p000 - 000 Acta Asiatica 118 - 2 Zimmermann
p000 - 000 Acta Asiatica 118 - 2 Zimmermann
Michael ZIMMERMANN
OUTLINE
This essay deals with how the Sanskrit term tathāgatagarbha is used in the
Mahāyāna text to which it gave its name: the Tathāgatagarbhasūtra. Whether the
Tathāgatagarbhasūtra is the oldest text in which the term appears or not, the
particular way in which it is applied there documents that the authors created or
made use of it while associating it with many diverse aspects, less in a philosophical
or abhidharma-like way but rather in a loose and associative style, opening the door
to incorporate the different connotations described metaphorically in the sūtra. The
rich illustrations found in the sūtra help us understand the early context in which the
term was coined and took shape. It will become evident that it is impossible to
reduce the genesis and meaning of the term tathāgatagarbha by way of a mono-
explanatory model.
After a decade of gentle slumber, recent years have witnessed a number of exciting
new publications on the Indian tradition of buddha-nature thought, in Sanskrit
commonly known by the Sanskrit terms tathāgatagarbha and buddhadhātu. The
event for which this contribution has been written aimed at shedding more light on
these two most important terms which were in use when talking about the idea that
all sentient beings carry something very precious within themselves that identifies
them as not being fully part of their samsaric environment, something that links
them to the same essential quality characteristic of buddhas, something that puts
them on the same level with all the buddhas. In general, this precious element is not
known to those who carry it within. And there exist various conceptualizations of
how this precious element is present in living beings. It could be present in them like
a seed that still needs to grow. But there are other descriptions that carry the idea that
1) This essay makes use of ideas and passages found in some of my previous publications
(Zimmermann 2002; Zimmermann 2014).
0020043019 東方学会 初 1∼ 15
8 月 30 日
ACTA ASIATICA 118 2 15 CS 5.5
2 MICHAEL ZIMMERMANN
this element is at all times fully developed inside living beings, and what is needed
is just its purification from impurities. Such impurities are described as being
attached to the element, but only peripheral and temporary, without ever affecting
the essence of that precious element itself.2)
Needless to say, the history of this general idea of sentient beings carrying a
precious essence―in Western publications often called “buddha-nature”―is
complex. The publication of Michael Radich’s monograph on the Mahāparinirvān ・
a-
mahāsūtra in 2015 did not make things easier, but rather led to a different theory
about the very origin of buddha-nature thought in India.3) Radich argued, based on
his own research and on several articles by Stephen Hodge, an independent scholar
based in the UK,4) that the Mahāparinirvān ・
a-mahāsūtra (MPNMS) is the earliest
text propounding that all living beings have buddha-nature, originating in its oldest
parts already as early as the second century C.E.5) Radich argues that the ideas that
led to the genesis of the term tathāgatagarbha originated as part of a wider pattern
of docetic Buddhology: ideas holding that buddhas are not really what they appear.
In doing this, Radich gave the whole discussion of the origin of this strand of
thought a new origin and assumed a new motive with regard to the creation of this
idea.
At the same time, Radich’s new chronology wiped out the hitherto assumed
history of ideas relating to buddha-nature, which had positioned the
Tathāgatagarbhasūtra (TGS) as the origin, with the śrīmālādevīsim ・
hanādasūtra
(śMS) and the Anūnatvāpūrn ・
atvanirdeśa (AAN) following it. Takasaki Jikido, the
“father of modern tathāgatagarbha studies,” had once introduced the idea of a core
triad of tathāgatagarbha-related sūtras (nyoraizō sanbukyō 如来蔵三部経),
comprising the AAN, the TGS and the śMS. He proposed a chronological sequence
beginning with the TGS, followed by the AAN, and, as the latest among the three,
the śMS.6) Takasaki’s arrangement of these texts as a triad has also recently been
fundamentally questioned by Jonathan Silk, in his monograph “Buddhist Cosmic
Unity,” published also in 2015.7) Silk’s research came to a different conclusion, based
on an analysis of the doctrinal content of the AAN and its literary style. For Silk
there is very little evidence that the śMS might be positioned chronologically before
the AAN. He further questions the very arrangement of these three sūtras as the
above-mentioned tathāgatagarbha triad. I think that Silk’s arguments are convincing,
2) Both of these conceptions are discussed in Zimmermann 2014: 514–517.
