
For example, both Vasubandhu and the Pudgalavādins can agree that pud-
gala is not the self.17 However, the ontological status of pudgala is itself
disputed. According to Vasubandhu, pudgala exists by reason of sharing a sim-
ilar existential status as collections of the five aggregates. By contrast, the
Pudgalavādins are portrayed as denying that pudgala shares a similar status of
existence as collections of the five aggregates. In particular, the Pudgalavādins
insist that pudgala exists by reason of being neither the same as nor different
from collections of the five aggregates, and therefore its existence is inexplica-
ble (avaktavya).
With the above overview in mind, three critiques can be formulated as to
the perspectives therein.
Critique one: Should pudgala be regarded as an entity possessing existence
as an intrinsic part of its nature? Just because the concept pudgala is a con-
cept, does not mean that pudgala is necessarily something like an entity right
over there. However, it seems that if one adopts the Abhidharmic orientation
of perspectives, taking phenomena as a starting point, and if one deems that
the referents of a great number of concepts – a mountain, a desk, a cat, etc. –
exist, since pudgala is also a concept, then it appears natural that the referent
of the concept pudgala should also exist. Surprisingly, it is in the same vein
that both Vasubandhu and the Pudgalavādins insist on the existence of pudga-
la.18 But let us briefly ponder this problematic view of existence. When one
says that a certain entity exists, one seems to assume that that very entity must
exist in spatio-temporal processes of becoming. If pudgala is taken in its con-
notation as continuous stream, and if pudgala exists, then how and in what
framework does pudgala exist? The existence of pudgala may be challenged
from three related levels. First, if the so-called existent happens to be an entity
and is framed by the spatio-temporal continuum, how can such a spatio-tem-
poral continuum as pudgala exist as an existent in a lower level? In other
words, as a continuous stream, pudgala appears as a continuous stream; but
this continuous stream cannot be downgraded to some level lower than con-
tinuous stream. If it were downgraded to an existent at a lower level it would
have lost its being as a continuous stream. Second, how can such a spatio-tem-
poral continuum as pudgala exist as spatio-temporal continuum at the same
level? In other words, as continuous stream, pudgala cannot be both continu-
ous stream and existent at the same level on account of the incompatibility of
continuous stream permeated by at least four-dimensional changes and exis-
tence remaining unchanged as permanent existence. Third, as continuous
17 Or translated as “a person is not a self”: see Duerlinger, 2003:124, 127, 199, 201–202, 280.
18 See Duerlinger, 2003: 2, 14, 24, 26, 35, 73, 131, 284.
Perspectives on the Person and the Self 407
stream, pudgala can only exist in a framework which is not only at a higher
level than pudgala but also encompasses pudgala. Since such a higher frame-
work by which pudgala can be accounted for is unknown to both Vasubandhu
and the Pudgalavādins, it will be pointless for them to say that pudgala exists.
Critique two: Should the assertion of existence/non-existence be applied to
a single item? Just like most of those who follow the Abhidharmic orientation
of perspectives, both Vasubandhu and the Pudgalavādins are inclined to apply
the assertion of existence/non-existence to a single item such as pudgala. Such
a convention is very common among ordinary people and is fully in accor-
dance with standard linguistic expressions. Nevertheless, it is one thing to say
that a single item exists, but quite another thing to examine if this type of
assertion is pertinent to more facets of reality. It is not that the assertion of
existence/non-existence is necessarily wrong, but the assertion is supposed to
reveal a more pertinent perspective on the reality it purports to represent,
rather than being confined within the walls of linguistic isolation and laby-
rinth. From the perspective of dependent origination (pratītyasamutpāda), an
item does not stand alone by itself, but arises from the condition of a priori and
surrounding context. If the existence/non-existence is asserted, it is better to
address the link between a certain item and its context.19 Obviously both
Vasubandhu and the Pudgalavādins fail to take dependent origination into
consideration when asserting the existence of pudgala, and, consequently, fail
to make it clear how pudgala can arise and cease. Without understanding the
mechanism of arising and cessation, the assertion of the existence of pudgala
begins with an attachment to a single conceptual item, and may end up with a
view of eternalism (śāśvatadṛṣṭi).
Critique three: How should the two concepts of pudgala and self be com-
bined to make assertions germane to the sentient world? Although both
Vasubandhu and the Pudgalavādins hold the same view that pudgala is not the
self, their divergent presuppositions, which on the surface appear to comply
with the thesis of not-self, definitely deserve further inspection. It is important
to know that both concepts of pudgala and self are not at the level that can be
perceived or observed. Rather, both concepts are convenient tools for verbal
communication and philosophical discourse. However, if philosophical dis-
courses merely remain at the level of conceptual items formulated in partisan
terms, such discussions will only continue to be entangled in seemingly unsolv-
able disputes. A feasible and better approach is to start by perceiving or
19 The general formula of this perspective runs thus: “When that exists (or is present), this
comes to be; from the arising of that, this arises. When that does not exist (or is not pres-
ent), this does not come to be; from the cessation of that, this ceases”.
408 Yao-ming Tsai
observing such factors as the five aggregates in the flow of life, and gaining
insight into the reality of the five aggregates, and explicating the relation
between the five aggregates, pudgala, and self.20
In order for the aforementioned approach to work effectively, the following
six steps should be implemented. First, perceiving or observing the constant
flux of the five aggregates may provide an insight into the reality that the five
aggregates do not exist as an independent or separate entity. In other words,
the five aggregates are empty (śūnya) of any independent or separate existence
or beingness (bhāvatā; vastutā) of their own. Second, due to fact of the imper-
manence and suffering, none of the five aggregates are qualified to be identified
as the self, whatever the self-concept may mean. In other words, the five aggre-
gates, either altogether or separately, are not the self. Third, if the phenomena
of the five aggregates are taken as individual, person or continuous stream,
then the term pudgala can be employed accordingly. Just as many other terms
can be used as media of indication, connotation or communication, so can the
term pudgala as long as it makes some sense for a given application. Fourth, no
matter what the term pudgala may mean in different contexts, as one of the
alternatives of the five aggregates in linguistic convention, the term pudgala
merely acts as a verbal designation. In reality, just as the five aggregates are
empty of the inherent existence of the five aggregates, so is the designated pud-
gala empty of the inherent existence of pudgala. Fifth, just as the five aggregates
are not the self, pudgala cannot be identified as the self either. Sixth, the rela-
tion between the five aggregates, pudgala, and self can therefore be explicated
simply and concisely as follows.
