`Atta, Niratta`, and `Anatta` in the early Buddhist literature
中華佛學學報第六期 1993.07 出版
(pp. 391-424)
Biswaded Mukherjee
Visiting Professor,
Chung-Hwa Institute of Buddhist Studies
391 頁
`Atta, Niratta`, and `Anatta` in the early Buddhist
literature
Biswadeb Mukherjee
Visiting Professor,
Chung-Hwa Institute of Buddhist Studies
Summary
The present article deals with the non-Buddhist
concepts of `atta` and their refutaion through the
doctrine of `anatta` . This article is divided into
five sections.
Section I discusses `Nikaya` passages which
states that the `atta` is not existent either in the
sphere of the Conditioned or in the realm of the
Unconditioned. It is nothing but a figment of
imagination.
Section II refers to the different opinions
held by scholars as to the proper implications of
the term `atta` . Except for the lone voice of E.
Conze who found similarity between the `atta` and
the `Purusa` of the `Samkhya` , all other scholars
who cared to ponder over this problem, seem to be
preoccupied with the relationship that could exist
between the `atta` and the `atman` of the `Upanisads` .
This may be due to the fact that these two
terms-- `atta` and `atman` --are philologically
identical. The philological identity led the
scholars readily to accept the view that these terms
are also philosophically identical without making
any further investigation into the exact
connotations of these two terms. The need for a more
objective study of the problem of equation of `atta`
and `atman` is obvious, and a suitable methodology
for such a study has been suggesed.
Section III delineates the main features of the
thoughts of the `Sasvatavadins` , `Ekaccasassatavadins` ,
the `ucchedavadins` and the followers of
the `Satkayadrsti` . They were the chief
propagators of the `atta` heresies rejected by the
Buddha.
Section IV, Part A contations the criticism of
all these heresies meant for the Buddhist monks. The
validity of such a criticism mainly rests on the
Buddha's claim to superior and higher knowledge of
which the heretics know nothing. Part B of the same
Section offers us the following two patterns of
reasonings which were ultimately meant for the
non-Buddhist ascetics and employed for the
refutation of the different `atta` heresies except
the `Ucchedavada` : i) To reject the opponent's view
by showing the internal contradictions, a method
which was later followed by `Nagarjuna` with
great success; ii) To temporarily accept a part of
the opponent's view in order to show that the view
as such is unacceptable. This may be taken as an
instance of skillfulness of means praised so much in
the `Mahayana` .
Section V comes to the conclusion that the
`anatta` doctrine was mainly formulated keeping in
view the `sasvatavada` and the `Ucchedavada` .
This is in conformity with some statements in the
buddhist texts that the buddha preached the `anatta`
doctrine as a middle way between these two extremes.
It is also clear that the `atta` concepts of the
`sasvatavadins` and the `Uccedavadins` cannot be
identical with the `atman` of the `Upanisads` . The
Buddhist sages were aware of the fact that the `atta`
rejected by the buddha is ' `ahankara-visayah` '
whereas the `Upanisadic Atman` is ' `Buddhanamvisayah` '.
393 頁
I
The doctrine of `anatta` forms the keynote of
the teachings of the Buddha and literally means that
the `atta` is non-existent. The denial of the `atta`
finds expression in the following statment ascribed
to the Buddha in the Samyutta `Nikaya`: "Sabbam
bhikkhave `aniccam/sabbam` bhikkhave
`dukkham/sabbam` bhikkhave `anatta`".(1) The chapter
just preceeding the one from which we have quoted
the three statements, contains such expressions as
"sabbe bhikkhave `jaradhammam`......`vyadhidhammam`
......`maranadhammam`......`samkilesadhamam`" etc.
(2) These statements leave no doubt that in all
these cases we have to take the term 'sabba' as
indicating the things condition. So we find that
'sabba' is a technical term and stands for the
'`samskrta dharma'`. This conclusion of ours is
confirmed by the definition of 'sabba' as given in
the Sabbavaggo of the `Salayatana-samyutta`(3):
"`Kin` ca bhikkhave `sabbam/cakkhu` ceva `rupa`
ca/`sotanca sadda` ca/`ghananca gandha` ca/`jihva
rasa` ca/kayo ca `photthabba` ca/mano ca `dhamma`
ca/idam vuccati bhikkhave `sabbam`" According to
this definition 'sabba' means eye and the foms, the
nose and the smells, the ear and the sounds, the
tongue and the tastes, the body and the tangible
things, the mind and the mental objects. Here the
term 'sabba'means twelve `ayatanas`. In a more
elaborate way the Buddha explains the same term
while giving instructions on the giving up of the
'sabba'. He states that eye, form, eye-contact,
eye-consciousness and the different feeling due to
eye-contact are to be renounced. Similarly he
preaches with references to other sense organs(4).
Thus the term 'sabba' fully covers the category of
'samskrta-dharma' and the statement "`sabbam
anatta`" means the conditioned is without any soul
────────────
(1) S. Vol.Ⅳ,p.28,ⅩⅩⅩⅤ,43-45(PTS,London, 1894)
(2) Ibid ⅩⅩⅩⅤ,34-38
(3) Ibid,p.15, ⅩⅩⅩⅤ,23
(4) Katamo ca bhikkhave `sabbapahanayo
dhammo/cakkhum bhikkhave `pahatabbam`/`rupa
pahatabba`/`cakkhuvinnanam pahatabbam`/
cakkhusamphasso `pahatabbo`/yam `pidam
cakkhusamphassopaccaya` uppajjati `vedayitam
sukham va dukkham va` adukkham `asukham va`/
tam pi `pahatabbam`/(`Samyutta`, Ⅳ, pp15-16,
ⅩⅩⅩⅤ, 24 (PTS, London, 1894)
394 頁
(soul).
But what about the Unconditioned？ Is it also
devoid of `atta`？ Walpolo Rahulo thinks that it is,
and refers to the three statements from Dhammapada
in support of his contention(5). These statements
which are very similar to the passage from the
`Samyutta Nikaya` quoted above are as follows： i)
sabbe `sankhara dukkha`, ii) sabbe `sankhara anicca`.
iii), sabbe `dhamma anatta`. Rahulo thinks that while
the phrase sabbe `sankhara` indicates the things
conditioned, the phrase sabbe `dhamma` refers to
both the Conditioned and the Unconditioned. But it
is difficult to accept this interpretation. We may
hold that the word 'sabbe' should have the same
meaning as the term 'sabbam'. viz. the conditioned
things, and it is unreasonable to assume that the
scope of the word 'sabba' while occuring together
with the term 'dhamma' is so extended as to include
even the opposite meaning, the Unconditioned.
Moreover the phrase 'sabbe dhamma' has been used
many times in the Buddhist scriptures to mean only
the Conditioned. The Karatalaratna, a comparatively
late work written by `Bhavaviveka`, described
'`sarvadharmah`' as being of dependent origin
(`sarvadharmah`......`pratityasamutpannah`......`nih-
svabhavah`).(6) Again in the Culasaccakasutta of the
Majjhima Nikaya (7) it is stated： "`rupam bhikkhave
aniccam, vedana anicca`,......`vinnanam aniccam`,
`rupam bhikkhave anatta`, `vedana`......`sanna`......
`samkhara`......`vinnanam anatta`, sabbe `samkhara
anicca`, sabbe `dhamma anatta` ti. "Here both the
phrases 'sabbe dhamma`, and sabbe `samkhara`
obviously refer to the five skandhas. That only the
conditioned things are meant here become evident
when in the same sutta (8) we meet with a discussion
on things which are both `anicca` and `anatta`.
`Nibbana` is not included in this discussion. The
`Upasivamanapuccha` of the Sutta `Nipata`(9)(1076)
states that in case of `nibbana` all dharmas are
destroyed (sabbesu
────────────
(5) Rahulo, What the Buddha taught (Grove Press,
INC.New York, 1962), pp.57-58； Dhammapada,
Chapter 20, verses 5-7
(6) `Swami Vidyaranya`, Bauddha Dharma O `Darsan`,
p.16(West Bengal State Book Board, Calcutta,
1984)
(7) Majjhima, Vol.I,p.280(Nava Nalanda Ed.),
p.228(PTS Ed.)； also see, `Samyutta` Vol. Ⅲ,
p.132
(8) Ibid. 282-87(Nava Nalanda Ed.)
395 頁
dhammesu upahatesu). Sabbe `dhamma` has been
explained as skandhas and `ayatanas` in the
Cullaniddesa,(10) making it clear that `nibbana` is
outside the scope of this phrase. In the
`Mahaparinibbanasutta`(11) Aniruddha explains
buddha's parinibbana as 'vimokkho' of consciousness
(pajjotass'eva `nibbanam` vimokkho cataso). The term
'vimokkho' has been explained by Buddhaghoso in the
Sumangalavilasini(12) as freedom from all dharmas
(vimokkho'ti kenaci dhemmena `anabarana`). Thus the
testimony of the `Suttanipata` and the `Maha
parinibbanasutta` confirms that the phrase '`sabbe
dhamma`' does not include `nibbana within its scope.
So the things conditioned were definitely held
to be devoid of `atta`. But was the nirvana, the
Unconditioned also bereft of `atta`？ Is the `atta`
non-existent with reference to the Unconditioned
also？ In the `Udana`(13) `nibbana` is described as
`anatta`which is rightly glossed as `atta-virahita`
in `Paramatthadipam`. Thus the `atta` is
non-existent either as a `samskrta-dharma` or as an
`asamskrta-dharma`. The `atta` is nothing but a
figment of imagination.
