Open Collections

Digital Himalaya Journals

Padma Dkar-po on Integration as Ground, Path, and Goal Broido, Michael M. 1985

Item Metadata


JSON: dhimjournal-1.0364945.json
JSON-LD: dhimjournal-1.0364945-ld.json
RDF/XML (Pretty): dhimjournal-1.0364945-rdf.xml
RDF/JSON: dhimjournal-1.0364945-rdf.json
Turtle: dhimjournal-1.0364945-turtle.txt
N-Triples: dhimjournal-1.0364945-rdf-ntriples.txt
Original Record: dhimjournal-1.0364945-source.json
Full Text

Full Text

Michael M. Broido
II.I: Yuganaddha ('integration) in general
That yuganaddha (Tibetan zung-'jug) is one of the most important
technical terms of the vajrayana is a platitude which needs no comment from me; one need only look at the writings of Prof. H. V.
Guenther or of Prof. Alex Wayman. At the same time it is hardly possible to think that this word has been adequately explained by either of
these distinguished authors or by anyone else. It is clear that the use of
this word by different1 Tibetan writers varies; as in Part I, the treatment
here will be restricted to Padma Dkar-po. However, Padma Dkar-po
himself has criticized various earlier Tibetan authors in this area,2 and
we will be able to bring some of the differences which he discusses into
sharp focus.
Both Guenther and Wayman have used a great variety of English
renderings for yuganaddha. In Guenther we find e.g. "unity",3 "unity
of opposites",4 "union of opposites",5 "unitive Being",6 "togetherness",7 "coupled together",8 "harmoniously blended",9 "harmonious
juxtaposition",10 "coincide",11 and "coincidence".12 Given the overriding importance of this term for Guenther, he is careless about his use of
these words to represent the views of different authors; for instance it
is Padma Dkar-po13 (and not the Bka'-brgyud-pas in general14) who in
criticizing Tsong-kha-pa, attributes to him, as I will show below, a view
of yuganaddha which might be interpreted as "harmonious juxtaposition", a phrase, however, which does not fit the attributed passage
very well. Most of Guenther's other renderings of yuganaddha represent either Padma Dkar-po's view (as he understands it) or Indian authors, such as Naropa, whose views Padma Dkar-po accepted.
In Wayman's writings we find e.g. "pair united",15 "pair combined",16 "pair-wise united",17 "combined together",18 and "coupling".19 On the whole these renderings represent (confusedly, as we
This is the second of two papers appearing in successive issues of the Journal of
the Tibet Society, the first being called Padma Dkar-po on tantra as ground, path and
goal. In the present paper, references to "Part I" are to this earlier paper.
will see) a "lexical" approach to the notion of translation, but insofar as
they represent the view of a specific author, it is Tsong-kha-pa (and, in
LW, his pupil Mkhas-grub-rje).
In the introduction to his translation of the hagiography of Bu-ston,20
Prof. D. Seyfort Ruegg has several times translated zung-'jug by "integration", representing the views of various Sa-skya and Dge-lugs authors whom he quotes at some length. For reasons given below I feel
that this word is a happy compromise between various Tibetan views
of zung-'jug, and so I will sometimes use ""integration" as a cipher for
yuganaddha or zung-'jug.
Methodologically, Seyfort Ruegg (and perhaps Wayman) is committed to a "lexical" notion of translation, in which one English word is
used uniformly to translate one foreign word. Guenther seems to dislike this approach, though it is not clear to me what he wants to put in
its place. In any case the lexical approach hardly constitutes a method of
translation, since it tells us nothing about which word to use uniformly
for each foreign word. When interpreting, I shall concentrate, on the
whole, on sentences, without worrying overmuch about uniformity;21
but when focussing attention on single words such as yuganaddha I
shall follow Gilbert Ryle's recommendation to consider carefully their
logical grammar: features such as the syntactic type (whether noun,
verb, predicate with 1,2,3 . . . places, &c). Ryle also emphasized the
semantic categories of words with which a word may be combined,
and though his notion of category is perhaps not clear enough to provide a basis for solving philosophical problems22 it is still of heuristic
value in demarcating specific philological ones. As Ryle showed in
English, these modest considerations may not tell us the meaning of a
word, whatever that is, but at least they may prevent us from making
Padma Dkar-po's own "etymology" of yuganaddha23 may be rendered thus: "yuga means 'joined' and naddha means 'non-dual', and so
yuganaddha is said to be "Integration".24 The phrase "and so" (de'i phyir)
is tongue-in-cheek. The syllable-by-syllable analysis is not meant to be
literal, but merely to give the right result.25
The full form of the Tibetan zung-'jug is found quite frequently, and
is zung-du 'jug-pa. The syntax of this full phrase gives us an important
part of its meaning, 'fug-pa is in general a verb or a verbal noun (Skt.