The five aggregates are at the level that can be and should be perceived or
observed. Although the five aggregates are impermanent and suffering, and in
reality are not qualified to be identified as the self, they are often individually
or collectively regarded as the self in linguistic convention and frequently
encountered in philosophical views as such. Although the five aggregates can
also be alternatively termed as pudgala to match such cognitive impressions as
individual, person or continuous stream, it is unnecessary, or even incorrect to
assert the independent existence of pudgala in itself. Therefore, both concepts
of pudgala and self can be switched to the five aggregates, which can be per-
ceived and understood as empty of such dharmas as the five aggregates. Since
the five aggregates are empty of said dharmas as the five aggregates, the so-
called five aggregates are in reality not the five aggregates. In light of this
20 Cf. bhagavān āha: tat kiṃ manyase subhūte atraiṣā saṃjñā samajñā-prajñapti-vyavahāraḥ
pañcasûpādāna-skandheṣu yad uta bodhisattva iti? (Wogihara, 1932: 71). See also Conze,
1975: 88.
Perspectives on the Person and the Self 409
Abbreviation
Bibliography
Kattackal, Jacob (1999). Indian Philosophy: Indian Religious Philosophies from Vedas to
Śankara-Rāmānuja-Śaivasiddhānta. Kottayam: Oriental Institute of Religious
Studies.
Ñāṇamoli Bhikkhu and Bodhi Bhikkhu (2005). The Middle Length Discourses of the
Buddha: A New Translation of the Majjhima Nikāya. Boston MA: Wisdom.
Sastri, N. Aiyaswami (1975). Satyasiddhiśāstra of Harivarman. Vol 1. Baroda: Oriental
Institute, Maharaja Sayajirao University.
Sastri, N. Aiyaswami (1978). Satyasiddhiśāstra of Harivarman. Vol 2. Baroda: Oriental
Institute, Maharaja Sayajirao University.
Sutton, Florin (1991) Existence and Enlightenment in the Laṅkāvatāra-sūtra: A Study in
the Ontology and Epistemology of the Yogācāra school of Mahāyāna Buddhism. Albany
NY: State University of New York Press.
Takahashi, Hisao et al. (ed.) (2006). Vimalakīrtinirdeśa: A Sanskrit Edition based upon
the Manuscript Newly Found at the Potala Palace. Tokyo: Taishō University Press.
Takakusu, Junjirō 高楠順次郎, Watanabe, Kaigyoku 渡邊海旭, and Ono Gemmyō 小野
玄妙 (ed.) (1924–1934). Taishō shinshū daizōkyō 大正新脩大藏經, Tokyo: Taishō
issaikyō kankōkai.
- T.2.99: Saṃyuktāgama, Za ahan jing 雜阿含經, Guṇabhadra.
- T.29.1558: Vasubandhu, *Abhidharmakośabhāṣya, Apidamo jushe lun 阿毘達磨俱
舍論, Xuanzang 玄奘.
- T.29.1559: Vasubandhu, *Abhidharmakośabhāṣya, Apidamo jushe shilun 阿毘達磨
俱舍釋論, Paramārtha.
- T.32.1646: Harivarman, Tattvasiddhiśāstra, Chengshi lun 成實論, Kumārajīva.
Tedesco, Pablo A. (1947). “Sanskrit Pudgala: ‘Body; Soul’”, Journal of the American Oriental
Society 67/3, pp. 153–250.
Tsai, Yao-ming (2006). “Interpretative Approach to Navigating Concepts Concerning
the Doctrine of Nondual Middle Way”, National Taiwan University Philosophical
Review 32, pp. 115–166.
Tsai, Yao-ming (2013). “Human Life from the Perspective of Buddhist Medicine: A
Philosophical Inquiry Based on the Vimalakīrti-nirdeśa-sūtra”, In The Conception of
the Human Person in Medicine: Exploring Boundaries between Traditional Chinese and
Western Medicine. Edited by Lukas Kaelin, Tsai Yao-ming et al. Vienna: Verlag
Österreich, pp. 117–139.
Tsai, Yao-ming (2014). “Language As an Instrument of Soteriological Transformation
from the Madhyamaka Perspective”, Asian Philosophy: An International Journal of the
Philosophical Traditions of the East 24/4, pp. 330–345.
Gowans, Christopher (2003). Philosophy of the Buddha. London: Routledge.
Vaidya, Paraśurāma Lakṣmaṇa (ed.) (1958). Avadāna-śataka. (Buddhist Sanskrit Texts,
no. 19). Darbhanga: The Mithila Institute.
412 Yao-ming Tsai
Winchester, Jules (2009). “The Self Concept, Culture and Cultural Identity: An
Examination of the Verbal Expression of the Self Concept in an Intercultural Context”,
The Linguistics Journal: Special Edition – Language, Culture and Identity in Asia. British
Virgin Islands: Linguistics Journal Press, pp. 63–81.
Wogihara, Unrai. (ed.) (1932). Abhisamayālaṃkārālokā Prajñāpāramitāvyākhyā: The
Work of Haribhadra together with the Text Commented on. Tokyo: The Toyo Bunko.
Wynne, Alexander (2007). The Origin of Buddhist Meditation. New York NY: Routledge.