Ⅱ
────────────
(9) PTS, London, 1913, The group of Discourses Ⅱ,
p.121 (PTS, London, 1992)
(10) Cullaniddesa, (PTS. London, 1988) p.177, also
see, The Group of Discourses Ⅱ p.376
(11) `Digha`, VolⅡ,p.157. The line "`pajjotasseva
nibbanam vimokkho cetaso`" is translated in
Dialogues of the Buddha, Pt. Ⅱ, p.176, as 'Even
as a bright flame dies away, so was the last
emancipation of his heart'. 'pajjota' is
translated as a lamp while '`nibbana`' is taken
to be blowing out of a lamp'. On the other hand
M.Walche understand 'pajjota' in the sense of
the 'Illumined One' and translate the same line
as 'By `Nibbana` the. Illumined's mind is freed'
(Thus Have I Heard, London, 1987； p.271). Both
these translations are not fully satisfactory
PTS translates 'eva' as 'like' while it would
have been better to render it as 'just, surely,
indeed'. Walche omits it. As 'pajjotassa
`nibbana`' is compared with 'vimokkho cetaso',
it would be more appropriate to translate as the
'nibbana of the Enlightened One'. So the sentence
may be rendered as follows： 'The `nibbana` of
the Enlightened one is, indeed, the release of
mind'.
(12) `Sumangala-vilasini`, PartⅡ, p.595(PTS, London,
1971)
396 頁
To know the reason behind the denial of `atta`
we have first to be clear about the exact
implication of the term `atta`. Scholars differ as
to the precise sense in which this term has been
used. But before their views were discussed it is
necessary to point out that the `prakrt` word `atta`
is the same as `atman` in Sanskrit. This
philological identification has led to philosophical
misunderstanding among the scholars. Some scholars
automatically take `atta` to be the `Atman` of the
`Upanisads`. They think that the philosophical
implications of these two terms-`atta` and
`atman`-are identical and consequently the doctrine
of `anatta` came to mean for them the refutation of
the `Upanisadic Atman`. The scholars belonging to
this
────────────
(13) P.Steinthal in his edition of the `Udana` (PTS,
London, 1982), p.80 adopts the following reading
of the relevant verse：
'duddasam anattam `nama`, na hi saccam
sudassanam `patividdha tanha janato`, passato
n'atthi `kincanan`' ti'.
In his translation F.L.Woodward reads 'anantam'
(infinite) in place of anattam (Verses of Uplift,
The Minor Anthologies of the Canon, Pt. Ⅱ, PTS
London, 1985, p.98). I prefer the reading
'anattam' due to the following considerations：
Steinthal gives the variant reading 'anattam'
given in the commentary (anatatan ti `pathati`)
called `Paramattha-dipani` which was copied in
Sinhalese script for the PTS. This transcript
which he calls 'C' often gives right information
and corrects the incorrect readings of the other
manuscripts (see, Stienthal, Udana, p.ⅤⅢ).
This commentary explains 'anattam' as
'`atta-virahitam`'. Another manuscript which he
calls 'B' as well as the Nava-Nalanda Ed. of the
`Udana` fallow this reading. There is no reason
why the reading 'ananttam' should not be
accepted. On the other hand it is difficult to
agree with Woodward who accepts the reading
'anantam', as it is found only in the
`Nidanatthakatha` of `Dhammapalacariya` (Ed.
F.L.Woodward, PTS, London, 1977, p.393).
`Nidanatthakatha` explains as follows：
'`niccasvabhavatta` anta-virahitam
amaranadhammam anirodham amatam ti attho'. 'As
nibbana is permanent by nature, it is without an
end, beyond death and cessation； it is
immortal'. It is obvious that this definition of
`nibbana` also excludes `atta`.
Actually speaking no scriptural evidence is
needed to prove that `nibbana` is devoid of
`atta`. The concept of `atta` is such that it is
automatically excluded from the domain of
`nibbana`.
397 頁
group have an unfaltering faith in the Buddha as the
unique preacher of the True Law, and as such harbour
no weakness for the honoured traditions of the
`Upanisads`. The view of Walpolo Sri Rahulo, an
worthy representative of this group, will be
pressented later. On the other hand those scholars
who hold both the Buddha and the `Upsnisads` in high
esteem do not subscribe to this view. Let us take
note of the opinion of Radhakrishnan (14) who
belongs to this second group of scholars. According
to him the Buddha clearly tells us what the soul
(`atta`) is not, though he does not give us any
clear account of what it is. It is, however, wrong
to think that there is no soul according to the
Buddha. In support of his opinion Radhakrishnan (15)
refers to the dialogue between the wondering ascetic
Vacchagotta which, according to him, shows that
something there is, though it is not the empirical
self. In another place (16) he refers to this
empirical self and states that it is the false view
that clamours for the perpetual continuation of this
small self, which the Buddha refutes. This, (17)
according to Radhakrishnan, also agrees with the
statement of the Buddha that the self is neither the
same nor different from the skandhas. He further
states (18)： "It is also clear that the reduction
of the self to a number of skandhas is not ultimate.
If the self is merely an impermanent compounded of
body and mind......then when it disappears then
there is nothing which is delivered......Freedom
becomes extinction. But `Nirvana` is timeless
existence and so the Buddha admit the reality of a
timeless self."
Radhakrishanan appers to be of the opinion
that the term '`atta`' which he translates as self
has been used in two different senses. One meaning
of `atta` is the small self or empirical self which
the Buddha rejects through the doctrine of `anatta`.
The other implication of the term `atta` is the ture
self which is same as `nirvana` and is accepted by
the Buddha. It is the identity of this true self
with other things that the buddha rejects.
We may now discuss the opinion of Rahulo who
represents the other group
────────────
(14) Indian Philosophy, Vol. I,P.386(George Allen &
Unwin, London, 1977)
(15) Ibid. p.386
(16) Ibid. p.388, note 1
(17) Ibid. p.387
(18) Ibid. p.451
398 頁
of scholars. While explaining the doctrine of
`anatta` Rahulo (19) refers to the different
concepts of `atman`, even the concept of
Brahman-`Atman` of the `Upanisads and holds that the
buddha was unique in rejecting all such doctrines
through the doctrine of `anatta`. "The idea of an
abiding, immortal substance in man or outside,
whether it is called `Atman`, I, soul, self, or ego,
is considered only a false belief, a mental
projection. This is the Buddhist doctrine of
`antta`(20). So it appears that by the doctrine of
`anatta` the Buddha was denying `atta` "which was
accepted by every other philosophico-religious
system". Rahulo seems to suggest in not so ambiguous
term that even the Brahman-`Atman` concept of the
`Upanisads` stands hereby negated.
It is clear that the concept of the
`Upanisadic Atman` acts as a powerful background in
in fluencing the formulations of both the
above-mentioned views. Conze, (21) on the other hand,
propounds a different theory which is not connected
with the `Upanisadic Atamn` in any way. He is not
sure what notions of an `ataman` were denied by the
Buddha, but he believes that these notions are of two
kinds, e.g. i) the ideas implied in the use of 'I'
and 'mine', and ii) the philosophical opinion of the
`Samkhya` and `Vaisesika`. The basic formula absence
of a self is confined to the five skandhas, and
nothing is said either way about its existence quite
apart from them. The Buddha never taught that the
self is not, only that it cannot be apprehended.
It is obvious that the solution of this knotty
problem hinges on the proper understanding of the
nature of `atta` that was rejected by the Buddha. It
is really strange that none of the scholars who
suggested the equation of `atta` either with the
`Upanisadic` Brahman-`Atman` or with some other
non-buddhist concepts ever tried to clearly enunciate
what this `atta` is. Any such suggestion without
first defining the concept of `atta` is bound to be
methodologically unscientific and unconvincing.
To arrive at a more objective conclusion the
following approach may be suggested. On the one hand
we should study those materials from the `Nikaya` and
────────────
(19) Rahulo, What the Buddha taught, p.51
(20) Ibid. p.55
(21) The Buddhist Thought in India, pp.38-39 (George
Allen & Unwin, London, 1962)
399 頁
`Agama` which are more or less descriptive in
nature. Such materials provide us with legends and
other informantions about the `atta` philosophy
dominant in the region where the Buddha was
preaching his `anatta` doctrine, and so engaged his
attention. The reason for doing so is the
justifiable surmise that the `atta` concept or
concepts, the negation of which forms the core of
buddhism, should find some mention in this
descriptive type of scripture. On the other hand we
should carefully scrutinise another type of material
contained in the Buddhist scripture which is more
critical and philosophical in nature, and acquaints
us with detailed arguments justifying the rejection
of the `atta` philosophy. Out of such study will
also emerge the `atta` concept or concepts which the
Buddha was rejecting. The study of these two types
of materials will enable us to form a complete idea
of the different `atta` concepts the Buddha
discarded as false views. It is only then we would
be in a position to reliably know whether the
`Upanisadic` concept of `Atman-Brahman` was rejected
or not.
The major non-buddhist concepts of `atta`
criticised in the `Nikayas and `Agamas` many be
broadly divided nito four main categories dealing
with a) `Satkayadrsti`, b), `Sasvatavada`, C)
`Ekaccassatavada`, and d) `Ucchedavada`. We will
discuss these heresies in the order given here.