vrt-, vrtti) and here is a verbal adjective (modifying a noun to yield another noun). Zung-du just modifies this verbal adjective (yielding another). (Of course the semantics do not follow this simple course; it is
partly for this reason that Padma Dkar-po's nirukta must be so non-literal — see note 25 — and also that lexical translation into English runs
into problems.) Also yuga or zung acts syntactically as a two place predicate (two things are joined). The logical grammar described so far may
be summed up in the schema:
(X,Y) zung-du 'jug-pa'i   Z (1)
where of course the semantics will impose restrictions on what can fill
the three places here labeled with variables X, Y, Z. In Tibetan the two-
variable slot (X,Y) is usually filled by some phrase denoting two things,
either in the explicit form X dang Y (X and Y e.g., thabs dang shes-rab),
yielding from (1) the schema
X dang Y zung-du 'jug-pa'i   Z (2)
or by contraction of such a phrase (e.g. thabs-shes, corresponding more
closely to the Sanskrit prajhopaya); or by a phrase of some other form
but similar function (e.g. bden-gnyis, Skt. satyadvaya). Most of these features are found in the phrase26
bden-pa gnyis-ga gnyis-su mi-phyed-pa zung-'jug-gi ting-nge-'dzin (3)
which may perhaps be translated "the samadhi which "integrates the
two satyas inseparably". This concludes the summary of the syntax of
On its semantics according to the schemata (1) and (2), the first thing
is to say something about the semantic character of what may replace
X, Y and Z. Now Z is normally omitted, but when it appears, the word
substituted is almost always, as in (3), ting-nge-'dzin (Skt. samadhi). I believe that any appearance whatever of zung-'jug or zung-du 'jug-pa (at
least in the kind of context considered in this paper) is to be taken as an
ellipsis of zung-du jug-pa'i ting-nge-'dzin (Skt. yuganaddhasamadhi). In
particular we see quite often the phrase
zung-jug-gi rim-pa   (yuganaddhakrama) (4)
which according to Padma Dkar-po28 is short for
zung-'jug-gi ting-nge-dzin-gyi rim-fa   (yuganaddhasamddhikrama). (5)
The expansion of (4) into (5) is not special to the yuganaddhakrama; he
describes all six (sic) stages of the Pancakrama as ting-nge-'dzin.29 Similarly, Padma Dkar-po's most detailed discussion of yuganaddha, that in
the gzhung-'grel,30 opens with the following verse from the Vajramala:
/phyag-rgya chen-po'i dngos-grub chel I zung-du 'jug-pa'i ting-'dzin-gyisl
Irnam-rtog med-par zhugs-nas nil Ithob-pa 'di-la the-tshom-med!
which means roughly: "If one enters into non-discursiveness by means
of yuganaddhasamadhi, without a doubt one will obtain the great siddhi
of mahamudra." Another standard phrase is zung-du 'jug-pa'i sku
where again the standard contexts31 make it clear that a samadhi is intended.32
The claim, then, is that yuganaddha (zung-'jug) is a technical term
used in Buddhist (and especially vajrayana33) soteriology; a samadhi, a
state of mind. As far as I know there is no colloquial use competing
with this technical use. From this point of view the problem of finding
a suitable translation is quite different from that of translating terms
like rlung (vayu).34
Now we turn to the main remaining feature of the schemata (1), (2),
namely the two variables X and Y, that is, the two items joined or
"integrated in yuganaddha. Padma Dkar-po has pointed out several
times that in this respect the terms zung-'jug and lhan-skyes (sahaja) are
very similar.35 In both cases the fundamental feature is the inseparability
of the two items so related; this feature is already present in the example (3). A further point upon which Padma Dkar-po not only insisted
but contrasted his view with that of others is that this inseparability is
non-contingent (ma-bcos-pu36). However, it seems that this non-contingency applies only to certain zung-'jug pairs, viz. those which are also
lhan-skyes, i.e. "born together", roughly the "positive" type discussed
below; these are the most interesting and controversial cases. Sahaja
(lhan-skyes) is a two-place relational attribute, saying of the two terms
that they are born together or arise or emerge together, rather than
separately.38 There is no one term with which it stands in a privileged
relationship (as does zung-'jug with ting-nge-'dzin), and this difference
emphasizes the extent to which lhan-skyes is a purely relational attribute, in contrast to the way zung-'jug stands for a kind of individual in
the broadest sense (viz. a kind of state of mind).
Next we can ask: is there just one yuganaddhasamadhi, of which the
different types listed in the standard sources (see below) are merely
different aspects? Or are there several different such samadhi? The texts
give no clear answer to this question. A possible line of attack is this:
we may look at the possible ways of individuating a thing such as a
samadhi. These will provide us with possible criteria on the basis of
which individual yuganaddha-samddhis might potentially be individuated. Then we can look at the actual ways in which yuganaddha is described in the texts, and see whether the differences in these modes of
description can be related in any way to those possible criteria. Two
kinds of criteria suggest themselves. First, there are purely physical criteria, such as are used to distinguish between more familiar states of
mind, such as sleep-states (frequency of the dominant electromagnetic
mode active in the brain, movements of the eyeballs, &c). It seems
perfectly plausible to suppose that some of the different types of dhyana
and samadhi distinguished in Abhidharma texts and elsewhere might
be individuated in this way. But when we review the different descriptive phrases used of yuganaddha, as I will do later, it begins to seem
very implausible that they could be correlated with such physical criteria. (Still, I have no reason to dismiss it as impossible.) Second, one
might try to use dispositional criteria. Such-and-such a kind of samadhi
will dispose the subject to behave in such-and-such ways . . . The only
criteria of this kind, relevant to yuganaddha, which I can think of are
those related to the powers and activities of the buddha and perhaps of
bodhisattvas on the highest levels. It is not clear to me that there are
any genuinely empirical distinctions in this area, but suppose there
are. It seems most unlikely (say from Padma Dkar-po's own descriptions) that these will be clearly related to the different pairs of
yuganaddha items listed in the Yuganaddhakrama, or to Birupa's snang-
stongi'gsalstonglbde-stongi'rigstong. It might seem somewhat more
plausible that such criteria might be related to the saiksalasaiksa (slob-
pa'ilmi-slob-pa'i) distinction in yuganaddha. But, as we see in the next
paragraph, there is independent textual evidence that these two form a
continuum of similar states. Accordingly, from now on I shall tentatively assume that we are dealing with one basic yuganaddha, perhaps
in slightly varying forms or aspects, and not with a family39 of different
ones, related by some family resemblance. Should this tentative assumption prove mistaken, it will not be difficult to reformulate most of
the remaining arguments.