Yoshimizu, Chizuko (2007). Clarification of the Critic of Logic by the Madhyamaka School
in Indian and Tibetan Buddhism. Final Report of Thematic Project, University of
Tsukuba.
Index Index 413
Index
All terms are listed in Roman alphabetical order. When both the Sanskrit and the Pāli term oc-
cur, the Sanskrit term is listed first. Tibetan terms are listed under the first occurring Tibetan
radical letter. Chinese terms are listed according to the order of the pinyin transcription system.
English translations of Buddhist technical terms are indexed under the Sanskrit/Pāli term.
abhāva 354, 355, 356, 370 *Abhidharmahṛdaya (Apitan xin lun 阿毗曇
abhāvavādin 356, 360, 361, 367, 369, 370 心論 ) 93, 284, 302
Abhayagiri 176, 196 Abhidharmakośa (Apidamo jushe lun 阿毗達
abhedyaprasāda / avetyaprasāda / 磨俱舍論 ) 10, 13, 15, 19, 91, 93, 175, 176,
aveccapasāda (four) 7, 8 51, 54, 71, 77, 177,
78, 79, 80, 81, 82, 83, 84, 85, 87, 88, 89, 178, 180, 183, 194, 195, 284, 308, 315, 321, 322,
90, 91, 92, 93, 94, 95, 96, 98, 99 324, 327, 329, 334, 337, 338, 339, 343,
Abhidhamm’-atthasaṅgaha 191, 192, 196, 198, 345, 346, 364, 405
203, 205 Abhidharmakośabhāṣya 9, 15, 20, 95, 158,
Abhidhamm’-atthasaṅgaha Commentary 159, 175, 177, 183, 193, 215, 224, 225, 226,
191, 198, 203 227, 228, 231, 232, 236, 237, 238, 245, 253,
Abhidhammāvatāra 169, 172, 175, 178, 180, 257, 259, 353, 357, 370, 383, 396, 397
187, 190, 191, 199, 210, 218 Abhidharmakośaṭīkā Tattvārthā 227, 228,
Abhidhammāvatāraṭīkā 169, 170, 173, 174, 175, 239
178, 180, 187, 190, 192, 193, 199, 201, 206, Abhidharmamahāvibhāṣā (Apidamo da
207, 209, 213, 216, 218, 219 piposha lun 阿毘達磨大毘婆沙論)
Abhidharma / Abhidhamma ix, x, xii, xiii, 19, 223, 224, 231, 235, 236, 238, 240, 376,
xiv, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 386, 392
15, 16, 17, 18, 19, 20, 29, 30, 31, 32, 33, 34, Abhidharmapiṭaka / Abhidhammapiṭaka 1,
35, 36, 37, 38, 39, 41, 42, 43, 50, 51, 65, 85,
128, 140, 213
88, 90, 91, 94, 98, 99, 108, 119, 120, 128,
Abhidharmasamayapradīpikā 19, 225
129, 132, 135, 140, 141, 169, 170, 176, 177,
Abhidharmasamuccaya 70, 257, 270, 306,
178, 191, 202, 205, 208, 209, 212, 214, 223,
335, 345, 379
224, 225, 226, 227, 228, 229, 230, 231, 234,
Abhidharmasamuccayabhāṣyā 147, 153, 379
236, 237, 239, 240, 241, 243, 244, 245,
Abhidharmasamuccayavyākhyā 228, 306
248, 249, 269, 282, 283, 284, 285, 286,
abhijn͂ ā / abbhijhā 64, 62, 63, 72, 172, 217
287, 288, 289, 291, 298, 299, 300, 301,
302, 303, 304, 305, 306, 307, 308, 310, 311, abhisamaya 231, 386, 693
314, 315, 317, 318, 319, 321, 324, 326, 329, Abhisamayālaṃkāra 59, 60, 61, 70, 83, 84
343, 344, 345, 353, 354, 360, 361, 362, Abhisamayālaṃkārālokā 84
363, 370, 372, 376, 378, 379, 380, 381, 382, Abhisamayālaṃkāravŗtti 59, 60, 61
384, 385, 387, 391, 392, 397, 405, 409 ābhogatā 331
Abhidharma School 13, 282 284 absence of dispositions (see anabhisaṃskāra)
Abhidharmic ix, x, 1, 5, 7, 8, 9,12, 13, 15, 17, acknowledging the cognition of the principle
19, 20, 158, 160, 161, 162, 164, 224, 226, (see dharmajñānakṣānti)
288, 289, 298, 301, 305, 307, 309, 311, 315, acknowledging the cognition of the principle
345, 396, 397, 405, 406, 407, 409 in its subsequent realization (see
Ābhidharmika 15, 31, 38, 201, 223, 227, anvayadharmajñānakṣānti)
234, 291, 298, 299, 300, 301, 325, 356, acquisition (see prāpti)
364, 365, 392 activity (see kāritra)
Abhidharmadīpa 176, 178, 232, 353, 370 actual perception (see pravṛttivijñāna)
© Koninklijke
koninklijke Brill
brill NV,
nv, Leiden,
leiden, 2016 | doi 10.1163/9789004279278_012
10.1163/9789004318823_016
414 Index
adhipatipratyaya 160, 161, 324, 325 ālambanapratyaya 146, 150, 160, 239, 324,
aduḥkhāsukha 68, 334, 335 325, 330
Advaita Vedānta 37, 38 ālambanasaṃpramoṣa 331
aeon (see kalpa) ālayavijñāna 9, 19, 146, 147, 148, 149, 150, 151,
aeon of destruction (see saṃvartakalpa) 152, 154, 156, 157, 158, 160, 163, 164, 229,
