The Presence of Buddhist Thought in Kalām Literature

The Presence of Buddhist Thought in Kalām Literature
Dong Xiuyuan
Philosophy East and West, Volume 68, Number 3, July 2018, pp. 944-973 (Article)
Published by University of Hawai'i Press
For additional information about this article

https://muse.jhu.edu/article/700765
Access provided by Peking University (8 Aug 2018 14:44 GMT)

THE PRESENCE OF BUDDHIST THOUGHT IN KALĀM LITERATURE
Dong Xiuyuan
Center for Judaic and Inter-religious Studies, Shandong University
dongxiuyuan2010@126.com
This paper1 is intended to examine the accounts of Buddhist thought in

Kalām literature and its influence on the early Mutakallimūn. I shall focus
on the Samaniyya’s view on epistemology, the Barāhima’s rejection of
prophecy, and the origins of Islamic Atomism. These seemingly separate
topics were all treated by Shlomo Pines throughout his academic career
spanning half a century. Pines, who made groundbreaking contributions to
each issue, did not establish a link among them. Based on the examination
of Buddhist literature and Kalām works, the present study hopes to shed
some light on the common sources of the Barāhima–Samaniyya doctrines
and Islamic Atomism.
I. The Samaniyya on Epistemology
It is generally accepted that the term Samaniyya (or Shamaniyya,

Sumaniyya), deriving from Śramaṇa (meaning “monk”), designates the
Buddhist thinkers in medieval Arabic Literature.2 In Kalām works, there are
reports of debates between some early Mutakallimūn and the Samaniyya.3
Pines calls our attention to the one related to the founder of the Jahmiyya,
Jahm b. Ṣafwān (d. 745):
It is reported that some Samaniyya said to Jahm b. Ṣafwān: Does the knowledge
of what is regarded as good (al-ma‘rūf) follow (yakhruj) from the five senses (al-
mashā‘ir al-khamsa)?
He answered: No.
They said: Tell us about the object of your worship (ma‘būd). Do you know
Him (‘araftahu) by any of them?
He answered: No.
They said: Consequently He is unknown.
Hereupon he remained silent and wrote about this to Wāṣil [b. ‘Aṭā’]. The latter
replied and said: You may posit a sixth [source of knowledge], namely
inference (dalīl). Hereupon they will say: He (God) does not follow either from
the senses or from inference (my emphasis). Thereupon I shall ask them whether
they distinguish between the living and the dead, between rational being and a
madman. There is no doubt about an affirmative answer. Now this is known by
944 Philosophy East & West Volume 68, Number 3 July 2018 944–973
© 2018 by University of Hawai‘i Press
means of inference. When Jahm gave this answer, they (the Samaniyya) said:
‘This is not your discourse (kalām)’. He informed them (of the facts). Hereupon
they went to Wāṣil, had a conversation with him, and were converted by him
to Islam.4
An earlier version of the same story is reported by Ibn Ḥanbal:

Some information of Jahm came down to us. He is from Tirmidh (my emphasis)
in Khurasan and a master of debate and Kalām. Most of his kalām is about
God. He met some people among the polytheists called Samaniyya who knew
Jahm. So they said to him: We are arguing with you. If our argument prevails
over yours, you should convert to our religion; if your argument prevails over
ours, we will convert to your religion.
So they argued with Jahm and said to him: Do you not assert that you have a
God? Jahm said: Yes. They said: Did you see your God? He said: No. They
said: Did you hear his word? He said: No. They said: Did you smell his smell?
He said: No. They said: Did you find him by taste? He said: No. They said: Did
you find him by touch? He said: No. They said: Then how do you know that he
is God?
It is said that (thereafter) Jahm was confused and did not participate in worship
for forty days. Then he came up with an argument just like that of the Christian
heretics who assert that the spirit (rūh. ) in Jesus the son of Maria is the spirit of
God from God’s essence . . . Jahm came up with an argument just like this
one. Thus he said to the Samaniyya: Do you not admit that there is a soul (rūh. )
inside you? They said: Yes. He said: Did you see your soul? They said: No. He
said: Did you hear its word? They said: No. He said: Did you find it by taste or
touch? They said: No. He said: Then similarly God does not have a face to be
seen, nor voice to be heard, nor smell to be smelled, and (then) he is unseen
(ghā’ib ‘ani ’l-abṣār) and not in any place.5
The views of the Samaniyya in both versions reflect faithfully the
Buddhist theory on the criteria of knowledge (pramāṇa). According to
Dharmakīrti (around 600–660), there are two and only two sources of
knowledge:
Right knowledge is two-fold: perceptual (pratyakṣam) and inferential (anumānaṃ).
Perceptual knowledge means here neither construction (kalpanā) nor illusion . . .
It is four-fold: (1) sense knowledge; (2) mental knowledge [that] follows [the first
moment of every] sense-cognition [which is thus] its immediately preceding
homogeneous cause. [The latter] is cooperating with [the corresponding moment
of] the object, which immediately follows the proper object [of sensation]; (3)
every consciousness and every mental phenomena [that] are self-conscious; (4)
the intuition of the saints (yogijñānaṃ) [that] is produced from the supreme state
of deep meditation on true reality (bhūtārtha).
Its (perceptual knowledge’s) object is the particular . . . that alone [which is

unique] represents ultimate reality. Because of its efficiency [to produce
Dong Xiuyuan 945

knowledge], it (the particular) is reality. Different from it is the universal. It is
the province of inferential knowledge.6
Dharmakīrti summarizes here the epistemological principles of hetuvidyā
(Buddhist logic) accepted by most Buddhist sects since that time. Perception
is the ultimate source of knowledge, which corresponds to reality, while
inference is by nature a mental construction whose validity depends on the
perceptual data it has synthesized.7 An absolutely imperceptible object
would be a pure illusion corresponding to nothing existent outside the
mind.8 Among the four classes of perceptual knowledge, knowledge of the
five senses is the primary and only undisputed one.9 In this vein, we find
that, during the debate on the existence of God, the Samaniyya began
exactly by asking about the sensory sources of the theistic claim. When the
Mutakallimūn resorted to inference as a sixth source of knowledge, the
Samaniyya would protest, as Wāṣil b. ‘Aṭā’ expected, that the existence of God
does not follow either from the senses or from inference, because what Jahm
tried to establish by inference is the existence of God as an incorporeal entity,
whereas in view of the Buddhists, to infer a nonsensible entity from the
sensible phenomena is invalid.10
Jahm’s final move based on the analogy of God and soul11 is also
reminiscent of the Buddhist position. This analogy can be found in
Abhidharma-mahā-vibhāṣā-śāstra12 (composed around the second century
C.E., hereinafter referred as Vibhāṣā):
Vasumitra said: There are unbelievers who affirm the existence of the soul
(shiwo 實我) by invalid reasoning: The soul is subtle and permanent everywhere;
it is either latent or active in the phenomena. When the unbelievers see that the
wind is blowing and the river is flowing, they say it is the soul that causes all
these phenomena. Their reasoning is not valid like inferring the effect of the
wind from the movement of the tree . . . those who assume “I am the agent”
insist that there is a real soul (shengyiwo 勝義我) inside the body, which is able
to affect, generate and transform all kinds of [psychical] things, and those who
assume “He is the agent” insist that there is a real soul outside the body, which
is able to affect, generate and transform all kinds of [physical] things.13
The Buddhists used the term ātman to denote both the soul inside the
human body and the soul of the universe. The exponents of these two kinds
of souls infer the existence of the transcendent entities from the sensible
phenomena (both physical and psychical). Vasumitra rejected the inference
as invalid, for he takes correct inference to be confined to the domain of the
perceptible. In Abhidharma-Kośa-śāstra (composed around the fourth to fifth
century C.E., hereinafter referred as Kośa), the author Vasubandhu exposed
the fallacy of this kind of inference in more detail:
How do we know that the word ‘soul’ (ātmā; wo 我) is only a designation for a
series of skandhas, and that no soul exists in and of itself? We know this
946 Philosophy East & West

because no proof establishes the existence of a soul apart from the skandhas,
no proof by direct perception (pratyakṣa), nor any proof from inference
(anumāna). If the soul were a real entity, separate like other entities, it would be
attained (i.e., known) either by direct perception as are the objects of the five
sense consciousnesses and the object of mental consciousness, or by inference,
as are the five indriyas . . . There is neither direct perception nor inference
(my emphasis) of a soul independent of the skandhas. We know then that a real
soul (ātmā; zhenwoti 真我體) does not exist.14
The direct object of Vasubandhu’s criticism are the Vātsīputrīyā, who are
the predecessors of the Sammatiya.15 The latter two are the only Buddhist
schools that affirm the existence of the soul. The Sarvāstivādins insist that the
existence of an entity can be established either by direct perception or by
inference based on perception. The object known by inference may not be
perceived but must be in principle perceptible. For example, we can deduce
the existence of the sense organ (indriya) from the occurrence of sense
consciousness, but cannot go further to affirm that there is a nonsensible
subject underlying all the sense organs and sensations. It is notable that
Vasubandhu’s phrase “neither direct perception nor inference” is strikingly
similar to that of the Samaniyya in al-Murtaḍā’s version cited above.16
Nevertheless, the Buddhist rejection of the soul is not without difficulties.
The tension between rejection of the soul and acknowledgement of the
subject underlying reincarnation has haunted Buddhist thinkers through the
ages. In fact, every Buddhist sect has its own alternative concept bearing
some functions of the soul.17 This was facilitated by the fact that most of
these sects accepted some form of inner consciousness as direct percep-
tion,18 which would in turn support the existence of a certain kind of mental
subject. The subtle difference between the soul and these alternative
concepts might hardly be detected by lay believers, let alone the early
Muslim audience. So Jahm’s response to the Samaniyya could be seen as an
echo of the soul-entity problem faced by the Buddhists themselves.
Related to this, the ninth-century Jewish theologian Dāwūd al-Muqammaṣ
addresses an unnamed objector (qā’il)—following a discussion of the
analogy between God and soul19—who resorts to direct perception
(shāhadnā)20 to support his affirmation that God is a substance with
qualities.21 He further claims that al-Muqammaṣ’s negative theology actually
compares (shabbaha) God to substance and that denial of the possibility of
comparison between God and creations will lead to denial of general
syllogism in all senses (qiyās ‘āmm bi-jamī‘ al-ma‘ānī).22 In view of this, al-
Muqammaṣ deems it imperative to clarify what the comparisons (al-ashbāh)
are, and then makes a move that is of great interest for our purpose: he says
that he would leave the discussion of this issue to the book he has begun
writing in refutation of the Buddhists (aṣh. āb al-budūd).23 It is the unmistak-
able affinity between the premises of the objector24 and Buddhist logic that
makes al-Muqammaṣ decide to deal with the topic in a monograph against
Dong Xiuyuan 947