According to Padma-Dkar-po, saiksayuganaddhai0 is the period from
the understanding of the radiant light until ultimacy, and corresponds
to the distinction of the bhdvanamarga (sgom-lam), during which one cultivates satya (already) freshly seen.41 Its counterpart asaiksayuganaddha
is just the buddha-level or ultimacy42 about which nothing more need
or perhaps can be said.43 But these are not two different yuganaddha;
rather the saiksa stage is already the real thing, as far as it goes, and he
criticizes the views of those who construct an artificial saiksayuganaddha
(see below).
The term yuganaddha has a long history in India in connection with
non-tantric Buddhist meditation, both in the Hmayana and the Mahayana. We find dhydna and samadhi used almost interchangeably, Tibetan
often using ting-nge-'dzin for both. Samatha (zhi-gnas, mental
quiescence) and vipasyand (lhag-mthong, insight) are first practiced separately and later joined together: yuganaddha, Tib. zung-'brel or zung-du
'brel-ba, and not zung-'jug. They are associated with one-pointedness
(ekagrata, rtse-gcig) of mind. In Tibet all these terms are prominent in
non-tantric mahamudra meditation (see note 45) and as such are frequently discussed by Padma Dkar-po, who was famous for his expert
descriptions in this area. Some details:
(a) Poussin (ADK vol. V, p. 131, n.2) gives references to the Pali literature with the spelling yuganandha, about which he is unhappy, e.g.
samatha-vipassanam yuganandham bhdveti, &c. He translates with
"attele". He also gives refs. to BCA, Siks., MSL &c.
(b) Vaibhasika texts such as Sarighabhadra's Abhidharmakosa-kdrika-
sdstra-bhasya (Sde-dge Bstan-'gyur, Mngon-pa, vol. khu 95bl), and the
Abhidharmadipa (author unknown, ed. Jaini) do mention one-pointed-
ness (ekdgratd) where one would expect it, viz. at the beginning of
chapter VIII (resp. 265a5 and p. 404) but without reference to
(c) Abhidharmakosa runs: samdpattih subhaikdgryam (Isnyoms-
'jug dge-ba rtse-gcig-pal). Vasubandhu comments: . . . / sa hi samatha-
vipasyandbhydm yuganaddha-vdhitvdd drstadharmasukhavihdra uktah (de ni
zhi-gnas dang lhag-mthong zung-du 'brel-bar ngang-gis 'byung-ba'i phyir
mthong-ba'i chos-la bde-bar gnas-pa dang/ lam sla-ba zhes gsungs-pas: Sde-
dge khu 66b4, quoted by Padma-dkar-po, Mngon-mdzod 152b4). According to the Vyakhyd, this is a sutra passage. The normal sense of vahin is
"conveying along" &c, and Poussin has accordingly rendered
yuganaddhavdhin (in a longer phrase taken from the Vyakhyd, Shastri ed.
vol. 4 p. 1128) by "atteles au joug". However, the Tibetan in the ADK
passage is ngang-gis 'byung-ba (spontaneously generated, cf Jaschke s.v.
ngang), and this is in broad conformity with the sense "causing, producing, effecting" for vahin (Monier-Williams, s.v.). So it is clear that
Poussin's colloquial translation receives little support from the Tibetan.
But every other occurrence of yuganaddha mentioned in paras, (a)-(g) is technical, relating to the samatha-vipasyana combination.
(d) Mahdydnasutrdlahkdra XIV.9ab runs:
yuganaddhasca vijheyo mdrgastatpinditam punah I:
/zung-du 'brel-ba'i lam ni del Ibsdoms-pa yin-par shes-par-byal',
and Vasubandhu's comment on this group of verses includes: samatha
manaskarah I vipasyana manaskarah I yuganaddha manaskarah I (confirming
the feeling associated with the word ekdgratd that we are not yet near
the goal). Sthiramati in his Sutrdlahkdra-vrtti-bhdsya (Sde-dge Bstan-
'gyur, Sems-tsam, vols, mi-tsi) comments on this verse (mi, 266a6):
"At the time when samatha and vipasyana have not yet been joined
together (zung-du ma-'brel-ba'i dus-na) they are "based (dmigs-pa;
dlambana in verse 10b) separately. Samatha is "based on words and
letters, while vipasyana is "based on (their) artha. But when they are
joined together (zung-du 'brel-ba) there is no separate "basis of
words and artha, the two are merged (bsdus-nas) and samatha and
vipasyana are unified (mgo mnyam-du 'brel-te), and the path based
on this is called the path of samatha-vipasyand-yuganaddha. By this
means, yuganaddha-manaskdra is taught."