aeon of formation (see vivartakalpa) 234, 242, 243
great aeon (see mahākalpa) Amituo jing 阿彌陀經 (see
incalculable aeon (see asaṃkhyeyakalpa) Sukhāvativyūhasūtra)
intermediate aeon (see antarakalpa) Amituo jing tongzang shu 阿彌陀經通贊疏
Afghanistan 29, 35, 48 309
Āgama 6 An Shigao 安世高 11, 248, 249, 250, 251,
aggregate (five) (see skandha) 252, 253, 254, 255, 256, 257, 258, 261,
agitation (see auddhatya) 262, 263, 264, 267, 268, 269, 270, 271,
Ahan koujie shi’er yinyuan jing 阿含口解十 272, 283, 285, 286, 301, 302, 303
二因緣經 251, 252,254, 255, 258, 259, An Xuan 安玄 285
260, 261, 266, 268 anabhisaṃskāra 331
ahrī 205 anābhoga 331
Ājīvika 3 anāgamya 332, 337, 362
ajjhatta- 113, 120, 125, 126, 127, 133 Anālayo Bhikkhu 136
Ānanda 3, 5, 126, 170, 211, 241
ajjhattā (dhammā) 120, 121, 122, 123, 125, 127,
ānantaryamārga 342
141
anapatrāpya / anapatrapa 205
ajjhattabahiddhā 7, 8, 17, 108, 109, 110, 111, 112,
anātmanaḥ 396
113, 114, 117, 118, 119, 120, 121, 123, 128, 129,
Anga-Magadha 36
130, 131, 132, 133, 134, 135, 136, 137, 138,
Aṅguttaranikāya 60, 71, 109, 110, 113, 115, 116,
139, 140, 141, 142
117, 126, 137, 261
ajjhattabahiddhā (dhammā) 114, 116, 120,
anityatā 223, 376, 378, 380, 393, 390, 391, 396,
122, 125, 127
398, 408, 409
ajjhattabahiddhārammaṇā (dhammā) An͂ n͂eyya-sattati-ṭīkā 169, 187, 189, 193
120, 122, 124, 125, 127, 128, 133, 135, 141 annihilationism (see ucchedadṛṣṭi)
ajjhattajjhatte 126 antarakalpa 326
ajjhattaṃ 7, 17, 108, 109, 110, 111, 112, 113, antonyms (pair ajjhattaṃ and bahiddhā) 111,
114, 115, 117, 118, 119, 121, 122, 123, 129, 131, 122, 139, 141
132, 134, 135, 136, 138, 139, 140, 141, 142 anumāṇa 136, 138, 377, 378, 380, 383, 384,
ajjhattārammaṇā (dhammā) 122, 124, 127, 388, 389, 391
128, 133 Anurādhapura 171
ajjhattika 115, 116, 120, 123, 126, 133, 140, Anuruddha 191, 192, 211, 219
217, 254 anuśaya 61, 76, 233, 341
ajjhattikabāhira 115, 117, 123, 134, 140 Anuṭīkā 170, 174, 213, 243
Akaniṣṭha 323 anvayadharmajñāna 342, 343
ākāra 173, 224, 240, 356, 380, 383, 384,385, anvayadharmajñānakṣānti 342
386, 392 ap 90, 176, 177, 183, 255, 287, 315
ākāśa 87 Apidamo da piposha lun 阿毘達磨大毘婆
ākāśānantayāyatana / ākāsānañcāyatana 沙論 (see Abhidharmamahāvibhāṣā)
124, 339 Apidamo xianzong lun 阿毗達磨顯宗論
akiṃcanyāyatana 339 (see *Abhidharmasamayapradīpikā)
akliṣṭājñāna 226, 241 Apidamo jushe lun 阿毗達磨俱舍論 (see
ālambana 239, 330, 384 Abhidharmakośa)
Index 415
Apitan ba jiandu lun 阿毗曇八犍度論 Atthasālinī 5, 119, 120, 121, 122, 123, 125, 126,
(see*Aṣṭagrantha / 127, 128, 134, 135, 141, 178, 185, 190, 196,
*Aṣṭaskandha(śāstra)) 201, 206, 209, 210, 212, 216, 244, 300
Apitan wufa xing jing 阿琵曇五法行經 283 Atthasamāsa 190, 191, 192
Apitan xin lun 阿毗曇心論 Atthuddhāra (see also Aṭṭhakathākaṇḍa) 123,
(see*Abhidharmahṛdaya) 127
apoha 388, 389, 391, 393 auddhatya 86, 211, 212, 213, 331
apramāṇa (four) 11, 52, 53, 54, 55, 57, 61, 62, aupacayika 242
64, 65, 67 69 avadāna 305
apratisaṃkhyānirodha 238, 341 āvajjana 153, 154, 270
artha 69, 379, 380, 388, 389 avaktavya 403, 406
Arthaviniścaya 65, 79 avasthānyathātva 235
ārūpya(samāpatti) four) 52, 53, 54, 55, 62, 64, avidyā 163, 198, 250, 255, 258, 259, 297, 306,
65, 69 341, 366, 371, 376, 377
ārūpyadhātu 324, 332, 339, 342, 383 awareness-holder dominating life 344
āryā meter 176, 197, 206, 207 āyatana 11, 65, 75, 76, 110, 115, 116, 117, 123,
Āryadeva 14, 307, 360, 361, 362, 370, 372 124, 126, 231, 245, 250, 253, 254, 256, 257,
[ārya]satya (four) 54, 55, 57, 62, 65, 68, 77, 259, 260, 264, 265, 266, 267, 268, 269,
85, 86, 98, 109, 110, 180, 193 224, 287, 299, 272 287, 300, 301, 308, 309 328, 329, 330,
342, 354, 363, 380, 382, 383, 384 364, 365, 376, 382
Āryasatyāvatāra 169, 176, 177, 178, 180, 181,
193, 196, 197, 214 Bactria 7, 29, 40, 41, 50, 93
asaṃhārya-dharma 66, 73, 74, 75, 98 bahiddha-
asaṃkhyeyakalpa 327, 328 bahiddhā (dhammā) 114, 116, 120, 121, 125,