the Buddhists, who provide a full-fledged version of the epistemological
view in question.
II. The Barāhima on Prophecy
In Kalām literature, the term “Barāhima” denotes an Indian sect that argued
against Islamic prophecy. Although the pronunciation of this term is close to
“Brāhmaṇa”, there is no apparent connection between the Barāhima’s
rejection of prophecy in Muslim reports and the actual doctrine of the
Indian Brahmans. Kraus argued that the alleged Barāhima’s intellectual
argument is in fact a fiction invented by the ninth-century freethinker Ibn al-
Rāwandī, who put his own heretical thesis in the mouth of the Barāhima as
a ruse against persecution by the Islamic authorities.25 Bruce Lawrence
basically accepted Kraus’ fiction-theory, while he also noted some differ-
ences between the heresiographer Shahrastānī’s (1086–1153) account of the
Barāhima arguments and Ibn al-Rāwandī’s.26
Shlomo Pines stressed the difficulty Kraus faced in his attempt to derive
the report of the Jewish thinker Sa‘adia (d. 942) on the Barāhima—who in
Sa‘adia’s account assert that Adam was the first prophet and his prophecy
cannot be abrogated—from Ibn al-Rāwandī’s work. Pines also pointed to a
similar claim of an Indian group called “Ibrāhīmiyya”27 recorded by the
ninth-century Zaidite scholar al-Qāsim b. Ibrāhīm. Based on these two texts,
Pines reasoned that the accounts of the Barāhima referred to by al-Qāsim and
Sa‘adia are not likely to be a fiction invited by Ibn al-Rāwandī but rather a
reflection on the Brahman Law of Manu (the first man in ancient Indian
mythology), of which Ibn al-Rāwandī made use to convey his own view.28
Sarah Stroumsa further cleared Ibn al-Rāwandī from Pines’ modified
indictment by analyzing some early Islamic and Jewish sources, particularly the
arguments of the Barāhima recorded in ‘Ishrūn maqāla of al-Muqammaṣ.
Stroumsa proposed that the debate between the Barāhima and al-Muqammaṣ
(both the former’s arguments against prophecy and the latter’s response) reflects
Buddhist thought, and that the discrepancy between the Kalām accounts of the
Barāhima and actual Indian conceptions, as well as the divergence of the two
versions of the Barāhima arguments, is the result of the debate strategy.29
Binyamin Abrahamov and Norman Calder criticized Stroumsa’s theory
and insisted that there is no evidence to support any real connection between
the Barāhima narratives and Indian thought (either Brahman or Buddhist).
They both looked west for the source of the Barāhima: for Abrahamov, it is
remnants of late antique paganism, known in Arabic historiography as the
religion of the Ṣābians; for Calder, it is the so-called “Abrahamic religion”.30
Stroumsa responded to these criticisms in her Freethinkers of Medieval Islam,
revealing that these two alternative candidates are even more obscure than
the Barāhima on which they are supposed to shed light, and their existence
just as controversial.31

In the present study, I endorse Stroumsa’s thesis that the various pieces
of information about the Barāhima do have sources in Indian thought in the
broad sense. Yet I agree with Abrahamov that the two trains of Barāhima
arguments (the intellectual arguments and the anti-abrogation arguments) are
mutually exclusive in their original forms.32 There is good reason to assume
that the Barāhima narratives have two different sources:
(1) The anti-abrogation arguments, as Pines argues, were originally
derived from a correct report of the Indian Brahman doctrine, and
then adapted to Biblical narratives;
(2) The intellectual arguments came from some Buddhist accounts of
Brahman thought and practice, as we shall see later.
The two initially conflicting sources were, in my view, integrated by
later Mutakallimūn into a coherent framework of heresiology. The account
of Sa‘adia is a key text, which sheds light on the process of the confluence:
I (Sa‘adia) perceive that some of them (1) say: “if the Brahmans (al-Barāhima)
should say to you: ‘we have received the order transmitted from Adam to wear
clothes made both of wool and flax . . . ’, you would not be in a position to
put forward (against this) the statement of the prophet who has forbidden these.
For they (2) would say to you: ‘Adam has told us that these (orders) will not be
abrogated.’ ”
These arguments are—may God direct you well—without foundation. It is they

(1) who ascribe (idda‘ū) these to the Barāhima. For the Barāhima merely
maintain that these things are permitted. We too acknowledge that they have
been permitted when this was the case, and similarly their prohibition [is
admissible] by reason (al-‘uqūl), inasmuch as man is permitted to refrain of his
own accord [from doing] these things if some advantage should accrue to him.
If a Brahman (Barahmī) should come forward in the future (musta’nifan) in order
to maintain these (affirmations) that have been ascribed to him, he should not
be permitted (to do) so. For he who abides by tradition (al-nāqil), must affirm
everyday that which he has affirmed the day before. He is not like (a person)
who consults his own judgment (al-murta’ī) and is permitted to say: something
that I did not grasp yesterday, became manifest to me today.33
This paragraph appears in the context of a debate on abrogation (naskh)

of the early revelations between Muslim theologians34 and Sa‘adia as the
representative of the Jews. Here the “they (1)” refers to the Muslim
theologians who argue for the abrogation, while the “they (2)” refers to the
Barāhima as reported by the Muslim theologians. The Muslim theologians
quote the Barāhima’s claim of Adam’s primitive revelation to refute Sa‘adia’s
defense for the continuing validity of Mosaic Law, because if he insists that
the revelation cannot be abrogated, he will have no reason to reject the
claim of the Barāhima. As a response, Sa‘adia points out that the view of
Dong Xiuyuan 949

Adam’s primitive revelation is not the Barāhima’s position but is ascribed to
them by the Muslims, whereas the original stance of the Barāhima is that the
behaviors that go against the Law are permitted by reason. Thus, Sa‘adia
identifies the rationalist position35 underlying the intellectual argument as
the Barāhima’s true position and accuses Muslim theologians of misattribut-
ing the anti-abrogation argument to the Barāhima.36 Furthermore, he reports
that the Barāhima might have accepted what is ascribed to them and made
use of this notion to justify their own opinions and practices based on
reason. As Sarah Stroumsa shows, this is a typical strategy of debate in
Kalām, called Ilzām.37 The next stage in the evolution of the Barāhima
narratives, which can be detected from al-Qāsim’s earlier report of
Ibrāhīmiyya/Barhamiyya,38 is that the Muslim authors or scribes replaced the
foreign name Barāhima with a derivative of a similar and familiar Arabic
one Ibrāhīm and then the prophet in question was transformed accordingly
from Adam to Abraham.39
What concerns us most here, among various versions of the Barāhima
accounts, is the intellectual argument, for which Ibn al-Rāwandī and al-
Muqammaṣ provide the earliest examples.40 Stroumsa found in al-Muqammaṣ’s
report of this argument two points that are reminiscent of the Buddhist views
on Bodhisattva. One is related to the qualification of the prophet:
If he deserved to become a prophet, then it is either by virtue of perfect
knowledge, or of perfect conduct, or both . . . no one can say that a prophet
deserves his status unless it be by virtue of perfect knowledge or perfect
conduct.41
The other lies in al-Muqammaṣ’s response to the criticism of the
Barāhima on prophecy:
For we say that there are two kinds of righteousness: general and particular.
General righteousness is the one incumbent upon (everyone, both) prophet and
non-prophet. It is what is incumbent upon us, as the obligation to cleanse
ourselves of all defects and to achieve virtues, their subsidiaries qualities and
the works they entail. Particular righteousness (al-ṣalāh. al-khāṣṣ) is a concern to
teach others and instruct them. It is incumbent upon prophets, priests and
Levites.42
It is notable that the two quotations of Kośa that Stroumsa compares with
these two points both come from the same paragraph in their original context:
But why do the Bodhisattvas, once they have undertaken the resolution to
obtain supreme Bodhi, take such a long time to obtain it? Because supreme
Bodhi is very difficult to obtain: one needs a great accumulation of merit and
knowledge (mahatāṃ hi puṇya jñāna sambhāreṇa), and of innumerable heroic
works in the course of three asamkhyeya kalpas. One would understand that
the Bodhisattva searches out this Bodhi so difficult to obtain, if this Bodhi were
the sole means of arriving at deliverance; but such is not the case. Why then do

they undertake this infinite labor? For the good of others, because they want to
become capable of pulling others out of the great flood of suffering.43
The concept of Bodhisattvas here—someone with perfect knowledge and
merit seeking supreme wisdom in order to save others from sufferings of this
world—fits both the Barāhima’s view on the qualifications of the prophets
and al-Muqammaṣ’s description of particular righteousness belonging to the
prophets. As for the phrase particular righteousness, I am inclined to believe
that the term al-ṣalāh. corresponds to puṇya (merit, well-being), and al-ṣalāh.
al-khāṣṣ to mahā puṇya (great merit). The latter also appears in another
place of Kośa dealing with Buddha’s altruism:
Great compassion is, by its nature, conventional knowledge. In the contrary
case, it would be, in its nature, absence of hatred as is ordinary compassion;
like ordinary compassion, it would not embrace all beings of the Three Dhātus,
it would not envision the three types of suffering. Why is the compassion of the
Blessed One termed “great”? . . . By reason of its factors, it is produced in fact
by a great provisioning of merit (mahā puṇya) and knowledge.44
Vasubandhu refers to Buddha’s great compassion (mahā karuṇā) embracing
all living beings as opposed to the ordinary compassion (karuṇāvat), and
attributes the great compassion to the great merit (mahā puṇya) and
knowledge.
It thus seems to me that there were some assumptions with Buddhist
overtones (namely, the altruism and special qualifications of the teachers of
truth) accepted by both parties of the debate which al-Muqammaṣ reported.
What is more, the rationalistic rejection of the divine revelation is a
distinctive position of Buddhism in contrast to most other Indian religions.45
The Sarvāstivādin school went so far as to state that not all the teachings of
Buddha are in accordance with the truth, suggesting that the law of reason
is above the authority of Buddha.46 As mentioned above, the Buddhist logic
ultimately acknowledges only two sources of knowledge: direct perception
and inference based on direct perception,47 neither of which validates the
claim of revelation. A similar line of thought can be found in the Barāhima’s
rejection of prophecy as summarized by the tenth-century Shī‘ite scholar Ibn
Bābūya:
Our opponents said that customs (al-‘ādāt) and direct perceptions (al-
mushāhadāt) disprove what you said about the unseen (al-ghaib).48 I said: “With
regard to the verses of the prophet, the Barāhima would say to the Muslims as
follows: none of you has perceived it (tushāhidūhā), so perhaps you followed
someone who is not worth following or accepted unreliable report (khabar).”49
The Barāhima here states that the Muslims cannot prove the truth of
prophecy by direct perceptions and that their belief in prophecy is probably
based on some unreliable report. The connection between these two
statements is not that clear. The logical gap may be bridged over with the
Dong Xiuyuan 951

eleventh-century Zaidite Imām al-Nāṭiq’s account of the common ground
shared by the Muslims and the Barāhima in their debate on prophecy:
As for the path to knowledge (ma‘rifa) of the messenger, it is either necessary