(Here, dmigs-pa = dlambana is not adequately represented by "based,
"basis, but this complex problem cannot be dealt with here. Yuganaddha
= zung-du 'brel-ba throughout.)
(e) Bodhicdrydvatara VIII.4a runs: samathena vipasyana suyuktah, and
the Pahjika comments: samathah cittaikdgratd-laksanah samddhih . . .
vipasyana yathdbhuta-tattva-parijhdnasvabhdvd prajhd I tayd suyuktah I
yuganaddha-vdhi-mdrga-yogena kurute klesdndm vindsam ... in full conformity with (c), (d).
(f) A single example from Padma Dkar-po must suffice. At Rnal-'byor
bzhi'i mdzub-tshugs 6a3 (i.e. in the rtse-gcig, ekdgratd section) he contrasts
zhi-lhag zung-'brel with zhi-lhag zung-'jug. Thus: zhi-lhag zung-'brel-du
bshad-pa'i lhag-mthong ni lhag-mthong gi don-spyi yin-lal zhi-lhag zung-'jug-
gi zhi-gnas kyang/ 'di'i skabs-kyi zhi-gnas ma-yin-nol. The contrast is just
what is suggested in the rest of paras (a)-(g) in relation to this article as
a whole.
(g) Summary: in translating yuganaddha (especially in a non-tantric
context) the Tibetans distinguished between zung-'jug (the goal, and
with little or no manaskara) and zung-'brel (joining or combination, not
the goal, connected with ekdgratd and manaskara and virtually always
used of the samatha/vipasyana combination). Both words are used almost exclusively of meditational states, of samadhi or its close relative
Though zung-'jug is mainly a term of the vajrayana, it also has a use
in non-tantric mahayana44 rather as does mahamudra.*5 With regard to
early Tibetan madhyamaka, in the oral tradition I have heard it said
that Pa-tshab Lotsawa, Rngog Blo-ldan Shes-rab, Phya-pa Chos-kyi
Sengge and others spoke a good deal of bden-gnyis zung-'jug and that it
was as a reaction against the over-emphasis on this connection between
the two satya in his own time that Tsong-kha-pa was moved to write
the very detailed separate treatment found in his Rigs-pa'i rgya-mtsho
and Dgongs-pa rab-gsal, against which Mi-bskyod Rdo-rje and other
Bka'-brgyud-pas and others later protested. Although Mi-bskyod Rdo-
rje does not seem to use the phrase bden-gnyis zung-'jug in the Dwags-
brgyud grub-pa'i shing-rta his account of the history of Tibetan madhyamaka is broadly comparable with this one, which however I am not in
a position to verify. Padma Dkar-po's summary of the tradition runs
Concerning the difference between siitra and mantra, Rgyal-dbang-rje
has said that the former is concerned largely with analysis and the
method of negation, while the latter is concerned with establishing
something positive. Accordingly the old translators of the laksana
(yana) said that zung-'jug is simply non-duality, as though (its substrate) did not exist; while when translating the mantras they generally indicated two existing things as (connected in) yuganaddha or
Padma Dkar-po himself, while broadly following Rgyal-dbang-rje's remarks, diverges somewhat from the view attributed to the earlier
translators. In his vajrayana works he uses yuganaddha pairs found
both in the Pahcakrama (see below) and in the tradition of Birupa
(snangstong &c, see above). The Pancakrama pairs are of both positive
and negative type. Now in his sutra (i.e. not vajrayana) works, Padma
Dkar-po typically considers zung-'jug pairs of the types which appear
as positive in the Pancakrama; these even play the main role.48 (Later
when we consider the ground/path/goal division in yuganaddha this
positive/negative distinction will come to seem less fundamental than
it does in these remarks. This is consistent with Padma Dkar-po's unwillingness to make a general distinction of Ita-ba (viewpoint) between
sutras and mantras.)
II.2: Yuganaddha as ground, path and goal
II.2.1: Preliminary survey
In Part I we sketched and quoted Padma Dkar-po's own summary of
the cig-car-ba's ground, path and goal in general, as he gives it in the
Gzhung-'grel. In the previous section, II. 1, we gave his view of
yuganaddha in general, using a variety of sources and arguments. Now
we bring the two together. As a preliminary we may note that in addition to yuganaddha, Padma Dkar-po has divided various other non-dual
conceptual categories49 by using the ground/path/goal scheme:
mahamudra (phyag-rgya chen-po),50 the radiant light (prabhdsvara, od-
gsal),51 and the middle (way, i.e. madhyamaka, dbu-ma).52 While a full
treatment of the ground/path/goal scheme would therefore include accounts of all these matters,53 yuganaddha is perhaps the most fundamental, and we confine attention to the others to the notes.