127, 135, 141
asaṃskṛta dharma 31, 67, 286, 365
bahiddhārammaṇā (dhammā) 120, 122,
Asaṅga 59, 241, 243, 306, 307, 335, 345, 379
124, 127, 128, 133
asatkāryavāda 359, 360, 372
bahiddhā 7, 17, 108, 109, 110, 111, 112, 113,
Aśoka 5, 65, 90
117, 118, 119, 121, 121, 122, 123, 127, 128, 129,
*Aṣṭagrantha / *Aṣṭaskandha(śāstra) 40, 41,
130, 131, 132, 133, 134, 136, 138, 139, 140,
42, 43, 283, 302
141, 142
Aṣṭasāhasrikā Prajñāpāramitā 49, 50, 62, 63,
bāhira 115, 116, 117, 119, 120, 123, 133, 134,
66, 67, 73, 76, 77, 103
135, 140, 179, 217, 253
astitva 237, 401
Bahuśrutīya 14, 354
Asvabhāva 241
Bajaur Mahāyāna sutra 7, 8, 48, 51, 77, 78, 88,
Aśvaghoṣa 175 89, 94, 98, 99
Aśvakarṇa 320 Bamiyan 47
Atiyoga 319, 329 Bannihuan jing 般泥洹經 253
ātman 158, 359, 363, 365, 367 Bechert, Heinz 170, 176, 177, 184, 189
ātmavāda 38 Beckwith, Christopher I. 39, 40, 41, 42, 43
Ātmavāda-pratiṣedha 397 becoming (see bhava)
atom (see paramāṇu) beingness (see bhāvatā; dharmatā)
attachment (see rāga) benzhi xiangfen本質相分 389
attentiveness (see saṃprajanya) Bhāṣya 177, 214
aṭṭhakathā 113, 141, 175, 178 Bhāṣyaṭīkā 177, 215
Aṭṭhakathā 190 bhava 250, 257, 260
Aṭṭhakathākaṇḍa (see also Atthuddhāra) 123, bhāvanā 212, 381
135, 141 Bhāvanākrama 331, 333, 345, 346, 331
416 Index
bhāvanāmārga 224, 343, 380, 383, 384 buliaoyi 不了義 (see neyārtha)
bhāvānyathātva 235, 240 Butkara I 29
bhāvatā 408
Bhāvaviveka 361 caitasika / cetasika 122, 125, 205, 308, 309,
bīja 147, 203, 206, 207, 223, 234, 242, 359, 381 310
bimba 389, 390, 391 Cakravāḍa 320
birth (see jāti) calm abiding (see śamatha)
bliss (see sukha) Candrakīrti 359, 361, 364, 370, 371, 372
bliss of veritable refinement (see Caraka Saṃhitā 37
praśrabdhisukha) Cardona, George 42
Bodhgayā 170 Catuḥśatakaśāstra (Bai lun 百論 ) 307, 360,
bodhi 60, 73, 74, 286, 288, 321, 375, 376, 379, 371, 370
380, 381, 384, 389, 392 Cāturmahārājakāyika 322
bodhipakṣika-dharma / bodhipakṣya- causal condition (see hetupratyaya)
dharma 53, 54, 55, 63, 64, 65, 67, 75, cause (see hetu)
77, 98, 245, 302 cause of maturation (see vipākahetu)
bodhisattva xiv, 49, 59, 70, 73, 74, 76, 77, 80, Central Asia ix, x, 6, 9, 11, 39, 40, 41, 58, 223,
81, 94, 97, 306, 307, 320, 321, 322, 327, 279
328, 340 cessation (see nirodha)
bodhisattva path 7, 49, 321 cessation due to deliberation (see
body of form (see rūpakāya) pratisaṃkhyānirodha)
Bojjhaṅgasaṃyutta 112 cessation not due to deliberation (see
Brahmā 60, 321, 323 apratisaṃkhyānirodha)
Brahmajālasutta 113 cessation of perceptions and feelings (see
Brahmakāyika 321, 323 samjñāvedayitanirodha)
Brāhmaṇa 3, 86 cetanā 153, 154, 212, 261, 262, 263, 264, 331
Brahmanism 3, 5, 10 Ceylon (see also Sri Lanka) 17, 18, 169, 170, 171,
Brahmanic(al) 5, 17, 29, 32, 36, 37, 38, 175, 176, 202,210, 213
321 chanda 68, 112, 114, 131, 132, 331, 328, 339
Bronkhorst, Johannes 7, 17, 18, 50, 53, 99 chanshu 禪數 301
Buddha ix, x, 1, 3, 4, 5, 6, 8, 13, 30, 31, 49, 50, Chang Ahan shi bao fa jing 長阿含十報法經
51, 71, 73, 79, 80, 81, 82, 83, 88, 89, 91, 92, 59, 250
93, 94, 95, 96, 97, 98, 127, 176, 224, 233, Chang’an 長安 283, 302
240, 241, 248, 258, 266, 279, 282, 297, Cheng weishi lun 成唯識論 228, 229, 239, 241
298, 299, 300, 301, 304, 305, 306, 307, Cheng weishi lun liaoyi deng 成唯識論了義
309, 310, 315, 316, 321, 322, 362, 366, 370, 燈 390
375, 382, 391, 392, 396, 398 Cheng weishi lun shuji 成唯識論述記 376,
Buddhabhūmiśāstra 391 380, 388
Buddhabhumyūpadeśa 377, 378, 379, 382, China ix, x, 10, 11, 13, 15, 18, 223, 224, 227, 270,
392 271, 279, 280, 283, 285, 286, 290, 296,
Buddhadatta 211 301, 302, 305, 310, 393
Buddhaghosa 8, 17, 72, 79, 80, 86, 118, 119, Chos-’byung me-tog snying-po 345
122, 125, 126, 128, 130, 134, 135, 136, 137, Chu 楚 279
138, 140, 141, 174, 183, 185, 194, 195, 197, Chu sanzang ji ji 出三藏記集 279, 283, 289
198, 199, 205, 336 Chunqiu 春秋 281
Buddhamitta 213 cittaikāgratā 333, 334, 335