(iḍṭirāran) or inferential (istidalālan). The former is rejected, because the men of
reason (al-‘uqalā’) dissent from each other regarding this. Thus, only inference
remains. It indicates two species: the intellectual (‘aqliyyan) and the heard
(sam‘iyyan). The intellectual one is rejected, because there is no path for the
intellect to attain the knowledge of prayer in such a way and that of fasting and
alms tax. Thus only the hearing (al-sam‘)50 remains.51
In what follows, the Muslims try to established hearing/report as a path

(ṭarīq) or cause (al-sabab) of knowledge other than direct perception (al-
mushāhada) which is obviously viewed as a model of the sources of necessary
knowledge.52 Accordingly, the stance of their opponents (the Barāhima) is that
direct perception and intellectual inference are the only two paths leading to
knowledge. We have seen the same epistemological position in the preceding
part on the Samaniyya. It should be noted that the rationalistic rejection of
prophecy and other versions of the Barāhima arguments are ascribed also to
the Samaniyya in some Kalām sources, and that both the Barāhima and the
Samaniyya are called “the learned of India” (‘ulamā’ hind).53
At the end of this section, I would like to address the question why the
Buddhist stance was presented in the name of the Barāhima/Brahmans. The
opposition of Brahmanes and Sramanes had been well known by the Greeks
since the first century B.C.E.54 The Kartīr inscription (the second half of the
third century C.E.) in Sassanian Persia juxtaposed these two Indian sects
(Shamans and Bramans).55 The account of Indian pilgrimage rituals by the
ninth-Century Nestorian ʿAbd al-Masīh. al-Kindī conveys this information
too.56 The Arabic authors thus may have known from various channels that
the Barāhima is an Indian sect distinct from the Samaniyya, though they had
few means of identifying the real doctrine held by the Brahmans.57 Here we
are confronted with a highly interesting phenomenon in intellectual history:
Buddhism originated in India where most of its scriptures were produced, but
it never saw such great success in its homeland as it achieved beyond the
Indian subcontinent. After its establishment, the classical writers of Buddhism
made a great effort to gain support from the upper castes of the Indian
society, an effort which is manifested in the fact that the Buddhist Scriptures
were addressed frequently to the Brahmans. In Vibhāṣā, the author describes
the Buddhist practitioners as zhen poluomen 真婆羅門 (“the true Brahmans”):
One who adheres to the dharma of Buddha is called Brahman. . . . The true
Brahmans are those who know that the dharma of extinction follows every
dharma of aggregation.58
Moreover, in the Sarvāstivādin edition of Udāna, Buddha redefines the

term of Brāhmaṇa with the doctrine and practice of Buddhism.59 In view of

this, it is likely that the early Muslims learned the intellectual argument
against revelation from Buddhist sources ascribed by the Buddhist informants
to the Brahmans. This theory can well explain the fact that the later
Mutakallim Shahrastānī presented the Buddhists (aṣh. āb al-bidada) as a
subcategory of the Barāhima.60
III. The Origins of Islamic Atomism
The doctrine of atom and its accidents (jawhar wa-a‘rāḍuhu) is the

cornerstone of the edifice of Kalām cosmology. Given its maturity in
framework and terminology, Islamic atomism is unlikely to be an original
accomplishment of the early Mutakallimūn.61 Most modern scholars naturally
followed the twelfth-century Jewish philosopher Maimonides in tracing it
back to Greek atomism.62 The theory of the Greek origins, however, always
encounters difficulties in explaining the fundamental differences between the
Greek and the Islamic atomism. As Pines and Wolfson pointed out,63 some
unknown development of atomism in late Antiquity or early Islamic period—
if there is such a Greek source—needs to be restored. Unfortunately, this
kind of evidence has not appeared so far. Alnoor Dhanani tried to confirm
Pines’ hunch of a possible influence of Epicurean minimal part64 by arguing
that the Kalām atom has magnitude.65 Leaving aside the methodological
weakness in his characterization of the early Kalām based on the rather late
sources of the eleventh century,66 there is still a sizable gap between the
theory of minimal part and Islamic atomism as acknowledged by Dhanani
himself.67 Tzvi Langermann draws attention to the Arabic translation of
Galen’s treatises containing his critique of atomism as the missing link
between the Greek and the Kalām theories.68 In a study on Epicurus and
Abū Bakr al-Rāzī, Lenn Goodman suggests that the Galen texts are most
likely to be the direct source of al-Rāzī’s atomism, while the atomism of the
Kalām is not the Epicurean one.69 Although the Arabic Galen corpus may
offer “a smoking gun” to identify the origin of al-Rāzī’s atomism,70 it fails to
account for the motivation of the early Mutakallimūn to base their dogmas
on such a physical theory, which had been thoroughly repudiated by the
Greek authorities in philosophy and sciences. The theological impetus to
establish God’s sovereignty provides too convenient an explanation,71 which
is in fact an inference from the role atomism played in classical Muslim
theology. Actually the main characteristics of Kalām atomism (e.g., the
transiency of atom-accident and the theory of accidents as a distinct category
of being), underlying its function to fit the doctrines of creation and of the
supremacy of divine will, cannot be found in any version of Greek atomism
but rather conflict with the latter’s basic premises. In other words, the Greek
theory reported by Galen had nothing really appealing to the early Islamic
theologians. On the contrary, it is the most alien one, among all Hellenistic
philosophical doctrines, to the principles of monotheism.
Dong Xiuyuan 953

In view of the difficulties of the Greek-origin hypothesis, Pines proposed
an alternative approach, turning to the east, namely India to look for the
sources of the Kalām theory. He shows in a convincing way that the points
where Islamic atomism deviates from the Greek theory are exactly the
common ground between the Kalām and the Indian atomism.72 At this
point, I would like to revisit Pines’ theory and offer several observations.
1. Epistemological presupposition of atomism. Most of the Mutakallimūn
(except Abū al-Hudhayl) believed that atoms can have sensible accidents
such as color, and some of them (Abū al-Qāsim al-Balkhī and the Ash‘arites)
even saw these sensible accidents as inseparable accidents of atoms.73 This
characterization of ultimate substance is unintelligible in any context of
Hellenistic philosophy, both atomistic and anti-atomistic. We can, however,
not only find its parallel in Sarvāstivādin classical texts,74 but also discover
the justification of this theory in Buddhist epistemology.
According to the Vaibhāṣikas, the Eastern branch of the Sarvāstivādin
school, atom (paramāṇu, “the finest”) is the indivisible unit of physical matter:
Atom is the smallest material particle, which cannot be divided nor broken nor
penetrated, and is neither long nor short, neither square nor circular, neither regular
nor irregular, neither high nor low. It is indivisible, invisible, cannot be heard nor
smelled nor tasted nor touched. So it is called “paramāṇu”, namely, the smallest
material particle. Seven atoms make up the smallest body visible to the eyes.75
This sounds exactly like Abū al-Hudhayl’s position: the atom has neither
magnitude nor secondary qualities. But the imperceptibility of the atom is
not a consensus among the Sarvāstivādins. What is more, it stands in tension
with the principle of Buddhist epistemology: if atom cannot be verified by
the senses, it will be in no way an ultimate reality. Thus, Vasubandhu
revised the classicical concept of atom, introducing a sensible aspect:
For us, the atoms, although supra-sensible, become perceptible (pratyakṣatvaṃ)
when they come together: the Vaiśeṣikas also attribute the power to create
coarse bodies to the united atoms.76
Saṃghabhadra (the fifth century C.E.), representative of the so-called

Neo-Sarvāstivādins, developed this theory and put forward a distinction
between two conceptions of the atom:
There are two kinds of atoms: one actual (dravya; shi 實), the other conceptual
(prajñapti; jia 假). What are they like? The actual one is the confirmed particular
physical being which is in a homogeneous aggregation (saṃhata; heji 和集)77
and thus can be attained by perception. The conceptual one is apprehended by
division, which is an object of inferential knowledge. The intellect divides the
aggregated physical matter to the utmost, and then distinguishes the atomic
characteristics such as color and sound. The result of this division is called the
conceptual atom.78

Saṃghabhadra argues that only the sensible atoms verified by perception
are real entities, while the atoms obtained purely by analysis, from which
the sensible characteristics have been abstracted, are just a kind of mental
construction. He further specifies the sensible aspect of the atom:
There is no atom that is not in a homogeneous aggregation. As they are always
in a homogeneous aggregation, atoms are not invisible. So the nature of atoms
has been established, which are definitely objects of the visual sense.79
The atoms here have sensible characteristics, and what their aggregation
manifests to the senses is the atom’s own quality in an amplified but still
homogeneous way. Therefore, the atom is qualified as an ultimate reality
according to the Buddhist criteria of right knowledge.80 A similar assump-
tion—namely, that only perception (Shāhad/Mushāhada) brings necessary
knowledge (‘ilm ḍarūrī) while knowledge from inference (istidlāl/dalāla)
cannot be necessary—can be detected in the discussion of the Mu‘tazilites
regarding human knowledge of God and regarding the sources of knowledge
about prophecy.81 In my view, it is the same epistemological stance that
justifies Kalām’s attribution of sensible accidents to the atoms.
2. A thoroughly atomistic understanding of being. The mainstream
opinion of the ninth-century Kalām is that the atom is unextended,82 which
later became a topic where the Basra and Baghdad schools of the Mu‘tazila
diverged.83 Furthermore, the accidents, for the Mutakallimūn, can last only a
single instant, while the substance cannot exist without accidents,84 from
which it follows that the atoms too (as elementary substances) are generated
and corruptible beings.85 Behind this transient character of atom-accidents is
an atomistic understanding of time.86 To the best of my knowledge, Greek
philosophers never went so far as to achieve such a thorough atomization of
both matter and time.
One of the main obstacles for a thoroughly atomistic understanding of
physical matter is the problem of contact. This objection is mentioned as early
as Aristotle. It can be summarized as follows: If the indivisibles are in contact
with parts of one another, they will be divisible; if they are in contact in
whole, they will merge together and take one and the same place.87 The
Mutakallimūn faced similar criticisms from the opponents of atomism.88 It is,
however, remarkable that the Kalām atomists did not address the general
question of atoms touching one another but focused on a more specific topic,
namely, one atom touching six other atoms.89 Abū al-Hudhayl, followed by
the Basra School, holds that one atom touches six other atoms so as to
constitute a three-dimensional body. On the other hand, Abū Ṣālih. rejects the
possibility of one atom in contact with two other atoms:
It is impossible for a third atom to join (yujmi‘) two others. Because each of the
two should occupy its counterpart. If one occupies it, the other will have no
position. For if the position of these two atoms are the same one, it (the third
Dong Xiuyuan 955