3) See Radich 2015.
4) See Hodge 2010/2012 and Hodge (unpublished).
5) Radich 2015: 99.
6) See Takasaki 1975: 39–126.
7) See Silk 2015.
and again, we can see how an established idea about the relative chronology
regarding some crucial texts in a tradition, after decades, tumbles and falls apart. A
new convincing model of how the intellectual history of buddha-nature unfolded as
a whole is yet to be penned and argued.
What do these new research results mean for the study of the history of terms
relating to buddha-nature in India? In order to clarify this point, let me explain a
little more about the assumptions made in past decades on this issue. In my own
research I had followed the approach taken by Takasaki Jikido, who had firmly
assumed that the TGS was the first scripture to propagate the idea of buddha-nature
in explicit terms. Takasaki’s main argument was based on the assumption that the
MPNMS contains a quotation from the TGS and thus the TGS must be the older text
of the two.8) In the process of editing the Tibetan and Chinese versions of the TGS I
also worked on an in-depth analysis of its content. What became clear on the basis of
a detailed look into the structure and terminology of the sūtra in all of its available
four versions was the following: as is widely known, the core of the TGS is formed
of a section of nine similes, starting with an introductory description of huge lotus
flowers in the sky, in which buddhas are seated, all with a beautiful appearance.
When suddenly the leaves of the lotus flowers start to wither, turning ugly and foul,
they close themselves and the buddhas become wrapped in the lotus flowers’ ugly
petals. This introductory imagery at the beginning of the text also forms the first
simile and defines the main structural pattern for the following eight other similes
(see Table 1).
All of these nine similes center around the idea that the buddha-nature of living
beings is hidden by ugly coverings, i.e., their emotional and intellectual defilements
(Skt. kleśa), but just like the buddhas within the flowers, their buddha-nature remains
unaffected by the ugly surroundings and, at the same time, invisible to other sentient
beings. To make things a bit more complicated, at first glance, the introductory
section of the TGS and the following exposition of nine similes seem to be from one
mold. The descriptions in the similes are all vivid and concrete. The similes appear
to be constructed in a unified fashion. They all start with the description of a more or
less well-known occurrence or process met with in daily life (with the exception of
the lotus flowers in the sky in the first simile), and then draw a comparison between
it and the spiritual sphere. At the end, the prime role of the Buddha is illustrated as
8) See Takasaki 1975: 128–190; for an attempt to date the TGS, see Zimmermann 2002:
69–75.
Table 1
that of teaching living beings about their unknown potential of buddhahood. From a
doctrinal point of view, too, there seems to be no reason to assume that the similes
were not the creation of a single author or group of authors put forward as a more or
less monolithic construct. Nevertheless, an in-depth examination of the first simile
leads to several observations that clearly go against the assumption that the
introduction and the nine similes had already been arranged in their present form
from the very beginning.
The most obvious manifestation of textual heterogeneity is the first simile itself.
After his announcement that he will expound the TGS, the Buddha sets about
formulating the first simile, which, regarding the upamān ・
a (the part of the simile
used to illustrate the subject matter), is taken from the setting of the introductory
section. Then, in the main section, he abruptly starts to recount this first simile once
again, this time with an emphasis on different elements. A close look at both
“versions” of this first simile, that is to say, the “old” and “new” versions, reveals
several inconsistencies which indicate that the two parts do not correspond
harmoniously. A detailed analysis leads to the conclusion that these inconsistencies
in the flow of the text can be somehow explained if we are to assume that here two
different lotus similes have been combined, with only the “new” simile using the
term tathāgatagarbha, while the “old” lotus simile draws upon the introductory
setting as described before. The “new” simile clearly follows the structure and
narrative line of the other eight following similes (2–9).