Buddhapālita 359, 361, 371, 372 cittānupassanā 7, 110
Buddhist logic 377 cittaviprayukta saṃskāra 308
Index 417
Cittuppādakaṇḍa 122, 127 cosmological 13, 280, 314, 318, 319, 320,
class (see jāti) 324, 338, 339, 345
cognition (see jn͂ āna) craving (see tṛṣṇā)
cognizing the principle (see dharmajñāna) Culla-Dhammapāla 211
cognizing the principle in its subsequent cumulative (see aupacayika)
realization (see anvayadharmajñāna) Cūlavaṃsa 169, 170, 171, 213
collocation (syntactic) 108, 109, 110, 111, 114, Cullaniddesa 109
116, 117, 119, 120, 122, 123, 134, 139, 140 cyclic existence (see saṃsāra)
Commentary on the Treasury of the Abhid-
harma (see Abhidharmakośabhāṣya) Da anban shouyi jing 大安般守意經 252
common knowledge (see saṃvṛti) Dao’an 道安 288, 289, 291, 302, 303, 304,
competence 334 305
concentration (see samādhi) Daodejing 道德經 254
concentration of cessation (see Dao di jing 道地經 250, 276
nirodhasamādhi) Daoism 12, 281, 282, 290
conceptual activity (see vikalpa) Da piposha lun 大毗婆沙論 301, 305, 357
conceptualism 376, 377, 392, 393 Dar-ma Khri-’u dum-btsan 346
conceptually existent (see prajñaptisat) darśana / dassana 69, 82, 96, 97, 111, 128, 177,
concomitant cause (see saṃprayuktakahetu) 182, 185, 187, 188, 194, 199, 212, 213, 214,
condition qua cognitive object (see 218, 219, 229, 366, 381
ālambanapratyaya) darśanamārga 224, 323, 342, 343, 383, 384
conditioned factor (see dharma / dhamma; Dārṣṭāntika 10, 177, 223, 225, 230, 231, 232,
saṃskāra) 233, 241, 242
conditioning (see saṃskāra) Dārṣṭāntika-Sautrāntika 174, 225, 230, 231,
Confucian(ism) 11, 12, 17, 18, 279, 280, 281, 233, 234, 239, 242, 243
282, 283, 286, 287, 289, 290, 291, 292 Daśabhūmikasūtra 80, 81
Confucius 12, 266 Dasheng fayuan yilin zhang 大乘法苑義林
consciousness (see vijn͂ āna) 章 298, 376
contamination (see kleśa) Dasheng fayuan yilin zhang buque 大乘法苑
contamination-hindrance (see kleśāvaraṇa) 義林章補闕 376, 389
contemplation (see dhyāna / jhāna) Dasheng yizhang 大乘義章 304
contemplation-as-cause (see Daśottarasūtra (see Chang Ahan shi bao fa
kāraṇadhyāna) jing)
contemplation-as-effect (see deed (see karma)
kāryadhyāna) delusion (see moha)
contemplation of feelings (see dependent origination (see
vedanānupassanā) pratītyasamutpāda)
contemplation of mental objects (see designation by provisional naming (see
dhammānupassanā) prajñapti / paññatti)
contemplation of mind (see desire (see kāma)
cittānupassanā) desire realm (see kāmadhātu)
contemplation of the body (see Dessein, Bart 87
kāyānupassanā) deterioration (see jarā)
conventional truth (see saṃvṛtisatya) Dhammajoti KL Bhikkhu 385
Conze, Edward 62, 63, 316 dhammānupassanā 7, 110, 114
co-occurring cause (see sahabhūhetu) Dhammapada 126
cosmology 318, 319, 320, 321, 322 Dhammapāla 116, 170, 171, 175, 186, 206, 211
Dhammarucikas 170
418 Index
Pakistan 29, 48 perspective 396, 397, 399, 403, 404, 405, 406,
pañcaskandha 9, 15, 110, 112, 121, 182, 193, 259, 409,
263, 287, 300, 301, 308, 328, 355, 362, Peṭaka 180, 222
363, 364, 365, 366, 367, 376, 382, 384, Peṭakopadesa 261, 263, 273,
390, 398, 399, 400, 403 405 406 408 409 phenomenological 13, 20, 258, 314, 319, 324,
Pañcaskandhabhāṣya 161 326, 345, 384, 385, 387
Pañcaskandhaka 9, 19, 146, 148, 149, 150, 152, phenomenon (see dharma)
153, 155, 156, 157, 158, 160, 161, 163, 164 Polonnaruva 213
Pañcaskandhakavibhāṣā 19, 146, 147, 148, Porāṇas 186, 187
149, 150, 152, 153, 154, 155, 156, 157, 158, post-enlightened cognition (see
160, 163, 164 pṛṣṭhalabdhajñāna)
*Pañcaskandhavivaraṇa 161 prabandha 227
pañcavastuka 300 pradeśa 34
Pañcaviṃśatisāhasrikā Prajñāpāramitā 59, prahāṇamārga 342
62, 63, 64, 65, 77, 100, 223, 276, 305 prajñā / paññā 69, 70, 73, 78, 79, 82, 86, 96,
Pāṇini 32, 33, 34, 113, 119, 226, 240, 291, 315, 329, 341, 383,
panjiao 判教 297, 311 385, 386, 387, 391
Papañcasūdanī 116, 118, 126, 128, 134, 135, 138, prajñāpāramitā 74, 77, 99,