Ⅲ
a)`Satkayadrsti`： In the `Samyutta Nikaya`(22)
we come across the following description of a false
view which later came to be known as `Satkayadrsti`：
"Those recluses and Brahmanas who while seeing in
various ways see the `atta`(soul), they actually see
the five aggregates of attachment or any of them.
Which five？ In this connection, monks, an uneducated
ordinary person......envisages matter as a soul, or a
soul possessing matter, or matter in a soul, or a
soul in matter. Or he envisages, feeling, perception,
the gormative forces, or consciousness as a soul,
possessed by a soul, in a soul, or soul in them.
Envisaging thus he gets a thought "I am"......Being
ignorant he thinks this, or "I am this" or "I shall
be" "I shall not be" or I shall be material "or" I
shall be having perception "or" I shall be
────────────
(22) `Samyutta`, Vol.Ⅲ. 46f.； Taisho, Vol.Ⅱ,
p.11b1ff. A.K. Warder, Indian Buddhism, pp.123ff
400 頁
without perception (`a-samjnin`)" or "I shall have
neither perception nor non-perception. But one who
is educated, abandons ignorance and gets knowledge
......and does not have any of these thoughts."
The passage quoted above shows that the `atta`
in which these recluses and `Brahmanas` believed
could be in four different ways connected with the
five skandhas and this connection is constant. The
existence of an `atta` without any reference to the
skandhas was unthinkable. This association of the
soul with the skandhas gave rise to certain
ego-centric thoughts in which both these elements-the
soul and the skandhas-are present, e.g. "I am this"
"I shall be material" etc.
Here we have a succinct description of a false
view that draws our attention to its two main
features： 1) `atta` refers to an individual soul
whose essence consists of an awareness of 'I' (22a)；
2) `atta` is invariably connected with the skandhas.
Its existence is always conceived in association with
skandhas. This heresy, as already noted, came to be
known as `satkayadrsti`(Pali： `Sakkayaditthi`). The
term `satkaya` refers to the five aggregates of
attachment (`pancupadanakkhandha`)(23) and
`Satkayadrsti` to a false view that fails to see the
skandhas merely as skandhas, but always views them
through the prism of an `atta`. The most subtle form
of `satkayadrsti` is discernible in the feeling "I
am" which lingers in anybody who is spiritually below
the status of an arhat. He may no longer consciously
believe in
────────────
(22a) In the Alagaddupamasuttam of the Majjhima
`Nikaya` (PTS), Vol. I, pp.138ff. (Nalanda Ed.),
Vol.I,pp.185ff. We read that the feelings of 'I'
and 'mine' are interdependent. And with "I" and
"mine" to be really and truly existing, the
possibility of this falso view would also be
there： "That is the world, that is `atta`. I
shall be after death, everlasting, abiding
eternal, unchanging and I shall remain as such
for eternity". The feeling of 'I' and 'mine'
being absent, the possibility of the false view
would not be there. So a staunch belief in the
reality of 'I' and 'mine' is the fulcrum on
which the illusory image of an eternal soul
rests.
(23) `Satkaya` or `Sakkaya` means
`pancupadanakkhandha` or five aggregates of
attachment： "`Panca` kho ime...`upadanakkhandha
sakkayo` vutto `bhagavata` seyyathidam
`rupupadanakkhandho`, `vedanupadanakkhandho`,
`sannupadanakkhandho`, `vinnanupadanakkhandho`,
`samkharupadanakkhandho` (Collavedallasutta,
Majjhima `Nikaya`, Vol.I,pp.369-70(Nalanda Ed.),
p.299(PTS)
401 頁
it, he may even intellectually refute it, but he is
not yet free from the vestige of this pernicious
heresy.(24)
We may refer to passage in the `Kosa`
literature that helps us to understand more clearly
the different modes of relationship existing between
the soul and the five skandhas (Conze, p.33). This
passage, like the `Samyutta Nikaya` account,
distinguishes between the twenty bases of grasping at
the notion of soul. One regards (1-5) the soul as the
five skandhas, as the flame of a lamp is identical
with its visual appearance； (6-10) the soul as
having or possessing the five skandhas, like the
shadow of a thing；(11-15) the skandhas in the soul,
as the scent in a flower；(16-20) the soul in the
skandhas, as the gem in the casket.(25)
────────────
(24) In this connection we may refer to the episode of
khemaka recorded in the `Samyutta Nikaya`, Vol.Ⅲ,
pp. 127ff.(PTS Ed.). Some monks ask Khemaka
whether he sees in the five aggreates of
attachment any self (`atta`) or anything
pertaining to self (`attaniya`). Khemaka replies
in the negative but at the same time admits that
he is not yet an arhat. "I am not an arhat free
from all impurities, Oh friends, with regard to
the five aggregates of attachment. I have a
feeling 'I am' but I do not clearly see 'This is
I am'". Then Khemaka explains what he calls 'I
am' is neither matter, feeling, perception,
mental formation (`sankhara`) nor consciousness
nor anything without them. But he has the feeling
'I am' with regard to the five skadhas, though he
could not see clearly 'This is I am'. He says it
is like the smell of a flower, it is neither the
smell of the petals, nor of the colours, nor of
the pollens, but the smell of the flower.
Khemaka further explains that this feeling 'I
am' disappears when one progresses further, just
as the chemical smell of a freshly washed cloth
disappears when it is kept in a box.
(25) See, E. Conze, Buddhist Thought in India, p.33.
It is to be noted that the same comparison of
`atta` with the smell of a flower also occurs in
the `Samyutta Nikaya` (see, note no. 24). We
should, however, observe that though Khemaka
feels 'I am' he does not treasure this feeling；
he wants to get rid of it as something unreal.
But to a follower of `Sasvatavada` the feeling
indicates something real which is the very
essence of his eternal being. The spiritual
experience is same in both the cases, though the
wisdom needed to interpret the experience
correctly is absent in case of the
`Sasvatavadins`.
402 頁
This account of the `Satkayadrsti` acquaints
us with the most basic form of `atta` heresy.
`Satkayadrsti` merely postulates a relationship
between the soul and the five skandhas. It neither
advocates the eternity of the soul nor holds the
soul to be co-terminous with the body.
`Satkayadrsti` is neither `Sasvatavada` nor
`Ucchedavada`. But the following statement recorded
in the `Samyutta Nikaya`(26) `ima` kho gahapati
ditthiyo `sakkayaditthiya` sati honti-shows that the
`satkaya drsti` is the root cause of all other
heresies, including the `Sasvatavada` and the
`Ucchedavada`. A person believing in a soul and
having desire for existence (`bhava-trsna`)
willingly believes that the soul survies the body
and accepts the false view of its permanetnt
existence (`bhava-drsti`). On the other hand, a
person thirsting for non-existence
(`vibhava-trsna`), willingly believes that death is
the ultimate end of every thing, and comes to
believe in the false view of non-existence
(`Vibhavadrsti`) the annihilation of soul after
death (`Ucchedavada`).(27)
b)`Sasvatavada`：The account of the
`Sasvatvada` as given in the `Brahma-jalasutta`(28)
may be abridged as follows：-
There are recluses and Brahmanas who accept
the philosophy of eternalism (`sassatavada`) and
proclaim that both the soul and the world are
eternal. And why they do so？ Some recluse and
Brahmana by means of exertion and proper mental
attention attains to such a cencentration of mind
that they are able to recall to mind many hundreds
of thousands of past births. And they think："There
────────────
(26) `Samyutta`, Vol.Ⅳ,p.287
(27) Lamotte, Indian Buddhism, pp.35,48,51. In this
connection we may refer to the
`Sattasavasutta`, Majjhima `Nikaya`, (Nalanda
Ed. Vol.Ⅰ,pp.12-13), which also states that it
is due to ego-centric thoughts (e.g. atthi me
atta etc.) that such a wrong view arises："This
is my self, which speaks and feels, which
experiences the fruits of good and bad actions
now here, now there, this self is permanent,
stable , everlasting, unchanging, remaining the
same for ever and ever." This false view is
given the designation of `Sasvatavada` in the
`Samyutta Nikaya`； "That is the self, that is
the world (loka). He takes rebirth, he is
eternal, stable, unchanging- this is the
eternalist veiw-point." cf, S. PartⅢ,pp. 98-99；
182
(28) `Digha Nikaya`, Vol. Ⅰ(PTS Ed. by T.W.Rhys
Davids)London, 1973, pp.13-14； T.W.Rhys Davids,
Dialogues of the Buddha, Part Ⅰ, pp.27-28(PTS,
London, 1977)
403 頁
I was born of such a name (`evam-namo`), of such a
lineage and caste (`evam-gotto`,`evam-vanno`).(29)
My food was such (`evam-aharo`). I was the
experience of such pleasures and pains
(`evam-sukha-dukkha-patisamvedi`) and had such a
span of life (`evam-ayu-pariyanto`). Falling from
there I was born here."(30) Thus they remember their
past existences in full detail. And each of them
says to himself： "sassato `atta` ca loko ca `vanjho
kutattho esikatthayithito`, te ca `satta sandhavanti
samsaranti` cavanti upapajjanti, atthi tveva
sassata-saman ti."(31) T.W.Rhys Davids translates as
follows： "The soul is eternal； and the world
giving birth to nothing new, is steadfast as a
moutain-peak, as a pillar firmly fixed； and that
though these living creatures transmigrate and pass
away, fall from one state of existence and spring up
in another, yet they are (atthi) for ever and ever."