What we can see from these (and some other) instances is that originally we had
to deal with two separate parts: on the one hand, the introductory section flowing
coherently into the “old” simile (without any mention of the term tathāgatagarbha),
and, on the other hand, a series of eight similes (again, without any mention of the
term tathāgatagarbha), which, for one reason or another, somebody wanted to
combine with the introductory setting and the “old” lotus simile. In order to
structurally adapt the “old” simile to the set of eight, a “secondary” or “new” lotus
simile was interpolated into the original one, which was structurally parallel to the
following eight similes. Without looking at the complete first simile in some detail,
hardly anybody can notice the fact that here two different accounts of the lotus
simile come together. The “new” lotus simile functions as the joint between two
parts, i.e., the introductory setting on the one hand and the eight following similes on
the other. More importantly, the “new” lotus simile introduces the compound
tathāgatagarbha, which does not show up in either of the two parts before and after
it.
For an understanding of the genesis of the term tathāgatagarbha and the wider
concept of buddha-nature these facts are, I think, quite suggestive. It is certainly
possible that the technical use of the term tathāgatagarbha as relating to buddha-
nature predates the TGS itself and that the sūtra’s aim was primarily to illustrate
lucidly how this term should be understood: a kind of subsequent explanation of an
existing idea already “in use.” In other words, someone wanted to give an
explanation intended to shed light on this term and its associated ideas through easily
comprehensible formulations in a kind of follow-up illustration. This explanation
now seems more likely in light of the research by Michael Radich on the MPNMS,
which has been discussed above. One of the questions to be asked in that case is why
throughout the nine similes of the TGS the term tathāgatagarbha is used exclusively
in the lotus simile. If the author(s) of the TGS were aware of the buddha-nature
tradition of the MPNMS and wanted to add another expression of this line of
thought, why would they not use this crucial term widely throughout all of the
similes? One possible answer could be that the similes of the TGS emerged from a
different geographical or social context: a context in which the TGS author(s) were
not aware of the tradition initiated by the MPNMS. Once they learned about the
similarity between their own ideas and those espoused in the MPNMS, they tried to
introduce the key term tathāgatagarbha into their own literary product, but did so
only subsequently, as I have shown above. Whether the similes used in the TGS at
that time were a rather fresh composition or whether they were part of an older
tradition cannot be decided at this time. Equally, the possibility that both sūtras, the
TGS and the MPNMS, originated in approximately the same period in different
contexts cannot be excluded. The scarcity of information about literary works in
matters of date and origin in ancient India is well known.
However, judging from the textual evidence in the TGS alone, this is not the only
possible scenario we can paint based on analysis. The term tathāgatagarbha could
equally be an ad hoc creation resulting from combining the two previously separate
parts (introductory setting and the “old” simile on the one hand, and the following
eight similes on the other). Alongside the process of joining the two parts, while
composing the “new” lotus simile, the term tathāgatagarbha could have been
created in its technical meaning relating to the buddha-nature of all sentient beings
as a fitting and comprehensive “catchword” for what was deemed to be the essence
of the set of similes. In both cases (case 1: the term tathāgatagarbha was inherited
from another text such as, e.g., the MPNMS; case 2: the term tathāgatagarbha was
newly coined by the composer(s) of the TGS), the sūtra’s introductory setting with
the first simile offers a plausible explanation for why the use or the creation of this
catchword would have been considered particularly fitting in this context. Here I
mean the several occurrences of the Sanskrit term padma-garbha (“lotus calyx”) in
the introductory setting. In these lotus calyxes (garbha), in the introduction of the
sūtra, full-fledged buddhas are seated in perfect meditation. The petals of the lotuses,
however, suddenly wither and turn into ugly, putrid, disgusting sheaths, covering the
beautiful buddhas still sitting motionless inside (garbha) the petal sheaths. The same
situation is then claimed to hold true for the buddha inside sentient beings: though
each sentient being hosts a buddha within itself, sentient beings are not aware of it
because this buddha is still covered by their emotional and intellectual defilements.