142 Prajñāpāramitāsūtra 48, 101, 103, 305, 321
Parākramabāhu II 169, 170, 174, 220, prajñapti / paññatti 174, 355, 362, 363, 369,
paramāṇu 34, 228, 239, 240, 326 372, 389, 390, 393, 405, 408
paramārtha 300, 362, 363, 364, 365, 366, prajñapticitta 365,
367, 368, prajñaptisat 315, 316, 355, 362, 363, 364
paramārthamukha 363 pramāna 359, 377
paramārthasat 363, 366, Pramāṇasamuccaya 388
paramārthasatya 299-300, 362, 363, 364, prāpti 229, 264
365, 368 prasāda 79, 85, 86, 92, 93, 334
Param’-atthavinicchaya 191 prasaṅga 359, 369, 372
Paranirmitavaśavartin 322, 323 Prasannapadā 359, 373
pariṇāma 233, 235, 242 praśrabdhi 331, 334
Parthia ii, 283, 285, 301 praśrabdhisukha 336
particular characteristic (see svalakṣaṇa) pratibimba 371, 389, 389-390, 391, 393
particulars 376, 378, 380, 389, 394 pratimokṣa 321
parts and the whole 354, 356, 356, 358, 360, pratipad (four) 57, 65, 68, 72,
372, 387 pratisaṃkhyānirodha 341
paryavasthāna 233 pratisaṃvid / paṭisambhidā (four) 54, 61, 63,
passing away (see vyaya) 64, 65, 69, 70, 77, 173, 174, 212
Patan͂ jali 34, 42, 50, pratītyasamutpāda 34, 65, 75, 148, 229, 245,
path of cultivation (see bhāvanāmārga) 248, 249, 251, 252, 253, 254, 255, 256,
path of relinquishing (see prahāṇamārga) 257, 258, 259, 260, 262, 263, 264, 267,
path of seeing (see darśanamārga) 268, 269, 270, 272, 287, 299, 366, 367,
path of utter liberation (see vimuktimārga) 369, 370, 371, 372, 401, 407, 409,
path without obstacle (see ānantaryamārga) pratyakṣa 231, 234, 235, 377, 380, 384, 388,
Paṭṭhāna 120, 128, 189, 190, 191, 192 390
Pengcheng 彭城 279 pratyavekṣajñāna 382
perception (see saṃjn͂ ā; vijn͂ āna) pravṛttivijñāna 146, 148, 150, 151, 152, 153, 154,
permanence (see nityatā) 155, 156
person (see pudgala) prayoga 70, 381, 382
Index 425
sukha 68, 69, 114, 130, 131, 216, 309, 334, 335, tathatā 378, 379, 380
336, 337, 338, 339 *Tattvasiddhiśāstra 14, 19, 353, 354, 355, 356,
sukhā vedanā 130, 131, 336, 338 357, 358, 359, 360, 361, 362, 363, 364,
Sukhāvativyūhasūtra 309 365, 366, 367, 368, 369
Sumaṅgala 5, 169, 170, 173, 174, 175, 187, 189, Taxila 29
190, 191, 192, 193, 199, 205, 207, 209, 213 tejaḥ 90, 205, 207, 287, 315, 323, 330, 388
Sumaṅgalavilāsinī 118 Thaneshwar 35, 36
(Su)Meru 319, 320, 322 Theravāda 6, 8, 17, 30, 58, 65, 66, 118, 119, 128,
summit (see mūrdhan) 136, 140, 141, 153, 165, 170, 184, 211, 225,
śūnya 14, 97, 136, 197, 217, 306, 342, 362, 363, 243, 244, 245, 299, 308, 338
366, 367, 368, 369, 371, 372, 383, 408, Theravādin 9, 98, 169, 171, 175, 176, 178,
409 184, 191, 196, 244
śūnyatā 62, 74, 95, 97, 98, 99, 126, 307, 329, Theriya 10, 169, 170, 194, 196, 202, 203, 205,
353, 356, 360, 362, 364, 365, 366, 368, 206, 208, 210, 214
369, 370, 371, 372, 378, 381, 383, 409 third order chiliocosm (see
śūnyatācitta 365, 369 trisāhasramahāsāhasralokadhātu)
suo zhi zhang 所知障 381 thirst / thirsting (see tṛṣṇā)
sureness (see prasāda) thorough limpidity (see samprasāda)
sustained contemplation (see vicāra)
thought concomitant (see caitasika)
sūtra / sutta xiv, 1, 3, 4, 7, 8, 13, 47, 48, 49, 50,
Three Kingdoms 253, 265, 266
51, 54, 70, 77, 78, 86, 88, 89, 90, 92, 94,
three types of consciousness 365
95, 97, 98, 99, 128, 180, 183, 198, 201, 229,
threefold nature (see trisvabhāva)
230, 232, 233, 282, 288, 292, 299, 300,
Thūṇā (see Sthūṇā)
301, 304, 305, 310, 315, 342, 356, 366, 398,
thusness (see tathatā)
400, 405,
tianming 天命 280
Sūtrakṛtāṅga, 34
*Sūtrālaṃkāravṛttibhāṣya 154, 162 Tiantai 天臺 297
Sūtrayāna 345 tianzi 天子 280
Sūyagaḍa 34, 35 ṭīkā literature 169, 171, 203, 206, 213, 243
svabhāva 14, 92, 226, 235, 236, 237, 240, 300, time 17, 34, 35, 118, 138, 184, 236, 237, 318, 326,
326, 362, 327, 340, 343, 353, 403
svalakṣaṇa 15, 74, 300, 314, 355, 375, 376, 377, Tīrthika 15, 74, 76, 367, 396, 397
378, 379, 380, 381, 382, 383, 384, 385, 387, tong yang 痛癢 250, 252, 253
388, 389, 390 tongda 通達 (see nirvedha)