(32)
Though this translation tallies with the
traditional Buddhist explanation, it is difficult to
accept it, as it raises some problems. From the
legend we have seen that the living beings in this
world could only remembers the details of their past
lives here. And an awareness of an unchangeable 'I'
(cf. I was born with such a name etc.) links all
these past lives together and gives rise to the
notion of an eternal soul (`atta`). The same soul
was thought to be born again and again as different
individuals. The `atta` and `satta` are not
identical. The skandhas are admitted to be different
in each of these births, though the `atta`
characterised by the awareness of an 'I' remains
unchanged.
The above translation from Rhys Davids of the
stock description of `Sasvatavada` is not in
aggreement with the implications of the
`Sasvatavada` legend. In this translation not only
the two verbs of motion (`sandhavanti` and
`samsaranti`) indicating repeated deaths and
rebirths and so underlining changes are connected
with the `satta`but also the verb of existence
(atthi) indicating permanency becomes associated
with the same `satta`. If we accept this translation
we have to admit that the `Sasvatavadins` thought
the concept of eternity to be compatible with the
────────────
(29) in the `Mahapadanasutta` we have `evam-jati` in
place of evam-vanno. Both the terms mean social
class or caste. (`Digha Nikaya`, Vol.Ⅱ.p8)
(30) `Digha Nikaya`, Vol. Ⅰ, p.13(PTS,London, 1975)
(31) Ibid, p.15
(32) Dialogue of the buddha, Part Ⅰ. p.28 (PTS,
London, 1977)
404 頁
notion of change. The soul's eternity was not at
variance with its identity with the changeable
skandhas. But there is no reason to believe that the
`Sasvatavadins` were so unrealistic as to support
such an unreasonable view. We have seen that
according to the `Sasvatavada` the eternal soul
characterised by a sense of 'I' was different from
the changing individuals. Moreover, other religious
groups contemporary of the `Sasvatavadins` were also
not supporters of such a view. The `Ucchedavadins`
(33) took the soul to be identical with the skandhas,
and consistent with this view they held that the soul
is destroyed together with the body. The
`Ekaccasassatavadins`(34) believed the `Mahabrahma`
to be eternal as they did not appear to have known
that the `Mahabrahma` was also identical with the
skandhas, created, and subject to death. All the
religious groups including the `Sasvatavadins` thus
seem to have held the view that the soul to be
eternal must have an existence independent of the
skandhas. So the translation making the `satta`
eternal cannot be accepted. The only other
possibility is to take 'atthi' in the sense of a
singular number and to connect it with 'atta`,
Then we get the following translation： "The soul
is eternal； and the world giving birth to nothing
new, is steadfast as a mountain-peak, as pillar
firmly fixed； and these living beings move on and
on, transmigrate, fall from one state of existence,
rise up in another, but (the soul) exists for
eternity."
For our purpose it is, however, important to
know how the buddhists would understand this
philosophy. From our discussion of the
`Satkayadrsti` it is clear that according to the
Buddhist's interpretation of the false view the soul
must be either identical with or closely related to
the Skandhas. The relation is such that the existence
of a soul apart from the skandhas is unthinkable. The
changes which the skandhas undergo even during one's
lifetime are obvious. Moreover the acceptance of the
repeated deaths and rebirths shows that the skandhas
do not remain unchanged. In short, according to the
buddhist interpretation, the `Sasvatavadins`
admitted the changeable nature of the skandhas,
accepted the close relationship existing bewteen the
soul and the skandhas, and yet held the soul to be
eternal (sassato). This also seems to be the opinion
of Buddhaghosa (35) as
────────────
(33) `Digha Nikaya`, Vol I, see, p.34； also see, p.12
(34) See, p.10 ff.
405 頁
expressed in the `Sumangalavilasini`. But, as already
shown, Buddhist presentation of the eternalism is
not the same as the original philosiophy of the
`Sasvatavadins`. The `Sasvatavadins` themselves
regarded 'atta' to be independent of body and mind,
i.e. of the skandhas, to use the Buddhist terminology.
This seems to be reason why the `Sasvatavadins`
themselves did not relate the soul to any of the
skandhas in the stock description of their own
philosophy. Even the logicians and thinkers (`takki`,
`vimansi`)(36) among them have nothing to say about
the relationship existing between the soul and the
skandhas.
Whatever may be the case, it is of no consequence
for our present study that the Buddhist presentation
of the `atta` of the `Sasvatavadins` does not tally with
the original concept. What really matters is the
idea the Buddhists had of such a concept. For when
they reject a concept of an eternal soul, they do so
in the light of their own understanding of it. So it
is important to take note of their idea of the
`atta` concept which may be described as follows：
The `atta` is an eternally existent individual soul
which is either identical with the skandhas or so
closely related to them as to be dependent on them
for its very existence. It neither attains
liberation nor gets annihilated.
c)`Ekaccasassatavada`：A concept of `atta`
different from that of the eternalists has been
recorded in the `Brahmajalasutta` of the
`Digha-Nikaya`.(37) This is the concept of an
eternal soul unique to the `Mahabrahma` as held by
the followers of the `Ekaccasassatavada`. The Buddha
in order to explain the origin of the
`Ekaccasassatavada` relates the following legend
about the past lives of the upholders of this
philosophy：
When the world system dissolves the beings are
reborn in the `Abhassara` world. They are made of
mind. Now the world system begins to come into
exitence and an empty `Braham-vimana`(Brahma-palace)
appears. One of the beings at the end of merits or
span of life falls from the `Abhassara` world and is
reborn in the Brahma-palace. There he lives, made of
mind (manomayo), feeding on joy (`piti-bhakkho`),
radiating light. etc. and thus does remain there for
a
────────────
(35) `Sumangalavilasini`, Vol. I, p.105(PTS,London,
1968)
(36) `Digha Nikaya`, Vol. I, p.16
(37) Ibid. pp.17-19；Dialogues of the Buddha, Part I,
pp.30-32
406 頁
long time. And he feels lonely and longs for
companions. Just then some beings die in the
`Abhassara` world and are born in the Brahma-palace.
They are just like the being who was first born. As
these beings were born after he desired for company,
he thought himself to be `Brahma Mahabrahma`,
Creator, Father of all that are and are to be. And
the other beings also thought that they have been
created by `Mahabrahma`. Their lives are of shorter
duration than that of `Mahabrahma`. They die in the
Brahma-palace and are reborn here in this world.
Here one of them leads a religious life and can
remember his past existence in the Brahma-palace,
but not beyond that. He thinks of `Mahabrahma` as
creator permanent, fixed, eternal, of a nature which
is not subject to transformation (nicco, dhuvo,
sassato, `aviparinamadhammo`) while he regards
himself and other beings as created by `Mahabrahma`,
impermanent, not fixed, eternal, and having the
nature of dying (`ancica`, `addhuva`, `appayuka`,
`cavanadhamma`).
The last part of the above account seems to be
a stock description which formed a part of the
original philosophy of the `Ekaccasassatavadins`.
According to their philosophy only the uncreated is
eternal while the created is impermanent. Only
`Mahabrahma` is eternal and the other beings are
impermanent. The information that `Mahabrahma` was
made of mind were most probably not included in
their philosophy, for it is not mentioned in the
utterance of him who remembers his past existence.
Otherwise we have to assume that according to the
`Ekaccasassatavadins` mind can be both created and
uncreated. So it appears that this extra bit of
information has been added in accordance with the
Buddhis dogma. Thus as per the understandings of
the Buddhists, the followers of the
`Ekaccasassatvada` believed in an eternal soul only
in case of `mahabrahma` and this soul was made of
mind, uncreated and immortal.
Two other groups of the `Ekaccasassatavadins`
(38) believed in many eternal individual souls. But
the eternity is not absolute in these cases； it may
be lost either due to the lack of self-controll or
due to envy. The fourth group of the
`Ekaccasassatavadins`(39) were comprised of the
logicians and the thinkers who concluded by
reasoning that there are two souls, one impermanent
and the other
────────────
(38) `Digha Nikaya`, Vol. I, pp.19-21
(39) Ibid, p.21
407 頁
permanent. The soul consisting of ear, nose tongue
and body is impermanent, not fixed, not eternal and
having the changeable nature. But the soul which is
thought or mind or consciousness (cittan ti va mano
ti va `vinnaran` ti va ) is permanent, eternal etc.
d)`Ucchedavada`： The seven groups of
`Ucchedavadins`(40) identified an individual soul
variously with physical body, or sensual desire, or
mind, or infinite space, or infinite consciousness,
or nothingness, or neither perception nor not
perception. The first group believed in one soul
made of gross matter while the other groups believed
in more than one souls. And all these groups upheld
the doctrine that the soul is annihilated with the
destruction of the body. It appears that even before
the Buddhists the `Ucchedavadins` denied the belief
in an eternal individual soul (`atta`) which was
identical with one of the skandhas. Therefore this
philosophy came to be known as the doctrine of
`niratta`.
So far we have discussed the `Satkayadrsti`,
`Sasvatavada`, `Ekaccasasvatavada` and the
`Ucchedavada`, and these are the only dominant
heresies regarding `atta` and relevant to our study
that have been recorded in the `Agamas` and
`Nikayas`.(41) Presumably these were the four main
type of heresies that attracted the attention of the
Buddha and the early Buddhists. It is obvious that
the `atta` concepts discussed in the `Ucchedavada`
and `Satkayadrsti` can in no way be connected with
the problem of `atta's` identity with the
`Upanisadic Atman`. It is only the `Sasvatavada` and
`Ekaccasassatavada` concept that deserves to be
considered in this connection.