Here, garbha, in both compounds, means “inside”: the interior space of a lotus
(padma-garbha) or the inside of living beings. On the basis of this lotus simile it is
thus clear how we should understand the term tathāgatagarbha when it is stated that
all living beings are tathāgatagarbha: i.e., that they all carry a tathāgata within
themselves. In terms of Sanskrit grammar, in this instance we are dealing with a
bahuvrīhi compound, an adjective referring to living beings, with garbha meaning
nothing more than “containing,” in the sense of containing a tathāgata―quite
different from what it can otherwise also designate, namely, “embryo, womb,
offspring, seed, essence, kernel, or member (of a family lineage).”9) This, however,
does not mean that the interpretation of the term tathāgatagarbha in the TGS as a
whole would be confined to this more technical or rather “weak” notion of garbha in
the sense of “inside” or “to contain” once the other eight similes were brought into
play. Even though the term tathāgatagarbha does not appear in any of the other
eight similes, the scenes depicted in these eight similes, illustrating how the buddha-
nature of living beings could be imagined, would almost certainly widen the range of
possible interpretations of the compound and shape its understanding.10)
All things considered, these attempts to explain the genesis of the term
tathāgatagarbha are highly speculative. They are indicative of the complexity of the
source material and the difficulties that come along with it when we aim for
straightforward conclusions. One of the conclusions we might certainly draw from it
is that it is probably not possible to decide with any high degree of probability where
the origin of some of the key words of buddha-nature thought lie.
Let us have another look at the above-mentioned contribution of Michael Radich and
its consequences for establishing the history of buddha-nature theory in India.
Radich puts forward a strong argument regarding the possible date of the MPNMS,
or, one should say, the two main “layers” of the MPNMS, of which one layer deals
with the eternity of the Buddha, stressing the dharmakāya-aspect (his vajrakāya,
“diamond body”), and the second layer emphasizes tathāgatagarbha, suggesting that
they are, also seen in light of the types of practices they condone, products of
different groups.11) It seems to be equally clear that the second layer dealing with the
buddha-nature of sentient beings is the later one of the two. Even though the first
Chinese translation is only attested at the beginning of the fifth century, internal
textual evidence suggests that it can be approximately dated to as early as the second
to early third century. I will not go into details of the flow of Radich’s arguments, but
at their center stands a prophecy as it is found in the MPNMS and a closely
associated group of Mahāyāna sūtras: the Mahāmeghasūtra, the
Mahābherihārakasūtra, and the An ・
gulimālīyasūtra.12) In this prophesy the Buddha
explains that
9) For a discussion of the terms tathāgatagarbha and garbha, see Zimmermann 2002: 39–50.
10) So, e.g., the eighth simile in the TGS of the embryo of a universal emperor in the womb
of a poor, depressed woman (Zimmermann 2002: 135–140). There, the compound garbha-gata or
garbha-stha appears.
11) See Radich 2015: 19–22.
12) See Radich 2015: 59–85.
“the text will first flourish for ‘forty years after my parinirvān
・
a’, but then it will
disappear. Subsequently, however, it will re-emerge when the true Dharma
(saddharma) has eighty years left in the world. It will flourish most in the first
forty years of this eighty year period. In other respects, this period will be a
terrible ‘endtimes of the Dharma’,…”13)
buddhadhātu, strongly associated with relic worship, hardly appears in the TGS. It is
mentioned one single time in a passage where the Tibetan reads rigs (for Skt.
gotra?).25) However, the quotation of this passage in the RGV reads dhātu, which is
also confirmed in one of the Chinese translations of the TGS. The reason here could
be similar to the reason given above: if the term buddhadhātu was heavily associated
with the relic context, the authors of the TGS would have tried to leave this
association behind. Dhātu, at the same time in the meaning of “element,” can carry a
rather substantial connotation, but can equally well have multidimensional
implications in terms like dharmadhātu (“essence of the dharma”) or sattvadhātu
(“realm of beings”), but of course also hetu (“cause”).