atom) will touch (māss) something larger than itself. If this is possible, it will be
possible to grasp the world in one hand. Therefore, he (Abū Ṣālih. ) said: it
cannot touch something larger than itself.90
Abū Ṣālih. ’s objection points to the basic structure of the atoms exposited
by Abū al-Hudhayl. For the latter, the six atoms are distributed around
another atom in three dimensions, and in each dimension, this central atom
is in contact with two other atoms. But Abū Ṣālih. only discusses one branch
of the question (i.e., the case of one atom touching two other atoms which
merge together), while his treatment of the other branch (the case of these
two atoms taking distinct positions) is missing. We find a full version of
this dilemma argument in Ibn Mattawayh’s report of the alleged opponents
of atomism:
The statement that one atom meets (yulāqī) six other atoms would be true only
if it meets them through its distinct sides (jihāt). If it meets them through one
single side, all the atoms will be on its right or its left or some side other than
these two. If this is not the case – namely, the sides are distinct – what this
atom meets is not that which the other atom meets. So it91 must be [a
compound of] multiple things. But this is impossible, for it is one single thing.92
Actually, this argument does not go against the concept of atom but only
denies the possibility of one atom touching more than one other atom. The
early Mutakallimūn of Basra, such as Abū al-Hudhayl, respond to the
objection by stating that the atoms touch one another through their sides
(jihāt), which are something other than the atoms themselves (ghayr al-
jawhar/al-’ajzā’), and that by the phrase something other than the atoms,
they mean some kind of geometrical boundary between the atoms, just like
the extremity (ṭarf) and the edge (h. add) of a body, which are something
other than the body.93
All the obscurities of the Mutakallimūn’s arguments mentioned above
will become clear, when we take into account the relevant points in
Sarvāstivādin atomism. The Sarvāstivādins assert that seven atoms constitute
a smallest body,94 which exactly corresponds to the one-touching-six
structure. It makes sense that the early Mutakallimūn learned this model
through the Buddhists they encountered (see below, Part IV) and then tried
to tackle its follow-up questions.
We find in both Vibhāṣā and Kośa that the Vaibhāṣikas treat the problem
of contact in a classical way:
The question is therefore posed whether the atoms do or do not touch one
another (spṛśanti; xiangchu 相觸). The Vaibhāṣikas of Kaśmir (Vibhāṣā, TD 27,
683a24) say that atoms do not touch one another; if atoms touch one another in
their totality, that is to say, the different atoms, would “mix with one another,”
that is, they would only occupy one place; and if atoms touched each other in
one spot, they would thus have parts: and atoms do not have any parts.95

Thus, the Vaibhāṣikas give a definite answer to the problem of touch:
the atoms cannot touch each other because they are impenetrable and have
no parts. There is no need for the Mutakallimūn to carry on discussing the
elementary question of two atoms touching one another, if they have taken
the framework of Sarvāstivādin atomism.
After establishing the impossibility of touch in its true sense, Vasumitra
makes a significant observation:
There is not, in reality, any contact (spṛśanti; 相觸). One says, metaphorically,
that atoms touch one another when they are juxtaposed without interval
(nirantara; wujian 無間).96
Vasubandhu approves this broad meaning of contact and elaborates it as

follows:
This opinion is the correct one. Otherwise, if atoms were to allow an interval
between themselves, since this interval would be empty, what would hinder the
progress of atoms into this interval? [There will be no interpenetration of the
atoms without interval,] for it is admitted that atoms are repulsive to each other
(sapratighā; youdui 有對) . . . If you admit spatial division to the atom, then an
atom certainly has parts, whether it enters into contact or not. If you deny it,
even if the atom enters into contact, there will be no such a problem.97
Vasubandhu explains the term nirantara as without anything including
space between the atoms. In this sense, it is possible for the indivisible
atoms to contact one another. As there is some kind of repulsive effect
between the atoms, they will not penetrate each other when juxtaposed
without interval.98 The state of nirantara is in fact a limit concept of
adjacency99: what remains between these atoms is just their contour lines
which, I tend to believe, are what Abū al-Hudhayl means when he speaks
of the sides through which the atoms touch one another.
As a response to Vasubandhu’s compromise on the possibility of contact,
Saṃghabhadra contends that the absolutely gapless adjacency is equivalent to
contact in its true sense, namely, overlapping with each other partially or
wholly, neither of which is acceptable. He reinterprets the term nirantara
(wujian 無間) as without one single atom of physical matter between the
atoms, and accordingly emphasizes that, as a result of repulsive force, there
must be space between the atoms adjacent (linjin 鄰近) to one another without
interval.100 It is not hard to see that Abū Sālih. shares the same stance with
Saṃghabhadra that the contact between the atoms is impossible in either sense.
What is more, the latter approaches the impossibility of multiple atoms taking
one position in a similar way when dealing with the problem of contact:
If it is possible for five atoms to share one position,101 why not hundreds,
thousands and millions? It is impossible for them to share the same position;
otherwise, the position of one atom will be able to contain all the atoms, and
then the world is equivalent to one atom.102
Dong Xiuyuan 957

In addition, the Sarvāstivādins also handle the question of touch from
another perspective:
The Bhadanta Vasumitra says: If atoms touched one another, they would
therefore endure two moments.103
Vasumitra’s statement presupposes that atoms cannot endure two

moments. It reflects a comprehensive atomistic worldview in which matter,
space, time, and even thought are all broken down into discrete indivisible
units:
An atom is the limit of physical matter (rupa); so too a syllable is the limit of
words, for example, go; and an instant, the limit of time (advan). What is the
dimension of an instant? If the right conditions (pratyaya) are present, the time
that it takes for a dharma to arise; or rather the time that it takes for a dharma in
progress to go [through the distance] from one atom to another atom.104
Therefore, the atoms as phenomena (dharma) come into being in one
moment and perish at once. They do not have a second moment to touch
each other. It seems that the Sarvāstivādins have gone so far in constructing
a thoroughly atomistic cosmology as to erase the distinction between the
atoms as substances and their accidents, a distinction which is fundamental
for the Mutakallimūn except Ḍirār ben ‘Umrū and his followers.105 This is
the reason why Pines preferred the Vaiśeṣika to the Sarvāstivāda among
Indian candidates for the origin of Islamic atomism.106 We shall explain this
point in what follows.
3. The distinction of substance and accident. Although the Sarvāstivādins
do not separate dravya (substance, shi 實) from guṇa (quality, de 德) and karma
(action, ye 業) in a Vaiśeṣika way, they have their own distinction between
substance and accident. The terms dravya and svabhāva (self-nature) are used
interchangeably to denote substance. The distinction between substance and
characteristic (lakṣaṇa, xiang 相) is discussed intensively in Vibhāṣā.107 Sub-
stance is the substratum in which lakṣaṇa appears, while one single substance
can bear different kinds of lakṣaṇas. In another context,108 the authors of Vibhāṣā
and Kośa present four types of categorical distinction adopted by the Vaibhāṣikas:
substance and mode (bhāva, lei 類), substance and characteristic (lakṣaṇa),
substance and state (avasthā, wei 位), and substance and relationship (anyo, dai
待). Vansubandhu finally inclines to the third one. He identifies state (avasthā)
with activity (kāritra, zuoyong 作用), revealing that the same substance can
exercise different activities in the three temporal modes (past, present, and future).
Furthermore, in the debates on the category theory in both books, the opponents
(mainly the Sautrāntika and their predecessors) raise the similar questions: given
the distinction of substance/characteristic (or substance/activity), what factor can
determine a certain characteristic (or activity)—rather than the alternatives—to
arise in the substratum.109 It is reminiscent of the crucial question of
particularization (takhṣīṣ) in Kalām.110

Hence, when the Sarvāstivādins state that the existence of the atom is
momentary, they are talking about its appearance or activity, not its substance.
It is worth noting that there is an affinity between the Vaiśeṣika and the
Sarvāstivāda in their logic and cosmology, which is a result of a long-term
dialogue between them.111 Just because of this similarity, the Sarvāstivāda treats
the Vaiśeṣika as a main rival among the non-Buddhist (especially Brahmanist)
sects. There are plenty of references to the Vaiśeṣika views in Sarvāstivādin
classical works from Vibhāṣā to Nyāyānusāra,112 for example, Vasubandhu’s
argument for the sensibility of atoms quoted above appears in a disputation
against the Vaiśeṣika’s doctrine on eternity of atoms.113 Therefore, the Vaiśeṣika
and the Sarvāstivāda, as sources of Islamic atomism, are not mutually exclusive.
It is possible for the early Muslims to have learnt both versions of atomism from
the Sarvāstivādin Buddhists. Furthermore, during the encounter of the Buddhists
and the Muslims (see below, IV), the former might have used the Vaiśeṣika, their
old rival, as a paradigm to understand and confront the challenges from the
monotheistic newcomer, particularly in view of the Vaiśeṣika’s theistic belief
that Maheshvara created the world with the atoms.114
IV. The Transmission Route
We have examined the argument of the Samaniyya, the accounts of the

Barāhima, and Islamic Atomism in Kalām, for all of which there are parallels
(the repudiation of ātman, rationalistic rejection of revelation, and the theory
of paramāṇu) in Sarvāstivādin Buddhist thought. Both trains of thought are
grounded on one and the same epistemological foundation that finds its
fullest expression in the Buddhist theory of pramāṇa. There is thus a
structural homology between the three views reported by the Mutakallimūn
and those of Sarvāstivādin Buddhism.
As we shall see in this section, the hypothesis of a historical link is
confirmed by the evidence of religious and intellectual contacts. Interestingly
enough, most of the evidence points to a single location: Balkh, the central
city of Tokharistan.115 Pines noticed this point and referred to the seventh-
century Chinese Buddhist pilgrim and translator, Xuan Zang’s travel report:
Speaking of the city of Balkh, he (Xuan Zang 玄奘) mentions the following
Buddhist buildings: a hundred monasteries, a temple and several hundred
Stūpas. Three thousand Buddhist monks studied the doctrine of the Hīnayāna
branch of Buddhism.116
In fact, this is not Xuan Zang’s only mention of Buddhism in Balkh. In

his biography, written by his disciples and assistants following his own oral
narrative, there is a more detailed description:
In the New Sanghārāma (納縛伽藍) there was (a priest) of the kingdom of Tcheka
who had studied the three Piṭakas belonging to Hīnayāna (小乘三藏); his name
Dong Xiuyuan 959