In course of our discussion we have noted the
following characteristics of an eternal soul as
envisaged by the followers of the `Sasvatavada` and
`Ekaccasassatavada`. The first group of
`Ekaccasassatavada` believed that only the
────────────
(40) Ibid, pp.34-35
(41) I have not mentioned the other `atta` heresies
described in the `Brahmajalasutta`, as they are
not radically different from what we have
discussed in the present article. All these
heresies are also based on the `Satkayadrsti`
and so share the characteristics of the
`Satkayadrsti`. These heresies must also be
partially similar to the `Sasvatavadins or the
`Ekaccasassatavadins`. I could not trace any
other sutta devoted to the rejection of any of
these heresies. The Buddhists themselves, it
seems , did not think it necessary to discuss
these false views separately and elaborately.
408 頁
soul of `Mahabrahma`, who is uncreated, is eternal.
But the `Sasvatavada` and the other groups of
`Ekaccasassatavada` recognise numerous, eternal,
individual souls who are either identical with the
skandhas or so closely related to them as to be
dependent on them for their very existence. The
second and third groups of the `Ekaccasassatavadins`,
however, are of the opinion that the eternity is not
absolute, for the beings die if they indulge in
sensual pleasures or are afflicted with jealousy. The
first type of `Ekaccasassatavada` explicitly states
that the souls, including that of the `Mahabrahma`,
are made of mind. This view seems to have been
accepted by the second and third groups of the
`Ekaccasassatavadins` who made the eternity of the
individual souls dependent on the purity of mind.
According to the fourth group of the
`Ekaccasassatavadins` which consists of the logicians
and the thinkers, the individual soul is identical
with citta or manas or `vijnana`. Thus we find that
the `Ekaccasassatavada` concept of soul belonged to
the sphere of mind and at least in most cases could
not rise beyond the `Rupaloka`. Moreover the eternal
souls are confined within space and time and are
characterised by the subject-object split.
The `Upanisadic Atman`, on the other hand, is
not an individual soul. The `Atman` is the supreme
reality, the only Being that is beyond speech, beyond
the reach of mind and the notion of space. And such
an ` Atman can by no strech of imagination be equated
with any of the skandhas. In short, the sassato
`atta` of the Buddhist scriptures and the `Atman` of
the Upanisads are two diametrically opposing points
of view.
So long we have discussed scriptural materials
which are more or less descriptive in nature. Now we
pay attention to a more critical type of `Agama
Nikaya` passages which acquaint us with the reasons
for the rejection of different types of `atta`
heresies, and thereby provide us with some extra
details regarding these heresies.
Ⅳ
We have seen that four kinds of relationship
between the `atta` and the skandhas were recognised.
Out of these only one type of relationship viz. that
of identity between the `atta` and the skandhas,
finds prominent mention in the `Nikayas` and the
`Agamas`. This appears to have been the dominant
heresy at the
409 頁
time of the Buddha. Only in a few cases the early
buddhist canon takes note of other types of
relationship, viz. `atta` has skandhaka or `atta` is
not a particular type of skandhaka. Again the nature
and contents of such a criticism of `atta` heresy
vary depending on the type of persons for whom it
was meant. The discourses which are held for the
benefit of the Buddhist monks are quite different
from those meant for non-Buddhist asceties.
with these preliminary remarks we will now
proceed to arrange the relevant materials according
to the type of heresies criticised and the type of
persons addressed.
A：Criticism of heresies meant for the Buddhist
monks
Ⅰ)Rejection of `Satkayadrsti`：The `samyutta`
passage referred to above is a plain form of
instruction discarding the heresy of `Satkayadrsti`.
The Buddha points out that some recluses and
Brahmans by erroneous observations of the skandhas
give rise to the false notion of a soul, but those
who are wise never make this mistake. Here the talk
is about the mere existence of a soul with reference
to the skandhas, and not the eternity or
impermanence of soul. It is apparent that the Buddha
is rejecting the `Satkayadrsti`. The Buddha does not
find it necessary to offer any extra argument in
favour of his view, as he addresses his own
disciples who have absolute faith in his wisdom.
Ⅱ)Criticism of `Sasvatavada`： In the account
of the `Brahmajalasutta` which we have already
discussed, the Buddha gives critical description of
the heresy of `Sasvatavada` in the form of a legend.
He points out that some recluses and Brahmanas
believe in eternal individual soul for they can
remember their past lives in this world. The sutta
implies that the memory of the past lives is genuine
though the interpretation of such experiences is
wrong. In this passage no logical argument is
advanced justifying the rejection of this heresy. It
will, however, be clear later from our discussion of
another passage from the same sutta that the Buddha
rejects this view on the basis of his more extensive
knowledge and higher knowledge.
Ⅲ)Criticism of `Ekaccasassatavada`：Like the
legend about the `Sasvatavada`, the legend
concerning the `Ekaccasassatavada` in the
`Brahmajalasutta` implicitly contains the criticism
of the concept of an eternal soul. The legend states
that
410 頁
the beings who died in the `Brahma` world were born
in this world. While in this world they could
remember their past birth in the `Brahma-vimana`,
but had no knowledge of their still earlier birth in
the `Abhassara` world, nor could they know that
`Mahabrahma` is also subject to death. This legend
suggests that the incomplete experience coupled with
imperfect knowledge led these beings to interpret
wrongly their genuine spiritual experiences.
Consequently they came to believe in the false
notion of an eternal `Mahabrahma` who is made of
mind. So the Buddha is here criticising
`Ekaccasassatavada` on the basis of his more
extensive knowledge.
Ⅳ)Criticism of heresies in general： There is
no legend criticising the `Ucchedavada`. This is
because the `Ucchedavada` tradition most probably
did not record any legend justifying this
philosophy. The `Ucchedavada` philosophy only
recognised the present life and denied the existence
of any life before birth and after death. So there
was no scope for the development of any such legend.
In the `Brahmajalasutta`, however, we come
across a general criticism of all the false views
noted in this sutta.(42) Here the criticism is based
on the more extensive knowledge and higher knowledge
of the Buddha. The passage in question states that
the Buddha knows of all these false views and also
knows of into what sort of future existences fall
those who grasp at and become attached to such
views. Knowing this and knowing (other) higher
things (`tan` ca ` Tathagato pajanati`, tato ca
`uttaritaram pajanati`) the `Tathagata` does not get
involved, he knows about extinction； and having
truly known the rising and passing away of
sensations (`vedananam samudayan` ca `atthamgaman`
ca), their taste, danger, their not being the refuge,
`Tathagata`, due to the non-attachment, is free
(vimutto).
These are those other things, profound,
difficult, to understand......not to be grasped at by
mere logic, which `Tathagata` having himself realised
and seen face to face, has set forth."
The higher knowledge which is profound and
beyond logic is different from the knowledge of
future existences and past lives of the beings. The
phrase- "Knowing this and knowing (other) higher
things" -clearly shows that two different types of
knowledge are referred to. The higher knowledge
refers to the
────────────
(42) `Digha Nikaya`, Vol.I,pp.16-17,21-22,24,28,29
etc.
411 頁
rising and passing away of `vadana` (sensation or
feeling), i.e. the knowledge of the
`paticcasamuppada` which only the Buddha possessed.
That the statement reggarding `'vedana'` refers to
the truth of the dependent origination is confirmed
by another passage of the same sutta (D. 1.3.71). All
the false views are rejected because they are
contradicated by the law of dependent origination. Of
all the links of the `Paticcasamuppada`, `Vedana` is
specially mentioned, for the awareness of 'I' is
directily dependent on `vedana`.
In corroboration of the the conclusion reached
above we may take note of another sutta where the
rejection of a false view on the basis of
`paticcasamuppada` is explicitly mentioned.
Ⅴ)Rejection of the `Ekaccasassatavada`： The
Buddha in his instruction to his disciple
`Kevattaputto Sati` is rejecting the
`Ekaccasassatavada` on the basis of
`paticcasamuppada`. (43) `Sati` gave out that it is
the self-same `vijnana` that passes from one
existence to another. Here `Sati` is speaking in
favour of `Sasvatavada`. This statement is
elaborated in the `Papancasudani`(44) to bring out
its full significance："That which speaks,
experiences now here now there, the fruits of his
good and evil actions is that consciousness
(`vijnana`) that I am speaking about.' The same
`vijnana` is enjoying the fruits of his actions.
This is the same as preaching `Sasvatavada` as it
will be clear from a passage of the `Sabbasavasutta`
of the Majjhima `Nikaya` and a passage from the
`Samyutta Nikaya`. The Majjhima passage runs thus
(45)：Atha va pana assa evam `ditthi` hoti：yo me
ayam `atta` vado vedeyya tatra tatra
kalyanapapakanam `vipakam patisamvedeti` so kho pana
me ayam `atta` nicco dhuvo sassato aviparinamodhammo
sassatisamam tath'eva thassati' 'Or a wrong view
occurs to him thus：the soul of mine that speaks,
experiences now here and now there the fruits of his
good and evil actions is indeed that soul of mine
that is permanent, stable, not subject to change,
that will stand firm for eternity.' The `Samyutta`
passage (46) states： One and the same person both
acts and
────────────
(43) M.I, `Mahatanhasankhayasutta`, pp.256ff. idam
`vinnanam sandhavanti`, `samsaranti anannam`.
ibid.p.256.