No doubt, all these attempts to explain and interpret key-words in the tradition and
my attempts to determine the circumstances of the genesis of a text like the TGS
must remain very speculative, with many factors in play that we simply do not know.
And that leads us back to the original question of this essay. As expressed in the title,
my aim was to show that it is impossible to reduce the genesis and meaning of the
term tathāgatagarbha by way of a mono-explanatory model. The situation we face
for the early intellectual history of Indian Mahāyāna Buddhism is very different, for
example, from that of research on a philosopher or religious leader who lived in the
1920s in Germany and came up with a new understanding of the psycho-spiritual
constitution of human beings. In this case, as a scholar of intellectual history, we
would have access to all his writings, in the best of cases even unpublished papers.
We would know about the place where he lived and its social, political, and
intellectual environment. We would know what kind of literature he consumed and
maybe even have access to his library. And we would know who his friends,
followers, and enemies were. In short, all this information would help us immensely
in an attempt to understand how this person’s ideas came into existence in the
particular context of his time, what his understanding of particular terms was, by
whom he might have been influenced, and where his own original contribution lies.
We would, under optimum conditions, be able to trace the major factors informing
his thinking and be well-equipped to discuss his contribution to the wider field of the
intellectual, social, and religious setting of his epoch.
It is obvious that we do not have resources of this sort pertaining to the
intellectual-religious history of India from around the first centuries of the Common
Era. This makes things highly speculative at best. We have to be very careful not to
oversimplify developmental complexities, many of which we do not yet know in
full, for the sake of a simple mono-causal explanatory model. Also, we must keep in
mind that most key terms carry multi-associative connotations―even more so during
one of the most vibrant, creative, multi-religious and, in a sense, international
periods of Indian intellectual history, as the first centuries of the Common Era no
doubt were. Radich is certainly right when towards the end of his monograph he
states: “However, I do not claim that the arguments of this chapter can exhaustively
account for the origins of tathāgatagarbha doctrine.”26)
Another way to illustrate this point is through a table showing the different terms
used to indicate the buddha-nature of sentient beings in the similes of the TGS.
Similes are, by their very nature, highly associative attempts to describe difficult
spiritual or metaphysical notions in more easily accessible patterns. The literature on
buddha-nature is, not surprisingly, full of such similes, and thus terminology brings a
high degree of associativeness with it, on the one hand opening the door to
imaginative contexts of diverse real-life situations, and on the other also leading to
associated issues of soteriological relevance that arise in this context. The TGS is a
brilliant example of this point, and this can be seen in the diversity of terms used in
order to give name to the idea of how the buddha-nature of all sentient beings can be
imagined.
8B: de bzhin gshegs pa’i rigs, but *tathāgatagotra: but quoted in RGV(V) as
Bth: de bzhin gshegs pa’i kham ・
s tathāgatadhātu
de bzhin gshegs pa; tathāgata
8C: chos nyid (twice) dharmatā (2 times)
CONCLUSION
At this early stage of buddha-nature thought it should be obvious that we are dealing
with a genre of literature that is rich in narratives and similes while relying on
terminology that does not sharply define how buddha-nature is to be understood.
This is common for the early stages of new spiritual or religious movements that are
still developing a well-structured system of thought. It is exactly this “not being very
precise” that may open the door to associating key terms from various kinds of
contexts, a strategy perhaps willingly employed in order to increase the potential to
propagate the new teachings more effectively among groups that would initially not
be interested in them.