Table 1 contains a list of all the pairs of items which appear in the
verses of the 'bras-bu skye-ba'i rim-pa (goal) section of the Gzhung,
"integrated in yuganaddha. Without exception, all these verses are taken
from the Yuganaddhakrama (col. 2); and of course all receive commentary in the Gzhung-'grel (col. 3). With the letters N and P in column 4 I
have attempted to distinguish whether the verse asserts the absence of
some pair of things discursively discrminiated (negative type) or the
presence of some kind of unity of two things (positive type). This classification is based entirely on the words (e.g. ekatvam, one-ness) found
in the verses themselves and describing the relational aspect of each
separate case of yuganaddha; the Sanskrit words (as printed by Poussin)
are given in column 5. The classification seems at first glance to correspond to that sketched by Padma Dkar-po in the remarks quoted at the
end of the previous section. Unfortunately it seems impossible to introduce a single clear and uniform distinction on the basis of these analyses. Basically this is because two things differ (or are the same) only under a description. For instance, one is tempted to take the first pair,
samsara I nirvana, as negative because the idea is to give up this dualistic
conception (kalpana-dvaya-varjana, rtog-pa gnyis-po spangs-pa). But when
we have done this, we are of course not left with just nothing; we are
left with the world as it always was before the samsara/nirvana descrip-
ground section
YK 2
YK 3
YK 4
grahyal grahaka
YK 5
path section
YK 20
YK 12
YK 11
YK 18
YK 8
satnapattyd, samasatah
YK 7
goal section
YK 9
kalpana nasti
YK 10
kalpanayd viviktatvam
YK 6
samvrtil paramdrtha
YK 13
YK 17
YK 15
dvaya varjitam
YK 19
abhinnam svabhavatah
YK 16
tion was imposed upon it, and it seems difficult to see how this can be
generally and clearly distinguished from what happens in the "positive" cases. Similarly I have marked the fourth pair (grahyalgrahaka, object/subject) as positive because the verse describes them as non-different (abhinna, mi-phyed); but actually what is claimed to be non-different
is not these things but the particular state of mind (buddhi, bio) relating
(to) them in yuganaddha. In fact virtually any one of the pairs can be
seen (sometimes with a little effort) as positive if one concentrates on
the things referred to or mentioned or described, and negative if one
concentrates on the descriptions. This shift of perspective can easily be
experienced with the two examples just discussed. A case which one
might think difficult to see as either positive or negative in this sense is
that of the two satya; but there is no doubt that Padma Dkar-po thought
it could be seen either way, as I have shown at some length elsewhere.54 A case which commonsense takes as 'obviously' negative is
rdgalaraga: how can opposites refer to the same thing? But this is precisely the mistake (leading to the nonsense of "unity of opposites," &c).
Of course desire and its opposite are not the same. But the paramananda
in which (according to the verse) both are absent ("negative version")
is simply a state without either ("positive version"). There is nothing
to distinguish the two versions other than what part of the description
you concentrate on. (A lot of mysticism boils down to confusions of
this kind.55) A case which commonsense takes as "obviously" positive
is prajhdlupaya. Is it not obvious that in the buddha-mind (or whatever)
prajna and upaya are both present, in some sense? Of course; but (as in
the rdgalaraga case), in order to see this case negatively one need only
note that in general prajna and upaya are not the same at all and that by
insisting "prajna, prajhd" one only drives out updya, &c, so that the
right kind of prajhd can arise only when we withold the description, as
the Prajhaparamita sutras never weary of insisting; this is the negative
version.56 The conclusion is clear: if the positive/negative distinction
between the different pairs of things "integrated in yuganaddha has any
significance at all, each case must be seen as regarded from some quite
narrowly restricted point of view. If anything general can be said about
these points of view, now is the time; unfortunately I have very little.57
It was suggested earlier that there seems to be little hope of distinguishing different kinds of yuganaddha on the basis of the different
pairs (say, as characterized in the different verses). We can now see
more easily why this is, at least in individual cases. Consider again the
first verse (Table 1) on samsara I nirvana. First take it negatively: we are
required to distinguish some kind of samadhi solely on the basis of the
absence of the notions of samsara and nirvana (say as against another
kind individuated on the basis of the absence of the notions of rdga and
ardga). It seems difficult to see how to do this; one is almost tempted to
say: it is easy to see that it is impossible. Similarly on the positive interpretation of the samsara I nirvana verse: it is understood that "samsara"
and "nirvana" describe the same world (as it were) and on this basis
alone we are to distinguish some samadhi. It seems clearly impossible. It
is not only that the characterization is much too slender; it seems to be
of the wrong sort altogether. Similar remarks apply to taking the positive and negative interpretations together. Of course it is possible that
a distinction may be set up when one or both of these interpretations is
supplemented by some other information; but to claim this is to give
up any claim to characterize anything on the basis of the verses. In any
case, it seems pointless to pursue this possibility in relation to Padma
Dkar-po, since nothing in his comments on these verses suggests anything remotely resembling a verse-by-verse distinction of this kind.