(44) `Papancasudam`,Vol.Ⅱ,p.305(PTS,London, 1979)
(45) Majjhima, Vol.Ⅰ,p.8(PTS,London, 1976)
(46) `Samyutta Nikaya`, Vol.Ⅱ,p.20
412 頁
experiences (the results)- this Kassapa which you
called at first 'suffering self-wrought' ammounts to
the Eternalist theory. (47)
It is also to be noted that `Sati` is
identifying the eternal soul with the `vijnana`
only. This reminds us of the view of the logicians
among the `Ekaccasassatavadins` who held that
indriya etc. are impermanent while the citta
(thought), or manas (mind), or `vijnana`
(consciousness) is eternal. So the eternalism which
`Sati` is preaching actually ammounts to the
`Ekaccasassatavada` of the logicians. To reject this
view the Buddha refer to the law of dependent
origination (`paticcasamuppada`). The Buddha
condemned the view of `Sati` and corrected him by
pointing out that `vijnana` can only originate
through cause and conditions (`annatra paccaya`
na'tthi `vinnanassa` sambhavo ti)(48) and so cannot
be eternal.
In the preceding pages we have dealt with such
criticisms of the false views regarding `atta` as
were meant for the loyal disciples of the Buddha.
For his disciples the unquestionable veracity of
these criticisms ultimately rests on their
unflagging faith in the claim of the Buddha to the
more extensive and higher types of knowledge. It was
not necessary for the Buddha to analyse critically
the different heresies in order to expose the flaws
in them, or to adduce extra reasons to justify their
rejection. The approach, however, changes when the
discourses are meant for the non-Buddhist ascetics.
B：Criticism of heresies meant for non-Buddhist
ascetics
Ⅰ)Rejection of `Sasvatavada`：(49) The
`Anatta-lakkhana` -sutta of the `Samyutta Nikaya` (
ⅢⅩⅩⅡ, 59)(50) while giving a clear exposition of
the `anatta` doctrine,
────────────
(47) Kindred Sayings, Vol.Ⅱ,p.16(PTS,London, 1982)
(48) Majjhima, Vol.Ⅰ,p.259. It appears that `Sati`
misunderstood the Buddhist doctrine of rebirth.
It is said in the `Samyutta Nikaya` (Ⅰ,122；
Ⅱ,67,103) that is only when `vinnana` obtains
a footing on something (`arammana`) that there
is the possiblity of the birth of a being, and
not otherwise, and the moment the `vinnana`
ceases (cuti), one is regarded as dead. Also
see, Nalinaksha dutt, Early Monastic Buddhism,
Vol.Ⅰ,P.255 (Calcutta, 1941).
(49) `Samyutta`, Vol.Ⅱ,pp.66-68(PTS.London, 1960)；
Vinaya, Vol.Ⅰ,P.13-14(PTS, London, 1964)
413 頁
adduces reasons for the denial of `atta` in the
following manner： Form (`rupa`) is not soul
(`atta`). If it were , this form could not turn
oppressive, and with regard to form it would be
possible to achieve the intention that "let my body
be thus, let my body be not thus". And so with
`vedana`, `sanna`, `samkhara`, and `vinnana`. "What
do you think, is form permanent or impermenet？" "It
is impermanet, Oh Lord". "But is the impermanent ill
(dukkha) or ease (sukha)？" "It is ill, Oh Lord".
"But is it fitting to consider that which is
impermanent linked to suffering, doomed to reversal
as 'this is mine, I am this, this is my soul'". "No,
indeed, Oh lord". And so for `vedana`, `sanna` ect.
Therefore whatever form there is, past or future,
inner or outer, gross or subtle, low or exalted,
near or far away, would be seen by right wisdom as
it really is, i.e. "all this form is not really mine,
I am not really this, this is not my soul (`atta`)."
And so `vedana`, `sanna` etc. Seeing this the
well-disciplined holy disciple become disgusted with
the skandhas."
From a careful study of the passage quoted
above the concept of the `atta` rejected here clearly
emerges. The `rupa` and other skandhas cannot be
`atta` for they turn oppressive and cannnot be
changed according to one's liking. So it follows that
te `atta` or the individual soul enjoys complete
self-mastery and remains ever happy. Moreover the
five skandhas are not `atta` for they are impermanent
and subject to change and suffering. This shows that
the `atta` concept rejected here was believed to be
identical with the skandhas, permanent, changeless,
happy and characterised by an awareness of an 'I' and
'mine'. So the concept of `atta` which emerges wholly
conforms to the `Sasvatavada`.
Two opposing trends of thought are discernible
in this sutta. One represent the viewpoint of the
`Sasvatavada` as presented above. The other trend
shows the Buddha's acceptance of the notion of I and
mine as a basis for further argumentation in order to
disprove the soul's identity with the skandhas. The
line of reasoning, partly explicit and partly
implicit, may be presented thus： the idea of 'I' and
'mine' is the characteristic feature of atta and
implies complete self-mastery of oneself. What one
refers to as 'I' and 'mine' must be its own master.
Moreover the individual soul is permanent. But the
body (`rupa`) is beyond one's control.
────────────
(50) `Digha Nikaya`, Vol.Ⅰ,pp.179,182 ff.；Dialogues
of the Buddha, Part Ⅰ, p.246ff. Also see,
A.K.Warder, Indian Buddhism (Motilal Banarasidas,
Delhi, 1991) pp.124ff.
414 頁
It grows, becomes old and ultimately dies without
anybody being able to do anything to arrest this
process. In short this philosophy upholds the
self-mastery and permanency of an individual soul
but advocates in the same breath the identity of the
soul with the five skandhas which are impermanent
and lack self-mastery. Thus it is clear that the
viewpoint of the `Sasvatavadins` suffers from
internal contradiction and cannot be accepted.
It should be noted that the Buddha did not
criticise this philosophy on the basis of any
Buddhist doctrine. The impermanence of body etc, is
a matter of common experience and no higher philophy
is needed to prove it. The Buddha neither referred
to the skandhas constituting a being nor to the law
of `paticcasamuppada` in order to deny the existence
of a permanent ego-centric entity. Rather he made
use of the notion of 'I' which is taken to be the
tell-tale sign of an individual soul to refute the
`Sasvatavada`. This approach appears to be all the
more intriguing as the Buddha is preaching to his
own disciples, and not to the non-Buddhist ascetics.
And from the dialogue between the Buddha and his
disciples it is obvious that his disciples has not
turned into supporters of this false view, but had
remained faithful to his teachings. Moreover it
would be preposterous to hold that the Buddha
actually supported the existence of a permanent
individual soul, although he denied its identity
with the five skandhas. What, then, is the reason
for this particular mode of preaching？ We can
reasonably surmise that he was instructing his
disciples how to meet the challenges of such an
erroneous view. The inadvisability of trying to
discard this false view by referring to the Buddhist
doctrine is obvious, for the non-Buddhists cannot be
expected to have any faith in the teachings of the
Buddha. So the Buddha was teaching his disciples how
such false doctrines could be successfully countered
even while basing one's argumants on such articles
of faith as formig an integral part of the
philosophy they were criticising. He was trying to
show how the logical implications of one aspect of
this false doctrine would render null and void
another aspect of the same doctrine. In other words
the Buddha was teaching his disciples to expose the
contradictions inherent in the doctrines of the
`Sasvatavadins` in such a way that even the
upholdres of this philosophy would be forced to
admit the justification of the criticism. In the
`Anatta-lakkhana-sutta` the point which the Buddha
drives home is that the logical implications of the
notion of 'I'
415 頁
which were acceptable to the protagonists of this
philosophy, would contradict the other part of the
philosophy, viz. the identity of the individual soul
with the skandhas. It is also clear that the sutta
was ultimately meant for the `Sasvatavadins`
themselves.
Ⅱ)a) The `Potthapadasutta` and the rejection
of `Satkayadrsti`： this sutta starts with a
discussion of the rising and the cessation of
perception (`sanna` or `abhisanna`) and then follows
it up with a judgement about the problem of identity
between the soul and perception. We will just have a
glimpse of the discussion as it will help us to
understand the nature of perception. The Buddha gave
a gradual discoures on the higher and higher stages
of perception leading to the summit of perception
(`sannaggam`) and then to the cessation of
perception (nirodha). Except the final stage, at
every other stage a perception of a lower type is
replaced by a perception of a higher type. The
Buddha showed that a certain type of perception
arose due to a certain type of thinking, a certain
type of mental training. When the thinking ceased,
the perception also ceased. With the cessation of
all thoughts, the possibility of the rising of any
new perception comes to an end.
Next `Potthapada` raises the question of a
soul. "Is perception the soul of a man, Sir, or is
perception one thing and soul another？" The Buddha
asked, "What, now, `Potthapads`, do you assume a
soul？" "I presume a gross soul Sir, material, made
of four great elements, feeding on solid food". "Yet
if your soul were gross, `Potthapada`, material,
made of four gross elements feeding on solid food,
in that case for you perception would be one thing,
soul another. Just let this gross soul be,
`Potthapada`, for then a man's perception occurs as
one thing, ceases as another thing". `Potthapada`
next proposes to assume a mental soul, with perfect
faculties, complete in its faculties. The Buddha's
objection remains the same： perception would then
occur as one thing but cease as another thing.