was Prajñākara. Hearing that there were many sacred traces of religion in the
country of Balkh (縛喝), he had therefore come to worship and reverence them.
This man was of singular wisdom and learning, so that as a youth he was
distinguished by his great sagacity. He had thoroughly sounded the nine
collections and mastered the four Ãgamas. The fame of his exposition of the
principles of the faith had spread throughout India. He was perfectly acquainted
with the Abhidharma of Hīnayāna: [Jñānaprasthāna written by] Kātyāyanīputra,
Kośa, the Shaṭpadābhidharma and other works. Hearing that the Master of the
Law had come from a distance to search for religious books, he was exceedingly
glad to meet him. The Master of the Law, in the course of his statement
respecting his doubts and difficulties about the Kośa and Vibhāṣā and other
books, asked him for some clarifications, and was answered in each case with
extreme lucidity. He remained here a month and studied the Vibhāṣā-śāstra.117
Xuan Zang’s report demonstrates that the teaching of Vibhāṣā (most likely
with Kośa) was alive and well in Balkh in the 620s, on the eve of the Arab
conquest (653–663/4).118 He consulted Prajñākara about Kośa and Vibhāṣā,
and studied Vibhāṣā for a month in the New Sanghārāma.119 Prajñākara was
a monk of Hīnayāna from Tcheka. All the classical works he mastered belong
to the Sarvāstivādins. We therefore have good reason to assume that he was a
Sarvāstivādin scholar. His pilgrimage to Balkh was not accidental. Bactria had
been a stronghold of the western branch of the Sarvāstivāda since no later
than the beginning of the Christian Era.120 Dharma-śresthin, the author of an
early Sarvāstivādin classical work, the Abhidharmahṛdaya-śāstra, is a native of
Balkh, and Ghoṣaka, one of the Four Masters of the Sarvāstivādins, also came
from Tokharistan.121 The archaeological evidence shows as well that the
Vinaya of Tokharistan belongs to the Sarvāstivādin school.122
In respect to Arabic reports, Ibn al-Nadīm provides an important clue for
the source of the Samaniyya:
The prophet of the Shamanīya is Buddha, and the majority of the people of the
land beyond the river (Transoxiana) were in accord with this doctrine before
Islam, in ancient times.123
This indicates that the Samaniyya are actually the Buddhists in

Transoxiana (Mā warā’ al-nahr), of which Tokharistan is a part. As
mentioned above,124 the Barāhima are closely related to the Sammaniyya
and are possibly the Buddhists who claimed to be “the true Brahmans”. An
account of Abū Mutī‘ al-Nasafī connects them to Balkh:
Abū Mutī‘ said: I had a disputation with the harābidha at Balkh; their
spokesman (mutakallimuhum) said that he does not recognize any prophet
except Ibrāhīm, for it is not fit for the Wise One to send messengers, then to
regret [his choice] and send others [in their stead].125
The point of view reported here is the third version of the Barāhima’s
rejection of Islamic prophecy. Abū Mutī‘ probably confused the learned

among the Buddhists with those of the Magians (the harābidha), because of
the syncretism of Buddhism and Zoroastrianism in Iranian Central Asia.126
After locating the encounter between the early Muslims and the
Buddhists, let us turn to the specific modes of transmission. There are two
complementary paths: one is inter-religious debate; the other is conversion.
Regarding the first way, we have Jahm’s disputation with the Samaniyya as
an example. According to Ibn Ḥanbal,127 Jahm came from Tirmidh in
Khurasan. The term Khurasan, in its early Islamic usage, denotes a vast
region east of western Persia.128 More precisely, Tirmidh (呾蜜) is a city of
Tokharistan. The inscriptions unearthed in Tirmidh reveal that it shared the
same Buddhist tradition with Balkh.129 Thus, the Samaniyya arguing with
Jahm are most likely to be the Sarvāstivādin Buddhists in Tokharistan. Pines
quoted Abū al-Ḥusayn al-Malatī to illustrate the influence of this debate
upon Jahm:
This is the sum of the discourse (Kalām) of the Jahmiyya. They were called after
Jahm, because Jahm b. Ṣafwān was the first who derived this discourse from
that of the Samaniyya, who are a subdivision (ṣinf) of the Iranians (al-‘ajam) in
the region of Khurasan.130
The Jahmiyya is not the only link between early Kalām and
the Buddhists. Al-Murtaḍā’s version of this debate and the disputation of al-
Naz. z. ām against the Samaniyya reported by al-Māturīdī show that the
Mu‘tazila were also exposed to the same challenge.131 In addition, al-
Baghdādī associates the Mu‘tazila with the Barāhima for their common
ground in religious rationalism and points out that the Mu‘tazila diverge
from the Barāhima on the possibility of prophecy.132
As for the transmission through the converts to Islam, Kevin van Bladel’s
study on the Barmakids’ Bactrian background sheds light on this point.133
The Barmakids, whose ancestor was the founder of the famous Naw
Bahār,134 came from Balkh. After conversion to Islam and entering the core
of power in the early Abbasid dynasty, they sponsored a series of
translations of Sanskrit scientific and literary works into Arabic and sent a
mission to India to learn about the religions there.135 It would be natural for
the Buddhist converts, represented by the Barmakids, to introduce their own
intellectual legacy, including epistemology and cosmology, into the nascent
Islamic theology.
By way of conclusion, we should return, as promised, to the question of
the transmission of Buddhist atomism to the Kalām physical theory. As
shown above, these two versions of atomism share the same assumptions,
focus, and even approaches to their problems. The impermanence and
limitedness of the world revealed by Buddhist atomism perfectly fits
the need of the Mutakallimūn to establish the absolute sovereignty of the
Creator. The difficulty that Buddhist thinkers faced in reconstructing the
natural and ethical order on the basis of a thoroughgoing atomism also left
Dong Xiuyuan 961

room for the Muslim theologians to introduce the transcendent cause. We
should nevertheless take into account the possibility that the atomism
transmitted by the Buddhists to Kalām might have contained theistic
elements from the Vaiśeṣika.
In light of all the above, I would like to propose a hypothesis as to why
the Indian or Central Asian sources have never been mentioned in the Kalām
discussions of atomism like in the Barāhima-Samaniyya narratives. When the
Mutakallimūn eventually developed their own theological–cosmological
system and accordingly their own intellectual identity, a process of rejection
began: they redrew the boundary between Islam and non-Islam, excluding
certain Buddhist elements (which they had assimilated) as pagan and heretical
thought. On the other hand, they chose to keep silent about the sources of
ideas that had already become integral parts of the framework of Kalām. I
suggest that atomism belongs to the parts whose origins were concealed,
while the thought reported in the names of the Barāhima and the Samaniyya
belongs to the parts rejected. Thus, the Barāhima–Samaniyya polemic
accounts appearing in the ninth and tenth centuries are products of an
ideological reconstruction and constitute a kind of etiological myth for Kalām.
Nonetheless, myth is never a creatio ex nihilo, but out of the prime matter of
historical experience, which remains to be detected by critical analysis.
Notes
1 – My thanks to Prof. Sarah Stroumsa, Prof. Lenn E. Goodman, Prof.
Steven Harvey, Prof. Frank Griffel, Prof. Binyamin Abrahamov, Prof.
David Shulman, Dr. Roy Vilozny, and Dr. Shalom Sadik for their
careful reading of an earlier draft of this article and for their valuable
remarks and suggestions. In addition, I would like to thank the Van
Leer Jerusalem Institute for providing me with ideal working conditions
to complete this article when I studied there as a Polonsky Academy
Fellow in 2015–2016. The research is supported by “The Fundamental
Research Funds of Shandong University” (No.: 2017TB0022).
2 – Daniel Gimaret, “Bouddha et les Bouddhistes dans la tradition
Musulmane”, Journal Asiatique 257 (1969): 291; Shlomo Pines, “A
Study of The Impact of Indian, Mainly Buddhist, Thought on Some
Aspects of Kalām Doctrines”, JSAI 17 (1994): 183, n. 6; Norman
Calder, “The Barāhima: Literary Construct and Historical Reality”,
Bulletin of the School of Oriental and African Studies, University of
London 57 (1) (1994): 47; Derryl N. Maclean, Religion and Society in
Arab Sind (Leiden: Brill, 1989), p. 5; E. C. Sachau, Alberuni’s India
(London: Kegan Paul, Trench, Trübner, 1910), xlv, p. 40.
3 – Shlomo Pines, Beiträge zur Islamischen Atomenlehre (Berlin: 1936,
Friedrich Wilhelms Universität, New York: Garland Publishing, 1987),

p. 132. On the debate between al-Naz. z. ām and the Samaniyya, see al-
Māturīdī, Kitāb al-Tawhīd, ed. Bekir Topaloghlu and Muhammed
Arushi (Beirut: Dar Sader, 2006), p. 221.
4 – al-Murtaḍā, Ṭabaqāt al-Mu‘tazila, ed. Diwald Wilzer (Wiesbaden:
F. Steiner, 1961), p. 34; Pines, “A Study of the Impact”, pp. 184–185.
The English translation is Pines’.
5 – Ibn Ḥanbal, al-Radd ‘alā al-zanādiqa wa-’l-jahmiyya, ed. S. Fawzān
(Kuwait: Gheras, 2005), pp. 196–198; my translation.
6 – F. Th. Stcherbatsky, Buddhist Logic, vol. 2 (An annotated English
translation of Dharmakīrti’s Nyāya-bindu, New York: Dover Publications,
1962), pp. 12–38; 法称 :《正理滴论》, 杨化群译, 《因明研究》
(1990.06.30): 368; 汤铭钧 : 《法称<正理滴论>梵汉对照和新译》, 《因
明》第二辑, 张忠毅编 (兰州 : 甘肃民族出版社, 2008), pp. 23–27. The
English translation by Stcherbatsky has been modified according to
汤铭钧’s edition of the Sanskrit text, and the serial numbers (1), (2), (3),
and (4) are added by me.
7 – Stcherbatsky, Buddhist Logic, vol. 1 (Delhi: Motilal Banarsidass
Publishers, 1993), pp. 222–223, 231; vol. 2, 405–406.
8 – Buddhist Logic, 1:506.
9 – Buddhist Logic, 2:26–33; According to the Tibetan tradition, the
Sarvāstivādin school did not accept the category of self-conscious
perception, see 章嘉·若贝多杰 : 《章嘉宗义·毗婆沙宗》, 牛宏译注,
《西北民族研究》, 04, (2006): 122.
10 – Buddhist Logic, 1:262–263; 2:104–107. By the term “sensible”,
I mean something that can be perceived by the senses; accordingly,
“nonsensible” means something that cannot be perceived by the senses.
11 – Ibn Ḥanbal’s attribution of this argument to the Christian doctrine of
incarnation is in fact irrelevant, because Jahm’s argument has nothing
to do with the spirit of God but only refers to human souls.
12 – Abhidharma-mahā-vibhāṣā-śāstra is no longer extant in the Sanskrit
original but only survives in the Chinese translations and the Tibetan
translation based on Xuan Zang’s (玄奘, 602–664) Chinese translation.
Because the three related Sarvāstivādin Abhidharmas (Vibhāṣā, Kośa, and
Nyāyānusāra) were all translated into Chinese by Xuan Zang and his
translation of the technical terms is basically word-for-word, in what
follows I will try to establish some of the Sanskrit terms behind the
Chinese terms in the translation of Vibhāṣā and Nyāyānusāra according
to the Sanskrit text of Kośa and Xuan Zang’s terminology in Chinese.
13 – 大正藏, 高楠顺次郎等辑 (东京: 大正一切经刊行会, 1922–1934), T27,
995c25-996a13; my translation.
Dong Xiuyuan 963