It seems that the usefulness of the positive/negative distinction
among the pairs of "integrated items is likely to be fairly restricted. So
we can perhaps sympathize with Padma Dkar-po's evident lack of in-
terest in pursuing systematically this distinction which he inherited
from Rgyal-dbang-rje and from the earlier translators. If nevertheless
we accept the limitation, we may record, as in Table 1, what seems,
from the relevant point of view, to be the positive/negative character of
each pair. When we do this, we notice something fairly striking. The
ground section consists of pairs which seem easy to interpret either
way, and the given interpretations are mixed up. The pairs of the path
section are all positive except for rdgalaraga (I will try to explain the
anomaly later). The goal section is virtually all negative apart from the
important case of the two satya; but this is not surprising when we reflect that they generalize many of the items found in the path section.
Do the materials recorded in Table 1 suggest any other distinction
between the ground, path and goal divisions? Two further points seem
worth recording. First, the path verses do concern various parts of the
path2 which are to be "integrated; this is especially clear with the first
four. Second, there seems to be no such ground of distinction between
the verses of the ground and goal divisions; this is strikingly illustrated
by Padma Dkar-po's use of the yuganaddha of the two satya both as the
goal (in the Gzhung-'grel) and as the ground (in the Nges-don grub-pa'i
So far in this subsection we have been considering the possible divisions of the yuganaddha section into ground, path and goal, or rather
the possible sense behind the division given in the Gzhung-'grel. During this procedure we have taken for granted the position given to this
section in the Gzhung-'grel as a whole, viz. that of the goal (strictly:
graded emergence of the goal) of the pathj of the cig-car-ba, taken as a
whole. Now we may recall what was hinted at in the introduction to
Part I, viz. that the distinction or contrast suggested by these two remarks is not as clear as one might suppose. That is: the cig-car-ba has
his ground, path2 and goal (viz. dngos-po'i gnas-lugs, lam.2 and 'bras-bu
skye-ba'i rim-pa), and this goal section is further divided into ground,
path and goal; and it is not at all clear that this second ground, path
and goal are different from the first set, even though on a crude set-
theoretic or bibliographical basis it would seem that they must be. I
now want to set aside these set-theoretical and bibliographic considerations and review briefly how the subject matter of the goah ({bras-bu
skye-ba'i rim-pa) in fact fits in with that of some of the earlier sections of
the Gzhung-'grel.
When one examines the subject-matter at all carefully, it becomes
clear that the topic of zung-'jug can easily be fitted into several places in
the Gzhung-'grel besides the one it actually occupies. (The following observations can be clarified by using Appendix B.) In the ground section, zung-'jug could come under lus dngos-po'i gnas-lugs (especially in
the section called mngon-par byang-chub-par gyur-pa'i rim-pa, 30bl); or it
could come under sems dngos-po'i gnas-lugs, either in parallel with the
whole of the existing section, or as part of the subsection 'bras-bu phyag-
rgya chen-po mngon-du 'gyur-ba, 128b3 ff. In the path section, it could
come under virtually any of the eight main sections of the rdzogs-rim
(completion stage), viz. gtum-mo, las-kyi phyag-rgya &c. Guenther has
rightly observed that the radiant light is the climax of the path. That
doctrine, as we saw earlier,58 is itself divided into ground, path and
goal; and the goal section, as set out in the Gzhung-'grel,59 is particularly closely related to the topic of zung-'jug. Now it is quite clear why
Padma Dkar-po, in spite of all these considerations, put the goal section of the Gzhung-'grel where he did (viz. as the goal of the entire cig-
car-ba section) and gave it the title which he did give it ('bras-bu skye-ba'i
rim-pa). In both cases, the matter is decided already in the text upon
which he is commenting, viz. the bka yang-dag-pa'i tshad-ma.60 The point
of my observations is not that he ought to have put it somewhere else.
It is rather that the structure of his materials and argumentation would
have allowed him to have it in several different places; and this fact is
consistent with the view that the different gzhi are all aspects of the
same ground, the different 'bras-bu of the same goal, &c. It is much
more difficult to reconcile the fact in question with the view that the
different 'bras-bu sections are about different goals. And to this extent,
that fact supports the first of these views against the second.61 Some
further detail on this point will be presented below.
Since the 'bras-bu skye-ba'i rim-pa section of the Gzhung-'grel is quite
complex, it would be tedious to pursue explicitly the parallels between
it and the goal division of the radiant light section. But it may be worth
setting out the parallel between the latter and what may be regarded as
a summary of the former, viz. the zung-'jug section of the Rim-lngar
'khrid-pa62 After different introductory remarks, both passages give a
short account63 of the mode of rising out of the radiant light64 including
identical summaries of the "reversed"65 passage through the four
lights, supported by identical quotations from the Carydmeldpaka-
pradipa.66 The Rim-lngar 'khrid-pa account then concludes with a summary of how this process differs for the cig-car-ba, thod-rgal-ba and rim-
gyis-pa67. The account under the radiant light concludes68 with a more
detailed version of how this happens for the cig-car-ba, most of which
occurs again in different bits of the 'bras-bu skye-ba'i rim-pa section,
though not in detail; the reader can work this out for himself.69
Other than what has been said, I know of no general principles
which clearly or explicitly underlie the division of the goal section of
the Gzhung-'grel into its own ground, path and goal. It remains therefore to set down what is found in these three divisions. To do so will
be, in a sense, to confront what was done in Part I with its proposed
application to yuganaddha. The general discussion of the present subsection suggests that this application will take two different forms. In
the first form, yuganaddha either is, or is closely related to, the goal as
set out in general terms in Part I. That there is such an application
seems too obvious to need detailed argument, and I shall take what follows as an illustration of this application, regarded as already known
to exist. But this is of course not the application for which most of the
general discussion of this subsection has been the preparation. Rather,
it has been preparation for the application, to the separate ground,
path and goal divisions of the yuganaddha section, of the ground, path
and goal notions in general; and here it is far from obvious or explicit
that such an application is intended, apart from the rather indirect evidence already presented. It therefore seems best not to assume or to
take for granted that there is such an application, but simply to present
a selection of materials, in the hope that the suggestion may be found
illuminating by the reader (as I personally feel it is illuminating). Thus
Part I provides a set of notions which may well underlie the threefold
division of the yuganaddha section, and Parts I and II may seem mutually illumined by the idea of this application. Like so many other proposals made by scholars working in this and related fields, the value of
the suggestion is perhaps best assessed in terms of its capacity to bring
order to the complex materials under discussion.