Lastly `Pottapada` proposes an immaterial soul,
consisting of perception. the objection of the
Buddha is still the same.
The denial of the identity between the soul
and perception is based on the following pattern of
reasoning. If the soul is material like body, then
the soul cannot be identical with perception. For in
that case the perception would be rising as one
thing and ceasing as another. It means that as long
as the perception is in existence, it can be assumed
to be soul (as body). But when it ceases to be, it
416 頁
can no longer be equated with the soul, for the
living material soul would be still in existence.
Hence the statement about the perception as rising
as one and ceasing as another. This reasoning also
holds good while showing the difference between the
perception and soul as mind, for the cessation of a
particular perception does not mean the annihilation
of mind. But this type of reasoning is not valid
when one assumes the soul to be made of perception.
In this case the cessation of perception will
automatically mean the annihilation of soul, and
consequently the argument that the perception ceases
as another is no longer tenable. This part of the
discussion appears to be a later mechanical
addition.
Whatever may be the case, it is obvious that
`Potthapada` is equating `atta` with a a skandhaka.
In his query about the identity of the soul with
perception as well as in his suggestion that the
soul may be made of four gross elements, or of mind,
or of perception, `Potthapada` is equating soul with
one of the skandhakas. He is not concerned about the
eternity of the soul or otherwise. In other words we
are dealing here with the `Satkayadrsti`, the root
cause of all other false views.
In rejecting the `Satkayadrsti` the Buddha is
not denying the existence of `atta` on the basis of
`paticcasamuppada` as he did while instructing his
own disciples. He is even accepting-for the sake of
argument, or we may say as a skillful means-
`Potthapada's` point of view regarding the existence
of soul or the constitution of soul, and then
rejecting the suggested identity by exposing the
logical inconsistencies involved in the suggestion.
He is showing that the acceptance of the view that
the soul is made of mind or four gross elements
contradicts the other aspect of the suggestion, viz.
soul is identical with perception.
b)`Mahanidanasutta` and the rejection of a
diiierent type of `Satkayadrsti`(51)
`Mahanidanasutta`(52) rejects a new type of
`Satkayadrsti` which may be formulated as soul
possessing feelings. The relevant passage is given
below："Herein, again, `Aanada`, to him who affirms
：'Nay, my soul is not feeling, nor is it non-
────────────
(51) For Sanskrit parallels to different `atta`
concepts discussed in the `Mahanidanasutta`,
see, `Nidanasamyukta`, C.B.Tripathi, Deutsche
Akademie der Wissenschaften Zu Berlin, IO,
1962.
(52) `Digha nikaya`, Vol.Ⅱ,ⅩⅤ.31,p.67(PTS London,
1982), Dialogues of the Buddha PartⅡ,p.64(Pts,
London, 1977)
417 頁
-sentient； my soul has feelings, it has the
property of sentience'─answer should be made：'My
friend, were feelings of every kind to cease
absolutely, then there being, owing to cessation
thereof, no feeling whatever, could one then say：
-"I myself am'？"
'No, lord, one could not.'
'Wherefore, `Ananda`, it follows that this
aspect：-`Nay, my soul is not feel feeling, nor is
it non-sentient； my soul has feelings, it has the
property of sentience'-does not commend itself".
Here the Buddha is teaching `Ananda` how to
refute the heresy of `Satakayadrsti`. The view that
the soul is not feeling, but possesses feelings
shows that the feeling is not intrinsic to `atta`
but external to it. It, therefore, follows that the
presence or absence of feeling should not at all
affect the basic nature of soul. But, as the Buddha
points out, our daily experience shows that in the
absence of all feelings there cannot be any
I-awareness which is the very essence of `atta`. So
the feeling is not external to soul； soul cannot be
regarded as having feelings. The accepted idea that
`atta` essentially means I-awareness goes against
the view that `atta` has feelings.
The `atta` heresy under discussion is
concerned with the relationship existing between the
soul and the skandhakas, and so falls under the
category of `satkayadrsti`.
It is by way of expediency that the Buddha
accepts I-awareness as the very core of an assumed
`atta` and thereby shows that the present heresy
suffers from internal contradiction. It is also to
be noted that the Buddha's instructions, though
addressed to `Ananda`, is ultimately meant for the
non-Buddhist ascetics.
Ⅲ) The `Mahanidanasutta` and the rejection of
the `Ekaccasassatavada`
The `Mahanidanasutta`(53) also contains
passages rejecting the heresy that the soul is
feeling. The relevant part of the sutta runs as
follows：-"Herein,`Ananda`, to him who affirms' My
soul is feeling'-answer should be thus made：-' My
friend, feelings is of three kinds. There is happy
feeling, painful feeling, and neutral feeling. Of
these three feelings, look you, which do you
consider your soul is？'
────────────
(53) Dialogues of the Buddha, PartⅡ,pp.63-64；
`Digha Nikaya`, Vol.Ⅱ,ⅩⅤ.28；29,pp.66-67
418 頁
'When you feel a happy feeling you do not feel
a painful feeling or a neutral feeling, you feel
just a happy feeling. And when you feel a painful
feeling, you do not feel a happy feeling or a
neutral feeling, but just a painful feeling. And
when you feel a neutral feeling, you do not feel a
happy feeling or a painful feeling； you feel just a
neutral feeling.'
"Moreover, `Ananda`, happy feeling is
impermanent, conditioned (`sankhata`), the result of
cause or causes, liable to perish, to pass away, to
become extinct, to cease. So too the painful
feeling. So too is neutral feeling. If when
experiencing a happy feeling one thinks -'this is my
soul' -when that same happy feeling ceases, one will
also think -'my soul has departed' -So too when the
feeling is painful or neutral. Thus he who say -'My
soul is feeling' -regards, as his soul something
which, in this present life, is impermanent, is
blended of happiness and pain, and is liable to
begin and end. Wherefore, `Ananda`, it follows that
this aspect -'My soul is feeling' -does not commend
itself."
The passage in question rejects the view that
the individual soul is feeling. The daily experience
of people shows that the feeling is impermanent, a
blending of happiness and pain, and subject to
origination and destruction. This characterisation
of feeling will be, in the main, also acceptable to
the `Ekaccasassatavadins`, for they, as we have
already noted, accepted the idea that the created is
impermanent. The feeling so characterised is
different from soul. The soul thus appears to be
permanent, beyond origination and destruction, and
experiencing unmixed happiness. It is identical with
one of the skandhas.
This concept of `atta` is, in all its
essential features, identical with that held by the
logicians among the `Ekaccasassadtavadins` who also
believed the individual soul to be eternal, and
identical with one of the skandhas belonging to the
sphere of mind.
It is to be noted that the Buddha is denying
the `Ekaccasassatavada` not on the strength of the
law of `paticcasamuppada` or any other Buddhist
dogma. It is obviously becasue his invoking of the
higher knowledge would cut no ice with the
non-Buddhists. He is, on the other hand, showing
that the philosophy under criticism suffers from
internal contradiction, and hence untenable. He
shows that the `Ekaccasassatavada` concept of
feeling is diametrically opposed to the
`Ekaccasassatavada` concept of soul, and so the
view that the soul is identical with feeling
419 頁
is to be discarded.
The `Mahanidanasutta` passage suports our
conclusion that the Buddha also taught his disciples
how to defeat the upholders of the `atta` heresy in
debate. Here the Buddha is teaching `Ananda` how to
refute the view of the `Ekaccasassatavadins`. In
other words his teaching is ultimately meant for the
followers of the `Ekaccasassatavada`.
Ⅳ) The `mahanidanasutta` rejects an atypical
`atta` heresy：In the `Mahanidanasutta`(54) we come
across a heresy which does not conform to any of the
types mentioned before. The relevant passage
recording a negative formulation the heresy is
given below：-
"Herein, `Ananda`, to him who affirms：'Nay,
my soul is not feeling, my soul is not
sentient'-answer should thus be given：'My
friend, where there is no feeling of anything,
can you there say：'I am？' 'You cannot,
Lord.' 'Wherefore, `Ananda`, it follows that
this aspect：'Nay, my soul is not feeling, my
soul is not sentient'-does not commend
itself."
The view that the individual soul is not
feeling and sentient is rejected because there
cannot be any awareness of 'I' without feeling and
sentience. Hence there cannot be any soul without
feeling and sentience. Here the criticism is based
on the assumption of 'I-awareness' as the essence of
te individual soul. The proposition admits of the
existence of an individual soul, asserts the soul's
difference from feeling and sentience, and is silent
about the question whether the individual soul is
eternal or not. This is a new type of `atta` heresy
which is neither `Sasvatavada` nor `Ucchedavada`.
Strictly speaking this negative formulation is not
even an illustration of `Satkayardrstti`, as it
denies, instead of affirming, the relationship
existing between the soul and the skandhas.
Against our contention one may argue that the
`formulation-atta` is not feeling - was merely used
to imply that the soul is identical with some other
skandha. In this case this implication should be
treated as the main proposition, which is different
from the given proposition. If this were so, the
Buddhist criticism
────────────
(54) Dialogues, Part Ⅱ,p.64； `Digha Nikaya` Vol.
Ⅱ,ⅩⅤ.30,p.67
420 頁
would have been directed against a proposition
asserting the identity of soul with skandhaka and
not against the formulation of the soul not being
feeling. But as this is not the case, we have to
take the given proposition as the main proposition.