14 – 大正藏, T29, 152b29-c09; Abhidharma-Kośa-Bhāṣyam, trans. Leo M.
Pruden (Berkley: Asian Humanities Press, 1988–1990), 1313–1314.
The Chinese translation of Kośa was made by Xuan Zang. All quotes of
the English translation of Kośa hereafter are from Pruden’s, occasionally
modified according to the Chinese– Sanskrit parallel text edited and
published by Research Institute of Sanskrit Manuscripts and Buddhist
Literature, Peking University (http://www.mldc.cn/sanskritweb/etext.
htm).
15 – It is noteworthy that the Sammatiya was the dominant Buddhist school
in Sind before and during the Arab conquest in the seventh and eighth
centuries. For the situation of the Buddhists in Sind at the time of the
conquest, see Maclean, Religion and Society, 1–12. In addition, ‘Abd
al-Jabbār reports a disputation between a qādī and a man of the
Samaniyya at the court of Sind during the rule of Hārūn al-Rashīd, see
Pines, “A Note on An Early Meaning of the Term Mutakallim”, Israel
Oriental Studies, I (1971): 231–232. According to the seventh-century
Buddhist traveler Yi Jing (義凈, 635–713), Hetuvidyā and Kośa
constituted fundamental subjects of the curriculum at Buddhist mon-
asteries in India at the time (671–685). One of the two monasteries he
mentioned as examples is at Valabhi where most Buddhists belonged
to the Sammatiya school (义净: 《南海寄归内法传校注》, 王邦维校注,
北京:中华书局, 1995, 131–132, 198; 玄奘, 辩机: 《大唐西域记校注》,
季羡林等校注, 北京: 中华书局, 1985, 911). Both J. Takakusu and
Wang Guowei 王邦维 see the term xuesheng 學生 (the secular students
at Buddhist monasteries) in Yi Jing’s report as a translation of the
Sanskrit word brahmacārin (I-Tsing, A Record of the Buddhist Religion
as Practiced in India and Malay Archipelago, trans. J. Takakusu, Oxford:
Clarendon Press, 1896, 106; 《南海寄归内法传校注》, 132).
16 – See note 4 above.
17 – Such as Mahāsavghika’s mūla-vijñāna and Sarvāstivāda’s anuprāpti,
see 吕澄: 《印度佛学略讲》 (上海: 上海人民出版社, 2005), 61–62.
18 – See note 6 above, (3), (4).
19 – al-Muqammaṣ, ‘Ishrūn maqāla, ed., trans. and annot. Sarah Stroumsa
(Leiden: Brill, 1989), pp. 212–221.
20 – For the Kalām usages of the term shāhad/mushāhada, see Josef van Ess,
Theologie und Gesellschaft im 2. un 3. Jahrhundert Hidschra. Eine
Geschichte des religiösen Denkens im frühen Islam (Berlin: Walter de
Gruyter, 1991–1997), 5:72–73, 80–81; ‘Abd al-Jabbār, Ah. mad ibn al-
Ḥusayn Mānakdīm, Sharh. al-Uṣūl al-Khamsa, ed. A.K. ‘Uthmān (Cairo:
Maktaba Wahība, 1996), 51. For its later evolution in Arabic theological
and philosophical contexts, see Diana Lobel, Between Mysticism and

Philosophy (Albany: State University of New York, 2000), pp. 90–102;
Dimitri Gutas, “The Empiricism of Avicenna”, Oriens 40 (2012): 428–429.
21 – al-Muqammaṣ, ‘Ishrūn maqāla, 222–223, para. 23.
22 – Ibid., 222–225, para. 25, 26.
23 – Ibid., 224–225, para. 26, n. 35. This term appears earlier in Stroumsa’s
new edition of the text in a discussion of soul as a noncorporeal living
thing (Provo: Young Brigham University Press, 2016, 240–241).
24 – I suggest that al-Muqammaṣ might have had Hishām b. al-Ḥakam and
his followers in mind. According to al-Ash‘arī, Hishām claimed that
God is a body (jism) and has all kinds of perceived accidents, and that
if there is no any kind of similarity (tashābuh) between God and
perceived bodies (al-ajsām al-mushāhada), it will be impossible to
indicate/prove (dall) Him, see al-Ash‘arī, Maqālāt al-Islāmiyyīn, ed. H.
Ritter (Istanbul: Devlet Matbaasi, 1929–1930), p. 32. For a Mutakallim,
a body is definitely a substance. Moreover, Hishām’s view on analogy
and its relation to proof is exactly what underlies the opinions of the
objector here. Considering that al-Muqammaṣ mentioned the anthro-
pomorphists (al-mushābiha) at the beginning of Chapter 10 (‘Ishrūn
maqāla, 210–211), it is possible for him to have revisited some aspects
of the thought of the notorious representative of this group from
another perspective. Hishām is said to have been a Jahmite before
conversion to Shīʿism and in his later life belonged to the circle of
Yah. yā b. Khālid al-Barmakī (EI, 2nd ed., “Hish̲ ̲ām b. al-Ḥakam”, by
W. Madelung, 496–497). Thus, he had plenty of opportunities to come
into contact with Buddhist thought. For the connection of the
Barmakids with Buddhism, see note 133 below.
25 – Paul Kraus, “Beiträge zur islamischen Ketzergeschichte: das Kitāb al-
Zumurrud des Ibn al-Rāwandī”, RSO 14 (1934): 356. The intellectual
argument reported by Ibn al-Rāwandī runs as follows: given that reason is
the greatest gift bestowed by God on human beings, the prophecy is
either superfluous if it conforms with reason, or impossible if it contra-
dicts reason (Kraus, 1934, p. 97). His contemporary al-Muqammaṣ,
provides two variants of this argument based on the same premise
of autonomy of human reason, see al-Muqammaṣ, ‘Ishrūn Maqāla,
254–259.
26 – Bruce B. Lawrence, Shahrastānī on the Indian Religions (The Hague:
Mouton, 1976), pp. 81–83.
27 – Binyamin Abrahamov later found al-Qāsim’s account of al-Barhamiyya
in K. al-radd ‘alā al-rāfiḍa (“The Barāhima’s Enigma: A Search for a
New Solution”, Die Welt des Orients, Bd. 18, 1987: 73–74), which
supports Pines’ assumption of al-Ibrāhīmiyya as a scribal adaptation of
Dong Xiuyuan 965

al-Barāhima (“Shī‘ite Terms and Conceptions in Judah Halevi’s Kuzari”,
JSAI, 2, 1980: 222).
28 – Pines, “Shī’ite Terms and Conceptions in Judah Halevi’s Kuzari”,
220–223.
29 – Sarah Stroumsa, “The Barāhima in early Kalam”, JSAI 6 (1985):
229–241.
30 – Abrahamov, “The Barāhima’s Enigma”, 84–89; Calder, “The Barāhima:
Literary Construct”, 49–50.
31 – Stroumsa, Freethinkers of Medieval Islam (Leiden: Brill, 1999),
146–148; Idem, “Soul-searching at the Dawn of Jewish Philosophy:
A Hitherto Lost Fragment of al-Muqammaṣ’s Twenty Chapters”, Ginzei
Qedem 3 (2007): 137–161.
32 – Abrahamov, “The Barāhima’s Enigma”, 76.
33 – Sa‘adia, K. al-Amānāt wa’l-I‘tiqādāt, ed. S. Landauer (Leiden: Brill,
1880), p. 139; Ibid., ed. J. Kafih (Jerusalem: Sura, 1960), p. 143. For
the English translation, see Pines, “Shī’ite Terms”, 221; the translation
has been slightly modified.
34 – Sa‘adia does not specify the religious identity of his interlocutor, but
naskh is a typical Islamic notion in both exegesis and inter-religious
debate, see Hava Lazarus-Yafeh, Intertwined Worlds: Medieval Islam
and Bible Criticism (Princeton, 1992), pp. 35–41. Mordechai Cohen
notes that Sa‘adia attributes the naskh argument to the Christians later
in al-Amānāt, at the end of 8:9, see Mordechai Z. Cohen, “Maimo-
nides’ Attitude toward Christian Biblical Hermeneutics In Light of
Earlier Jewish Sources”, in New Perspectives on Jewish-Christian
Relations: In Honor of David Berger, ed. Jacob J. Schacter and E.
Carlebach (Leiden: E.J. Brill, 2012), p. 462. I take this move of Sa‘adia
as a device to avoid persecution from the Muslim authorities.
35 – This position is expressed positively by the Zaidite Imām al-Nāṭiq’s
(d. 1033) report of the Barāhima: “Their law is based on the intellect
(al-‘aql): they necessitate whatever the intellect necessitates, prohibit
whatever the intellect forbids, and approve whatever the intellect
approves.” (Camilla Adang, Wilferd Madelung, Sabine Schmidtke,
Baṣran Mu‘tazilite Theology: Abū ‘Alī Muh. ammad b. Khallād’s Kitāb
al-uṣūl and its reception – A critical edition of the Ziyādāt Sharh. al-
uṣūl by the Zaydī Imām al-Nāṭiq bi-l-h. aqq Abū Ṭālib Yah. yā b. al-
Ḥusayn b. Hārūn al-Buṭh. ānī (Leiden: Brill, 2011): 135; my translation).
36 – Actually, the Muslim theologians were trying to integrate the two
versions of Brahmin thought they received from different sources.
37 – Stroumsa, “The Barāhima”, 236, 241.