The reader may find Table 1 a helpful summary of the main verses in
the goal section of the Gzhung, and the following subsections will contain brief accounts of Padma Dkar-po's comments on a selection of
these verses. Textual extracts in Tibetan, corresponding roughly to
these accounts, are given in Appendix C. GROUND-YUGANADDHA
Samsara/nirvana ('khor-ba, myang-'das)
After defining these terms, Padma Dkar-po says that when the discursiveness (spros-pa) which holds samsara and nirvana to be different
has been abandoned, then no matter how one properly {yang-dag-par)
analyses the samadhi of him who does this properly, since the artificial
imposition of duality is broken up these things become one, and so one
speaks of samsara and nirvana as "integrated.
The talk of two things becoming one is found already in the verse
(see Appendix C: dngos-po gcig gyur-pa, ekibhavah) and derives from a
confusion between two things and two descriptions of a thing. In other
contexts Padma Dkar-po advises abandoning the conflation of these
subsumed in the notion of a don spyi.
Klesa/bodhi (nyon-mongs, byang-chub)
When the essence of klesa is cognised, by that very fact it becomes
thoroughly illuminated and this is bodhi. When they are inseparable, like water and ice, klesa is made the path and this is called srid-
zhi zung-'jug.
Of course Padma Dkar-po's remark applies only to the cig-car-ba.70
Grahyal grahaka (gzung-ba, 'dzin-pa)71
To claim that there is a substance (dngos-po) involves the judgement
(rtog-pa) that it has qualities (rnam-pa-dang-bcas-pa); to claim that there is
no substance (dngos-med) is to base it on emptiness. These errors occur
through taking what is to be seen (blta-bya) as an object (gzung-ba) and
what sees it (Ita-byed) as a subject ('dzin-pa). When one rises beyond
these errors into the sphere of pratydtmddhigamajndha (so-so rang-rig-pa'i
ye-shes), the object (yul) which is appearance rises as gdangs from the
sphere of the void, and the owner of the object (yul-can) understands
rnatn-shes (vijhana) as rang-rig-pa'i ye-shes, and one speaks of snang-stong
zung-'jug and rnatn-shes dang ye-shes zung-'jug.72
Utpattikramalutpannakrama (bskyed-rim, rdzogs-rim)
When non-dual awareness occurs at the time of the utpattikrama,73 it
has the same pure taste as the awareness of the diety in its own unborn
incessancy (gdangs), and this making the two stages non-dual is called
their "integration.
Pindagrahal anubheda (ril-por 'dzin-pa, rjes-su gzhig-pa)
To destroy74 the illusory body75 all at once is pindagraha; to destroy
vessel and contents76 separately is anubheda. To enter the void, to stay
in the radiant light, and to rise from it as the pure and illusory (body) in
this way is all done to bring about the rising of the radiant light, and
since when this is done there is no movement away from the gshis,77
the sphere of the radiant light, there is snang-stong zung-'jug.
Svddhisthdna/prabhdsvara (bdag byin-brlabs, 'od-gsal)
When the depth of mind as a feature78 is not made manifest by
anything (else)79 it is abhisambodhi. What is fit to be the example for
all clarity is the svadhisthdna-diety. When that clarity generates it-
self unhindered from that depth, and that depth enlightens that
clarity, there is gsal-stong zung-'jug.80
Prajhopaya (thabs dang shes-rab)
Here upaya is the white dharma of charity &c. while prajhd stands
for prajnaparamitd. To abide in any great realization which acts by
grasping all the other pdramitas with an essenceless prajhd is prajho-
Sunyatdlkaruna (stong-nyid, snying-rje)
The object or the basis is the sixteen voidnesses, and the subject or
what is based on the former is the sixteen compassions. When one
knows the path on which subject and object or basis and what is
based on it are indivisible, one enters upon action out of desire
&c, and this is called siinyatakarunayuganaddha. By this desire one
cultivates the opposite of the path of action out of great desire, and
so it is called a buddhagocara. GOAL-YUGANADDHA
Sdsvatoccheda (rtag-pa dang chad-pa)
To view (things as) existent is eternalism; to view things as non-existent is nihilism. By whatever word(s) one gets rid of all attachment to the extremes (mtha) of existence, non-existence, and lack
of either of these, one consecrates the madhyamaka path in which
one does not fall into any extremes, and the wise know this as the
yuganaddhakrama itself. Any other explanation of "lack of extremes" which really contains attachment to some extreme represents a failure to understand the character of yuganaddha.