Though the negative formulation states that
the individual soul is not the same as feeling, it
should not be interpreted to mean that the soul is
not connected with any of the five skandhas. For the
`Samyutta Nikaya` passage quoted above shows that
the negative formulation of this type is also the
result of the `Satkayarsti`.
To sum up, the `atta` concepts held by the
followers of the `Sasvatavada`, `Ekaccasassatavada`,
`Ucchedavada`, and `Satkayadrstti` mainly have been
criticised and rejected in the `Nikayas` and
`Agamas`. None of these concepts are identical with
the `Upanisadic Atman-Brahman`.
Except the `Ucchedavada`, all other concepts
regarding `atta` have been criticised in two
defferent ways. When the criticisms rejecting the
`atta` heresies were addressed to his own disciples,
the Buddha referred to his more extensive knowledge
that could see farther into the past and future
lives of the beings, and his superior knowledge of
the law dependent origination (`Paticcasamuppada`).
But when the criticisms were meant for the
non-Buddhist ascetics all references to these two
types of knowledge were avoided. The Buddha by way
of expediency makes use of a part of the opponent's
proposition for the sake of further argumentation,
and thereby lays bare the internal contradictions
involved in the proposition. This method of reductio
ad absurdum which the Buddha introduced, was later
applied by `Nagarjuna` with great success.
The `Ucchedavada` was criticised only on the
basis of the Buddha's extensive knowledge and
superior knowledge but not on logical grounds, as
the `Ucchedavada philosophy does not suffer from any
logical inconsistencies.
Ⅴ
We have shown that the early Buddhist
scripture rejected any notion of `atta` which was
either identical with the skandhas or dependent on
them for its very existence. The Buddha specially
took great pains to discard the `Sasvatavadins` and
the `Ucchedavadins` who preached the philosophy of
`atta` and `niratta` respectively. The Buddhists
were aware of this fact and gave expression to it in
421 頁
canonical and non-canonical texts. We will cite a
few quotations from the Buddhist texts in support of
our view. In the `Dutthatthakasutta`(NO.8) of the
`Suttanipata` we read the following verse；(55)
"attam nirattam na hi tassa `atthi` adhosi se
` ditthimidha sabba`"
Here the belief in `atta` and `niratta` are
counted among the false views (`ditthi`) which a
follower of the Buddha has given up. Again the
following two verses from Lalitavistara(56) and
`Madhyamikakarika`(57) while confirming the Buddhist
rejection of both `atta` and `niratta` give some
extra information regarding these heresies：
Ⅰ)`astinastivinirumktamatmyanairatmyavarjitam
prakrtya jatinirdesam dharmacakram
ihocyate`.
Ⅱ)`astiti sasvatagraho nastityucchedadarsanam
tasmadastivanastitve nasriyeta vicaksana`.
From the verses quoted above it is evident
that the existence of an eternal indireidual soul
(`atta`) was preached in the philosophy of the
`Sasvatavada` and term 'asti' in its vocabulary did
not convey the ordinary meaning of mere existence
but acquired the special sense of eternal existence.
On the other hand the `Ucchedavada` which preached
the philosophy of `niratta` (`nairatmya`), coined
the term'nasti' to signify simultaneous annihilation
of the temporarily existing `atta` and the physical
body. It is in these special senses that the Pali
Buddhist texts use these two terms 'asti' and
'nasti' while recording the dialogues between the
Buddha and the non-Buddhist ascetics. Any lack of
awareness of the special imports of 'asti' and
`'nasti'` may result in drawing wrong conclusions
from such dialogues.
As the ego-centricity and absolute dependence
on the Skandakas-the two
────────────
(55) PTS, London, 1913； also see, The Group of
Discourses, Ⅱ(PTS, London, 1992) p.304
(56) Leffman, Lalitavistara, p.426
(57) `Madhyamikakarika`, 15.10
422 頁
hallmarks of the concept of `atta-can` be no means
by associated with the `Upanisadic` `Atman-Brahman`,
it is but natural that the `Nikaya` criticisms do
not contain any reference to the `Upanisadic`
concept. The post-canonical early Buddhist texts
also never confused `atta` with the `Atman-Brahman`
of the `Upanisads`. In the `Vajracchedika
Prajnaparamita`(58) `atman` is used together with
such terms as `jiva`, sattva and pudgala. All these
terms are put in the same category and are used to
denote different aspects of the same ego-centric
entity. 'Sacet Bodhisattvasya sattva `samjna`
pravarteta na sa bodhisattva iti vaktavya. tat kasya
hetoh？ na sa `Subhute` bodhisattvo vaktavyo yasya
`atma-samjna` pravarteta, sattva-`samjna` `va`
`jivasamjna` `va` `pudgala-samjna` `va` pravarteta'.
'If in a Bodhisattva the perception of a being
should take place, then he could not be called a
Bodhisattva. And why？ He is not to be called a
Bodhisattva in whom the perception of a soul, or a
being, or a living being, or a person would take
place'. (59) Similary in the early `Madhyamika`
literature we read：'Ko 'yam `atma` yo
`'hankaravisayah'`, which may be translated as
follows： What is this `atman` that is the domain of
ego？(60) Here `atman` (Pali：`atta`) is
characterised by `ahankara`. In this context we may
refer to the definition of 'upadhi' in
`sopadhisesanirvanam as given by `Candrakirti`(61)
in the `Prasannapada`： tatra upadhiyate asminn
`atmasneha ity `upadhih`,
`upadhisabdenatmaprajnapti-nimittah
pancopadanaskandha` ucyante'. Here the concept of
`atma` is held to be caused by the five aggregates
of attachment. The same idea is expressed by
Buddhaghosa (62) while defining `'attabhava'`.
`'Attabhava'` vuccanti `sariram`. `Khandha-pancakam`
eva `va`, tam `upadaya pannatti-matta sabhavato`'.
Conze translates： 'Personal existence means the
dody, or the five Skandhas together, because
dependent on them this mere concept comes about'.
(63) Thus it is clear that these early Buddhist
texts while discussing the concept of `atta` never
thought about the `Upanisadic Atman`, but remained
faithful to the `Nikaya`
────────────
(58)`Vajracchedika Prajnaparamita`(E.Conze, Rome,
Is. M.E.O. 1957) p.29
(59) Also see, ibid.pp.66-67
(60) Poussin, `Mula-Madhyamaka-Karikas`, Ch.18,p.340
(61) Conze, ibid. p.94； Poussin, ibid. p.519-520
(62) Visuddhimagga(h.C.Warren, 1950)Ⅸ,54
(63) `Vajracchedika Prajnaparamita`,p.100
423 頁
characteristics of `atta`, viz. the ego-centricity
and its invariable relationship with the skandhas.
This point has been clearly stated by Vasubandhu. In
his `Vimsatika-vrtti` Vasubandhu(64) states：`'Yo
balair dharmanam` svabhavo `grahyagrahakadih`
parikalpitas tena `parikalpitenatmana tesam
nairatmyam`, na tv `anabhilapyenatmana yo buddhanam
visayah` 'It is because of that imaginary `atman`
which the ignorant people think to be the things'
self-being consisting of subject and object etc.,
that the things are devoid of `atman`, not because
of the ineffable `Atman` which is the domain of the
Buddhas'. The `atta` rejected by the Bhddhas is
"`ahankara-visayah`" whereas the ineffable
`Upanisadic Atman` is `'buddhanam visayah'`.
────────────
(64) `Vimsatika-vrtti`(Ed. S.Levi, paris.1925)p.6
Key words: 1.`Atta` 2.`Niratta` 3.`Anatta` 4.`Upanisad`
424 頁
初期佛教文獻的`atta、Niratta 和 Anatta
穆克紀
中華佛學研究所專任客座教授
提要
本文共分五節，討論外道持的 `atta` (我) 觀究竟如何
，佛教又如何用 `antta` (無我) 論來破它。
第一節引《阿含》和《尼柯耶》，肯定在有為界與無為
界都沒有 `atta` ，它只不過是想像的產品。
第二節介紹近代學者對 `atta` 一詞函義的種種看法。
其中只有 Conze 一人發現，佛教講的 `atta` 和數論所謂的
`purusa` 有相似處。其他所有思考過此一問題的學者或許
由於 `atta`、 `atman` 二詞對等，單單關心 `atta` 與奧義
書 `atman` 有何等關聯，而未經系統的研究，確信二者之間
有關係。此一觀念顯然有必要重新評估。本節則提出較客觀
、適當的研究方法。
第三節簡要敘述持 `atta` 見的常論者和持 `niratta`
見的斷論者主要思想。
第四節分析佛教如何破 `atta` 的信念。 在反駁 `atta`
的基礎上分別擬構出常論者和斷論者所持的觀點，並發現
佛教用三種推理方式來破 `atta` 。 尤其破常見及身見者的
方式顯示辯論的兩種重要的特色：一、以揭發對方說法內在
矛盾來駁斥。這個方法後來的大乘佛教用得非常成功。二、
先接受對方部分說法，把它當做進一步顯示該說法謬誤的基
礎。這或許可以看做大乘佛教所推崇的「方便」。
第五節指出，佛陀提 `anatta` 的中道是針對常論者和
斷論者的極端見解。同時證明常論、斷論所謂 `atta` ，並非
奧義書的 `atman` 。