Obviously such circumstances make it very difficult to decide from which specific
background the newly introduced terminology has developed. Even for the “creators”
of a specific term, the situation might change in a moment after the term was
introduced: other equally interesting models of understanding and interpreting the
term might arise and stand juxtaposed to each other. This is the reason that I think
tathāgatagarbha is not always tathāgatagarbha, as expressed in the title of this
essay. The understanding of this term might vary, depending on the context in which
it is used and on the imaginative world of the listener or reader. I suggest that there
is no need at all to attempt to apply mono-causal explanatory models. Rather, as
scholars of Buddhist studies, we should be proud that we are dealing with a rich,
imaginative, and often little regulated tradition, which―perhaps precisely because
of this―proves to be a very creative and adaptable strand of religious thought today,
just as it has been throughout the last two thousand years.
Abbreviations
Bth Second Tibetan translation (paracanonical) of the TGS in the Newark manuscript
Kanjur from Bathang (see Zimmermann 2002)
AAN Anūnatvāpūrn・
atvanirdeśa
MPNMS Mahāparinirvān ・
amahāsūtra
RGV(V) Ratnagotravibhāga(vyākhyā) (text quoted from Johnston 1950)
Skt. Sanskrit
śMS śrīmālādevīsim
・
hanādasūtra
TGS Tathāgatagarbhasūtra
References
Hodge, Stephen. 2010/2012. “The Mahāyāna Mahāparinirvān ・
a-sūtra: The Text and Its
Transmission.” Corrected and revised version of a paper presented in July 2010 at the
Second International Workshop on the Mahāparinirvān ・
a-sūtra, Munich University.
Hamburg: Zentrum für Buddhismuskunde, 2012. https://www.buddhismuskunde.uni
-hamburg.de/pdf/5-personen/hodge/the-textual-transmission-of-the-mpns.pdf. Accessed
20 November 2018.
― (unpublished). “Who Compiled the MPNS, Where & When? The Geographical and
Historical Origins of the Tathāgata-Garbha Doctrine.” Unpublished paper.
Johnston, Edward H., ed. 1950. The Ratnagotra-vibhāga Mahāyānottaratantraśāstra. Patna:
Bihar Research Society.
Jones, Christopher V. Forthcoming. The Buddhist Self: On Tathāgatagarbha and Ātman.
University of Hawai‘i Press.
Radich, Michael. 2015. The Mahāparinirvān ・
a-mahāsūtra and the Emergence of
Tathāgatagarbha Doctrine. Hamburg Buddhist Studies Series 5. Hamburg: Hamburg
University Press.
Ruegg, David Seyfort. 1969. La théorie du tathāgatagarbha et du gotra. Paris: École
Française d’Extrême-Orient.
Shimoda Masahiro 下田正弘. 1997. Nehangyō no kenkyū―Daijō kyōten no kenkyū hōhō
shiron 涅槃経の研究 ― 大乗経典の研究方法試論 [A study of the Mahāparinirvān ・
asūtra:
With a focus on the methodology of the study of Mahāyānasūtras]. Tokyo: Shunjūsha 春
秋社.
Silk, Jonathan. 2015. Buddhist Cosmic Unity: An Edition, Translation and Study of the
Anūnatvāpūrn ・
atvanirdeśaparivarta. Hamburg Buddhist Studies Series 4. Hamburg:
Hamburg University Press.
Takasaki Jikidō 高崎直道. 1975. Nyoraizō shisō no keisei―Indo daijō Bukkyō shisō kenkyū
如来蔵思 想の形成 ― インド大乗仏教思想研究 [The formation of tathāgatagarbha
thought: A study of Indian Mahāyāna Buddhist thought]. Tokyo: Shunjūsha.
Zimmermann, Michael. 2002. A Buddha Within: The Tathāgatagarbhasūtra, The Earliest