38 – See notes 27, 28 above.
39 – In Bāqillānī’s account of the Barāhima, all these three opinions (the
rationalist, the Adamist, and the Abrahamist) appear, see Abrahamov,
“The Barāhima’s Enigma”, 75. Besides, Shahrastānī notes the “error” to
trace the name Barāhima back to Ibrāhīm (Abraham), see Lawrence,
Shahrastānī on the Indian Religions, 38.
41 – al-Muqammaṣ, Twenty Chapter, 254–257.
42 – Ibid., 260.
43 – 大正藏, T29, 63c16-c23; Pruden, 480–481.
44 – 大正藏, T29, 141a13-17; Pruden, 1143–1144.
45 – Stcherbatsky, Buddhist Logic, 1:65–67.
46 – Samayabhedo paracanacakra śāstra, 大正藏, T49, 16c07-09.
47 – Stcherbatsky, Buddhist Logic, 1:74; 2:12–13; 法称: 《正理滴论》, 368.
48 – For the Mutakallimūn’s endeavor to infer the unseen (al-ghā’ib) from
the perceptible (al-shāhid), see Josef van Ess, “The Logical Structure of
Islamic Theology”, in Logic in Classical Islamic Culture, ed. G. E. von
Grunebaum (Wiesbaden: O. Harrassowitz, 1970), pp. 34–35.
49 – Kraus, “Beiträge zur islamischen”, 353; Ibn Bābūya, Kamāl al-Dīn, ed.
Husein al-A‘lamī (Beirut: al-Mu’sasat al-A‘alā li-l-Maṭbū‘āt, 1991), p. 88.
50 – For discussion of hearing (revelation and tradition) as a source of
knowledge, see J.R.T.M. Peters, God’s Created Speech (Leiden: Brill,
1976), pp. 95–100; Abrahamov, Islamic Theology (Edinburgh:
Edinburgh University Press, 1998), pp. 12–13.
51 – Adang, Madelung, Schmidtke, Baṣran Mu‘tazilite Theology, 136; my
translation.
52 – Ibid., 137 ff., notably 150, 152.
53 – V. Minorsky, “Gardīzī on India”, Bulletin of the School of Oriental and
African Studies, University of London 12 (3/4) (1948): 630, 633; al-
Bazdāwī, Kitāb ’Uṣūl sl-Dīn, ed. Hans Peter Linss (Cairo: Dār Khayā’
al-Kutub al-‘Arabiyya, 1963), pp. 90–91. For the English translation,
see Calder, “The Barāhima: Literary Construct”, 47.
54 – Gimaret, “Bouddha et les Bouddhistes dans la tradition Musulmane”,
291–292.
55 – Encyclopaedia Iranica, XV, “Kartīr”, by P. O. Skjærvø, 611.
56 – Lawrence, Shahrastānī on the Indian Religions, 76–77.
Dong Xiuyuan 967

57 – al-Maqdisī, Kitāb al-bad’ wa-’l-tārīkh (Cairo: Maktaba al-thaqāfa wa-’l-
dīniyya, 1980), I: 197–199.
58 – 大正藏, T27, 400b19-c03; the Chinese translation is made by Xuan
Zang. The English translation is mine.
59 – 大正藏, T04, 798a01-799c03. For the English translation, see F. L.
Woodward, The Minor Anthologies of the Pali Canon (London: Oxford
University Press, 1948), pp. 1–12.
60 – Lawrence, Shahrastānī on the Indian Religions, 41,108; Stroumsa, “The
Barāhima”, 240.
61 – Pinpointing Islamic atomism’s originality, Josef van Ess still suggests us
to look to Iran for its possible sources but provides no substantial
evidence, see van Ess, Theologie und Gesellschaft, 4:459; “Sixty years
after: Shlomo Pines’s Beiträge and half a century of research on
atomism in Islamic theology”, Proceedings of the Israel Academy of
Sciences and Humanities vii (2002): 31–32. It is remarkable that van
Ess notes the epistemological axiom underlying Islamic atomism
(Theologie und Gesellschaft, 1:452, 2:14, 397–399, 4:645; “Sixty years
after”, 28–31), which he calls sensualism (I rather preferred the term
analytical empiricism, see note 80 below). He also points out that the
Mu‘tazilites accused the Zanādiq of this view while it is actually not a
typical stance of the Manicheans (“Sixty years after”, 31). It would be
understandable if we take into account the fact that the Manicheans
adopted considerable elements of Buddhism before and during this
period, see Samuel N. C. Lieu, Manichaeism in Mesopotamia and the
Roman East (Leiden: Brill, 1997), pp. 25, 35, 259–260, 263; David A.
Scott, “Manichaean Views of Buddhism”, History of Religions 25 (2)
(1985): 104–110.
62 – Maimonides, The Guide of the Perplexed, I 71, trans. S. Pines
(Chicago: The University of Chicago Press, 1963), p. 178; T. J. De
Boer, History of Philosophy in Islam (London: Luzac & Co. Ltd, 1903),
pp. 57–58; H.A. Wolfson, The Philosophy of the Kalam (Cambridge,
MA: Harvard University Press, 1976), pp. 466–471, 727–729. It is
interesting to note that Richard Sorabji favors the Greek origin of
Islamic atomism but does not exclude the possible Indian (especially
the Vaiśeṣika) influence, see R. Sorabji, Time, Creation and the
Continuum (London: Duckworth, 1983), pp. 397–400.
63 – Pines, Beiträge zur Islamischen Atomenlehre, p. 99; Wolfson, Philosophy
of the Kalam, pp. 472–486.
64 – Pines, Beiträge zur Islamischen Atomenlehre, p. 98.

65 – Alnoor Dhanani, The Physical Theory of Kalām (Leiden: Brill, 1994),
pp. 100, 191.
66 – For a discussion and a case study of the methodological limit of this
approach, see Michael Schwarz, “Can We Rely on Later Authorities for
the Views of Earlier Thinkers”, Israel Oriental Studies I (1971): 241–248.
67 – Dhanani, The Physical Theory, pp. 191–192.
68 – Tzvi Langermann, “Islamic Atomism and the Galenic Tradition”, Hist.
Sci. xlvii (2009): 277–291.
69 – Lenn E. Goodman, “How Epicurean was Rāzī”, Studia graeco-arabica,
5 (2015): 252–257.
70 – Ibid., 256.
71 – Wolfson, Philosophy of the Kalam, pp. 468–471.
72 – Pines, Beiträge zur Islamischen Atomenlehre, pp. 102–123.
73 – al-Ash‘arī, Maqālāt, pp. 302, 311–312, 315–318; al-Naysābūrī,
K. Masā’il fi al-Ikhtilāf bayn al-Basriyyīn wa-l-Baghdādiyyīn, ed.
A. Biram (Berlin: H. Itzkowski, 1902), pp. 43–44; al-Baghdādī, K. Uṣūl
al-Dīn (Istanbul: 1928), pp. 33, 41–42, 56–57. According to Ibn
Mattawayh, most of the Basrian Mu’tazila held that atoms are by their
nature perceptible so that we can know them necessarily, see Ibn
Mattawayh, al-Tadhkira fī Ah. kām al-Jawāhir wa-l-a‘rāḍ, ed. Daniel
Gimaret (Cairo: Institut français d’archéologie orientale, 2009),
pp. 12–13, 87–88; see also Dhanani, The Physical Theory, pp. 141–
145. Besides, the consensus among the Mutakallimūn that God sees
the individual atom (al-Tadhkira, 87) is reminiscent of the Buddhist
doctrine that the atoms can be seen by the divine eye (divyaṃ cakṣuh. ;
天眼. Vibhāṣā, 大正藏, T27, 702a10-11,684a18-21).
74 – Vibhāṣā, 大正藏, T27, 64a26-b01; Kośa, 大正藏, T29, 66c17-22,
Pruden, 493–494; Abhidharma-nyāyānusāra-śāstra (the fifth century
C.E., hereafter referred as Nyāyānusāra), 大正藏, T29, 351a.
75 – Vibhāṣā, 大正藏, T29, 702a6-10; my translation.

76 – Kośa, 大正藏, T29, 66c12, Pruden, 493.
77 – By the term homogeneous, I mean that the whole and the parts share
the same characteristics. In Nyāyānusāra, Saṃghabhadra distinguishes
two kinds of aggregation (和集 and 和合) according to the criterion of
homogeneity (i.e., whether the parts have the sensible characteristics
manifested by the whole), see 大正藏, T29, 350c19-351a29.
78 – Nyāyānusāra, 大正藏, T29, 522a06-09; The Chinese translation is
made by Xuan Zang. The English translation is mine.
Dong Xiuyuan 969

79 – 大正藏, T29, 351b05-07.
80 – It should be noted that perceptibility is a necessary condition for reality
but not the unique criterion. There is another complementary principle,
namely, the affordability of analysis: When something is divided
physically or mentally, if its idea (manifested in a set of perceptible
characteristics) remains, its existence is real; if its idea disappears, its
existence is virtual (Kośa, T29, 116b11-b26; Pruden, 910–911). This
concept of reality may be characterized as “analytical empiricism”.
81 – Sharh. al-Uṣūl al-Khamsa, pp. 48, 51–52, 57; al-Nāṭiq, Baṣran
Mu‘tazilite Theology, 136; see also Pines, “A Study of the Impact”,
189–193.
82 – Maqālāt, pp. 307–308, 315–316.
83 – Masā’il, 38, 41–42. I tend to consider the opinion held by the Basra
school, which is focus of Dhanani’s study (The Physical Theory,
100–113), as a later development due to the influence of the Falsafa.
84 – Maqālāt, pp. 358, 570–571.
85 – al-Muqammaṣ, ‘Ishrūn maqāla, 106–109; Bāqillānī, K. Tamhīd, ed.
Ahmad Ḥaidar (Beirut: Mu’sasa al-Kutub al-Thaqāfiyya, 1987), pp. 41–43.
86 – Maimonides, The Guide, I 73, 196–198; For Maimonides’ possible
sources on this issue, see Michael Schwartz, “Who Were Maimonides’
Mutakallimun”, part 1, Maimonidean Studies 2 (1991): 176–182.
87 – Aristotle, Physics, 6.1, 231b1 ff.
88 – For discussions on al-Naz. z. ām’s and the Arabic Aristotelians’ criticism
of atomism from this perspective, see Pines, Beiträge zur Islamischen
Atomenlehre, 8, 11; van Ess, Theologie und Gesellschaft, 3:309;
Dhanani, The Physical Theory, 167.
89 – Maqālāt, pp. 302–303; Ibn Mattawayh, al-Tadhkira, pp. 72, 81–82, 89.
90 – Maqālāt, p. 302. Pines, Beiträge zur Islamischen Atomenlehre, p. 8.
91 – It is the atom that touches two or six others.
92 – Ibn Mattawayh, al-Tadhkira, p. 89.
93 – Ibn Mattawayh, al-Tadhkira, p. 89. According to al-Naz. z. ām, the
exponents of the theory of one atom touching six others hold that the
sides are accidents of the atom but not the atom itself, see Maqālāt,
p. 316. For a discussion on the divergence between the positions of
the early and the late Basrian Mu‘tazila, see Dhanani, The Physical
Theory, pp. 168–169.