svarūpa 237, 371, 377, 380, 388 transcendence obtained visibly in this life
Śvetāmbara 34, 35 (see dṛṣtadharmanirvāṇaprāpta)
synod (see saṃgīti) transcendence with a remainder of the
syntagm 17, 108, 117, 139, 140 aggregates (see sopadhiśeṣanirvāṇa)
transcendence without remainder of the
Taiping dao 太平道 281 aggregates (see nirupadhiśeṣanirvāṇa)
Taizi ruiying benqi jing 太子瑞應本起經 transformation (see pariṇāma)
252, 264, 276 Trāyastriṃśa 321, 322, 323
Tamil country 9, 17, 170, 171 triad (ajjhattaṃ, bahiddhā, ajjhattabahiddhā)
Tang 唐 dynasty 13, 229, 284, 290, 298 109, 110, 112, 113, 117, 119, 133, 139, 140, 141
tantra xiii Triṃśikā 152, 154, 155, 321, 323
tantric 320, 321, 329, 341, 343, 344, Triṃśikāvijñaptibhāṣya 152, 155, 165, 229, 239,
tathāgata 51, 59, 64, 65, 67, 74, 77, 98 245
Tathāgata 7, 60, 63, 75, 95, 77, 95, 126, 127, Tripiṭaka / Tipiṭaka 1, 108, 111, 126, 130, 229
304, 306, 321, 365, 368 trisāhasramahāsāhasralokadhātu 319
Index 429
trisvabhāva 390 332, 342, 345, 353, 357, 383, 396, 397,
tṛṣṇā 53, 71, 116, 197, 198, 199, 205, 250, 254, 398, 405, 406, 407
255, 256, 257, 258, 260, 264, 266, 267, Vātsīputrīya 245, 353, 397
268, 270, 272, 316, 332, 333, 334 vāyu 183, 287, 315, 320, 323, 327
tṛṣṇotpāda / tṛṣṇāmūlaka-dharma (four) 57, Veda 117,
71 Vedic xii, 3, 4, 5, 8, 17, 117, 139
truth (see satya; tathatā) vedanā 66, 110, 121, 129, 130, 131, 133, 135, 182,
Tunnyun 229, 194, 195, 204, 235, 237, 250, 251, 252, 253,
Tuṣita 321, 322, 323 258, 260, 263, 268, 272, 308, 328, 333,
twelve sense-spheres (see āyatana) 336, 362, 367, 398, 399
two truths 362, 363, 364, 369, 370, 372, 384 vedanānupassanā 7, 110
vedanopekṣā 335, 336, 337
ubiquitous cause (see sarvatragahetu) veritable refinement (see praśrabdhisukha)
ucchedadṛṣṭi 401 Vibhaṅga 54, 55, 65, 69, 71, 100, 120, 121, 128,
Udāna 126 129, 130, 131, 132, 133, 134, 135, 139, 140,
ultimate truth (see paramārthasatya) 141, 174, 181, 212, 245, 268, 299
unbreakable confidence (see abhedyaprasāda vibhāṣā 10, 40, 41, 42, 43, 231
/ avetyaprasāda / aveccapasāda) Vibhāṣā 40, 41, 42, 42, 231
understanding (see prajñā / paññā)
vicāra 68, 218, 238, 332, 333, 334, 335, 336,
unity 240, 354, 355, 357, 360
337, 338, 340, 341
universal characteristics (see
Vidiśā 9
sāmānyalakṣaṇa)
Vigrahavyāvartanī 355, 358, 359, 362, 367
unlimited (see apramāṇa)
vihāra 40
upādāna 121, 198, 249, 250, 251, 253, 256, 257,
vijñāna 53, 57, 72, 124, 138, 148, 149, 150, 151,
258, 260, 261, 264, 267, 268, 356, 400
152, 155, 158, 159, 160, 161, 162, 164, 185,
upādāyarūpa 287
Upaniṣad 3 186, 187, 188, 195, 196, 202, 203, 204, 208,
Upaniṣadic 139 212, 213, 228, 229, 235, 239, 240, 250, 252,
upapatti 211, 332 255, 257, 259, 260, 263, 266, 270, 310,
upāya 72, 364, 378, 393 325, 328, 329, 330, 334, 339, 340, 356,
upekṣā / upekkhā 68, 69, 115, 204, 211, 212, 357, 398, 399
216, 331, 332, 333, 334, 335, 336, 337, 339 vijñānānantyāyatana / viññāṇañcāyatana
ūṣmagata 381 124, 125, 135, 339
Uttarakuru 321 Vijñānavādins 178
vijñaptimātra 152, 242
Vaibhāṣika 10, 11, 40, 178, 184, 224, 225, 226, Vijñaptimātra 389, 392, 393
227, 228, 229, 230, 231, 233, 234, 235, 236, vikalpa 153, 162, 355, 384
238, 240, 257, 326, 340, 353,356, 357 Vimalakīrtinirdeśa (Weimojie jing 維摩詰經 )
Vaidalyaprakaraṇa 358 81, 106
vaiśāradya (four) 60, 61, 63, 64, 65, 66, 70, vimāna 323
76, 77 vimuktimārga 342
Vaiśaravaṇa 322 Vimuktisena 83, 84
Vaiśeṣika 36, 37, 359, 372, 397,405 Vimuttimagga 196
vāsanā 150, 151, 224, 234, 241, 381 Vinataka 320
vaśitā (four) 59, 60, 61, 64, 66, 70 Vinaya 1, 4, 5, 13, 54, 77, 92, 93, 177, 304
Vasubandhu 9, 10, 15, 91, 92, 93, 94, 106, 146, vipākahetu 324, 325
148, 149, 150, 158, 163, 175, 177, 178, 223, vipākaja 242
224, 227, 228, 229, 230,231, 232, 233, 235, vipākaphala 315, 343, 344
236, 238, 241, 284, 315, 324, 328, 329, 331, vipaśyanā / vipassanā 118, 119, 340
430 Index