95 – Kośa, 大正藏, T29, 11c4-7, Pruden, 120.
96 – Ibid., T29, 11c24, Pruden, 121.
97 – Ibid., T29, 11c25-29, Pruden, 122.
98 – Saṃghabhadra develops the notion of repulsion (有對) into the concept
repulsive force (有對勢力/有對力) by which the adjacent atoms affect
one another without touch, see 大正藏, T29, 373b27-c03.
99 – It is remarkable that Ḍirār understands the aggregation of the
accident-atoms as adjacency of the side (mutajāwira al-ṭaff) but not
interpenetration (mudākhala), see Maqālāt, 318. For a debate among
the later Mutakallimūn on the definition of composition (ta‘līf) as
adjacency, see Ibn Mattawayh, al-Tadhkira, pp. 290–291.
100 – Nyāyānusāra, 大正藏, T29, 373b13-27. The divergence on the mean-
ing of nirantara between Vasubandhu and Saṃghabhadra might also
have set the stage for the disputation among the Mu‘tazilites on the
possibility of an atom taking its place on the junction between two
other atoms, see Masā’il, p. 84. And when discussing the latter
question, the Mu‘tazilites also referred to the impossibility of
interpenetration (tadākhul) of the atoms, see al-Tadhkira, p. 91.
101 – This statement is an inference from two propositions of some
Vaibhāṣikas that Vasubandhu referred to in Kośa: Atoms themselves
are not impenetrable (T27, 390a02; T29, 3c02-03) and One atom of
generated matter is dependent on a tetrad of primary elements (T27,
663c07; T29, 3c14).
102 – Nyāyānusāra, 大正藏, T29, 372b19-21.
103 – Kośa, 大正藏, T29, 11c22-23, Pruden, 121.
104 – Ibid., T29, 62a18-22, Pruden, 474.
105 – Maqālāt, pp. 305–306, 316–318.
106 – Pines, Beiträge zur Islamischen Atomenlehre, p. 112.
107 – 大正藏, T27, 201b23.
108 – 大正藏, T27, 396a13-b23; T29, 104b28-c29, Pruden, pp. 808–810.
109 – 大正藏, T27, 202a18; T29, 105a09-16.
110 – Maimonides, The Guide, I 74, 218–219. For the earlier sources of
this argument, see Michael Schwartz, “Who Were Maimonides’
Mutakallimun”, part 2, Maimonidean Studies 3 (1992–1993): 159–160.
111 – 吕澂: 《印度佛学源流略讲》, 133; 木村泰贤: 《小乘佛教思想论》,
演培译, 《谛观全集》译述二 (台北: 天华出版公司, 2008), 183; Arthur
B. Keith, Indian Logic and Atomism (Oxford: Clarendon Press, 1921),
pp. 44, 121–122, 148, 168–169, 182–184, 192, 207, 212, 232, 234.
Dong Xiuyuan 971

112 – 大正藏, T27, 587a01, 729c27; T29, 24b02, 400c02, 408a28, 541b13.
113 – 大正藏, T29, 66b07-66c23.
114 – Keith, Indian Logic, pp. 214–218. This may offer an explanation for al-
Muqammaṣ’s statements that there are monotheists (muwah. h. iduhum)
among the Buddhists and that the Barāhima falls within the category of
monotheists, see ‘Ishrūn maqāla, pp. 224–225, 254–255. Al-Baghdādī
also deems the Barāhima as monotheists, see K. Uṣūl al-Dīn, p. 26.
115 – Another possible location of the encounter is Sind, which requires
further study, see note 15 above.
116 – Pines, “A Study of the Impact”, p. 182.
117 – 慧立、彦悰: 《大慈恩寺三藏法师传》 (北京: 中华书局, 2000), pp. 32–
33; Idem, The Life of Hiuen-Tsiang, trans. Samuel Beal (London: Kegan
Paul, 1911), pp. 50–51. The translation is slightly modified.
118 – EI, 2nd ed., “Balkh”, by R. N. Frye, 1001. According to Huichao (慧
超) who visited Tokharistan in 720s, the residents there still adhered
to Hīnayāna Buddhism under the rule of the Arabs at the time when
he traveled, see 慧超: 《往五天竺国传笺释》 (张毅笺释, 北京: 中华书
局, 2000), p. 96.
119 – According to Yi Jing’s 《大唐西域求法高僧傳校注》 (王邦维校注, 北
京: 中华书局, 1988, 11), the name of this “New Monastery” is “納婆
毗訶羅” (nabo-biharo), which is doubtless a transliteration of “Nava
Vihāra”, just as Naw Bahār. Richard Bulliet (“Naw Bahār and Survival
of Iranian Buddhism”, Iran 14, 1976:140–146) observed that there are
a number of places bearing the name Naw Bahār in Khurasan and
Central Asia, which reveals a route of Buddhist transmission from east
to west. He assumed that the spread of Buddhism from Balkh into
the Sassanid Persia was initiated by some specific sect of Buddhism.
I tend to associate it with the Sarvāstivāda, for Xuan Zang reported
that during the time of his travel there were still two or three
Sanghārāmas and hundreds of monks in Persia, who were all
Sarvāstivādins (《大唐西域记校注》, 1939).
120 – Charles Willemen, Bart Dessein, Collett Cox, Sarvāstivāda Buddhist
Scholasticism (Leiden: Brill, 1998), pp. 102–105.
121 – 普光: 《俱舍论记》, 大正藏 T41, 11c13; Tāranātha, History of
Buddhism in India, trans. Lama Chimpa and Alaka Chattopadhyaya
from Tibetan (Delhi: Motilal Banarsidass Publisher, 1990), p. 96.
122 – B. A. Litvinsky, “Outline History of Buddhism in Central Asia”, in
Kushan Studies in USSR (Calcutta: Indian Studies, 1970), p. 121.
123 – Ibn al-Nadīm, K. Fihrist, ed. S. A. Fuād (Cairo: al-Furqan, 2009),
3:422; The Fihrist of al-Nadīm, trans. Bayard Dodge (New York:
Columbia University Press, 1970), pp. 824–825.

125 – Stroumsa, Freethinkers, p. 151.
126 – According to the archaeological discoveries in this area, Buddha was
portrayed as the monarch of the universe after the model of Ahura
Mazda, see Melikian-Chirvani, “The Buddhist Ritual in the Literature
of Early Islamic Iran”, in South Asian Archaeology 1981, ed. Bridget
Allchin (Cambridge: Cambridge University Press, 1984), pp. 276–277;
David Alan Scott, “The Iranian Face of Buddhism”, East and West 40
(1990): 60. Moreover, the Kara-tepe Buddhist inscription in Tirmidh
(the third to fourth century C.E.) combines the name of Buddha and
Mazda, reading as Buddha–Mazda, see Boris Stavinsky, “‘Buddha-
Mazda’ from Kara-tepe in Old Termez (Uzbekistan): A Preliminary
Communication”, The Journal of the International Association of
Buddhist Studies 13 (2) (1980): 91. In addition, the report on the
Barmakids in Faḍā’il-i Balkh also mistakes the famous Buddhist
temple Naw Bahār as a sanctuary of the Magians (Mughān), see
Arezou Azad, Sacred Landscape in Medieval Afghanistan (Oxford:
Oxford University Press, 2013), pp. 78–79.
128 – EI, 2nd ed., “Khurāsān”, 55–56, by C. E. Bosworth.
129 – 玄奘, 辩机: 《大唐西域记校注》, 季羡林等校注, 103, 104 注 (一).
130 – Pines, “A Study of the Impact”, p. 184.
132 – K. Uṣūl al-Dīn, p. 26.
133 – Kevin Van Bladel, “The Bactrian Background of the Barmakids,” in
Islam and Tibet: Interactions along the Musk Routes, ed. Anna
Akasoy, Charles Burnett, and Ronit Yoeli-Tlalim (Farnham: Ashgate
Publishing, 2011), pp. 43–88.
134 – See note 119. If the thesis of the link of Vihāra and Madrasa in
Tokharistan (Litvinsky, “Outline History”, 124–130) and the possible
influence of the Sarvāstivādin dialectical method upon Medieval
Arabic and Latin scholasticism (Christopher I. Beckwith, Warriors of
the Cloisters: The Central Asian Origins of Science in the Medieval
World, Princeton: Princeton University Press, 2012, 50–75) could be
proved by more concrete evidence, it would be great support for the
viewpoint proposed here, though I have reservations about Beckwith’s
model of “recursive argument”.
135 – Van Bladel, “The Bactrian Background”, pp. 74–85.
Dong Xiuyuan 973

The Presence of Buddhist Thought in Kalām Literature

Uploaded by

Document Informationclick to expand document information

Document Informationclick to expand document information

Copyright:

Available Formats

The Presence of Buddhist Thought in Kalām Literature

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Presence of Buddhist Thought in Kalām Literature

Uploaded by

Copyright:

Available Formats

The Presence of Buddhist Thought in Kalām Literature

Published by University of Hawai'i Press

For additional information about this article

Access provided by Peking University (8 Aug 2018 14:44 GMT)

This paper1 is intended to examine the accounts of Buddhist thought in

I. The Samaniyya on Epistemology

It is generally accepted that the term Samaniyya (or Shamaniyya,

An earlier version of the same story is reported by Ibn Ḥanbal:

Its (perceptual knowledge’s) object is the particular . . . that alone [which is

Dong Xiuyuan 945

946 Philosophy East & West

Dong Xiuyuan 947

II. The Barāhima on Prophecy

948 Philosophy East & West

These arguments are—may God direct you well—without foundation. It is they

This paragraph appears in the context of a debate on abrogation (naskh)

Dong Xiuyuan 949

950 Philosophy East & West

Dong Xiuyuan 951

As for the path to knowledge (ma‘rifa) of the messenger, it is either necessary

In what follows, the Muslims try to established hearing/report as a path

Moreover, in the Sarvāstivādin edition of Udāna, Buddha redeﬁnes the

952 Philosophy East & West

III. The Origins of Islamic Atomism

The doctrine of atom and its accidents (jawhar wa-a‘rāḍuhu) is the

Dong Xiuyuan 953

Saṃghabhadra (the ﬁfth century C.E.), representative of the so-called

954 Philosophy East & West

Dong Xiuyuan 955

956 Philosophy East & West

Vasubandhu approves this broad meaning of contact and elaborates it as

Dong Xiuyuan 957

Vasumitra’s statement presupposes that atoms cannot endure two

958 Philosophy East & West

IV. The Transmission Route

We have examined the argument of the Samaniyya, the accounts of the

In fact, this is not Xuan Zang’s only mention of Buddhism in Balkh. In

Dong Xiuyuan 959

This indicates that the Samaniyya are actually the Buddhists in

960 Philosophy East & West

Dong Xiuyuan 961

962 Philosophy East & West

Dong Xiuyuan 963

964 Philosophy East & West

Dong Xiuyuan 965

966 Philosophy East & West

Dong Xiuyuan 967

968 Philosophy East & West

75 – Vibhāṣā, 大正藏, T29, 702a6-10; my translation.

Dong Xiuyuan 969

970 Philosophy East & West

Dong Xiuyuan 971

972 Philosophy East & West

Dong Xiuyuan 973

You might also like