Samvrti/paramdrtha (kun-rdzob, don-dam)
This case is discussed at great length in my paper (1983a).
Supta/prabuddha (gnyid, gnyid sad-pa)
Here "sleep" (gnyid) means the stage of un-knowing, while to
wake up (gnyid sad-pa) means a one-sided knowing. The state of
neither sleep nor waking is that where one has unequivocally and
permanently risen up (sangs-pa) from this sleep of un-knowing. It
is like a lotus flower upon which a bird sits (?) as if upon a broad
(rgyas-pa) field, and so an awareness which cognises everything
cognisable is called broad. When these two are "integrated (zung-
du 'jug-pa) there is buddhahood (sangs-rygas).
Kdrya/kdrana (bya-ba, byed-pa)
Since this yuganaddha82 is the essence (bdag-nyid) of all the good of
others, action and agent are inseparable.
II.3 Yuganaddha revisited
One may well feel that the account of yuganaddha given so far is lacking in something. Part I dealt with ground, path and goal in general,
illustrating the great importance, for Padma-Dkar-po, of the notion of
dngos-po'i gnas-lugs.83 So far, Part II has presented a variety of facts
about yuganaddha and some of Padma Dkar-po's explanations of the
Gzhung verses on it, without, one might feel, really saying what it is;
and in another sense, the notions of ground, path and goal have perhaps been plausibly applied to something, but one is not quite sure
what. (This is at any rate my personal reaction to the account in the
Really, we still need a general account of yuganaddha. Now, in addition to three accounts divided into ground, path and goal in the way
we have seen,84 Padma Dkar-po has given several general accounts,85
but they are too brief and cryptic for a reader who does not already
have some partial grasp of his idea of yuganaddha, such as is provided
by the materials already surveyed; these illustrate the general accounts, which in their turn enable us to see what lies behind the more
specific ones. In this section I will extract from the general accounts, especially the one in the Phyag-chen gan-mdzod, a synthetic general account which I hope will fill the gap.86 Padma Dkar-po is a quite ahisto-
ricist writer, in that he hardly ever defines his position on anything by
refuting the views of earlier writers (as, say, Tsong-kha-pa so often
does); but on difficult matters like this one he has sometimes clarified
his views by contrasting them with those of others. Later I will show
him doing this; to start off with, it seems a good idea to mimic the procedure by saying a few things about what his notion of yuganaddha is
not, especially in relation to some of the proposed translations mentioned in section ILL
There seems to be nothing mystical87 about yuganaddha, and the materials surveyed give no reason to translate it by "the union of
opposites" (Guenther88). For instance, rdgdrdgayuganaddha, far from being the union of rdga and its opposite araga, is something lacking in
both. Nor is yuganaddha a logical relationship such as "coincidence"
(Guenther89). (It is almost too obvious to be worth saying that it is not a
matter of logic that action and agent are "integrated in yuganaddha. And
even if "coincidence" is not taken as a logical term, it is really obvious
that it cannot apply, other than metaphorically,90 to cases such as rdgal
araga and pudgalanairdtmydldharmanairdtmya. Similarly for "identity".)
Again, yuganaddha is supposed to transcend various specific dualities,
and perhaps all dualities whatever, but that does not mean that it is
something transcendental in the sense of being beyond experience.91
In particular, it does not seem at all like Kant's "intelligible intuition".92
It may be worth examining a little more closely the claim that
yuganaddha is part of experience. First, it is not claimed that it is a part
of ordinary everyday experience (for no samadhi is). Second, the claim
is not that it constitutes some kind of (necessary? Kantian?) ground of
experience.93 Thirdly, the claim is not that it is some fundamental kind
of experience from which all others are derivative in some sense.94 The
status of yuganaddha as experience begins with the observation, much
too neglected by everybody writing on the subject, that it is a samadhi
(II. 1). No claim is made that this samadhi is totally without concepts; but
what is important to see now is that even a samadhi which was totally
lacking in concepts would not thereby be removed from experience,
any more than a similar dream would be. Such a dream might perhaps
be seen by us as an uninteresting part of experience, because not connected with the rest in the right ways. But Buddhists do not see this
lack of connection as making such experiences uninteresting, indeed
the very opposite (hence the interest in dream as one of the six topics of
Naropa and in other ways). The claim, then, is that Padma Dkar-po's
notion of yuganaddha is consistent with our interpreting it as a kind of
experience, a samadhi experienced perhaps under unusual conditions
of meditation, abhiseka &c. or by unusually gifted people, but in no way
radically different from other experiences enjoyed by perfectly ordinary
Guenther has frequently used the word "unity" for yuganaddha (see
note 3); the points made in the last few paragraphs are sufficient reason for rejecting it as a quasi-literal translation, but I find it not without
appeal as a metaphor (cf. note 90). What is this appeal based on? Attempts to expand it into such phrases as "unity of experience" and
"experience of unity" seem to lead to nothing.96 I think some kind of
case might be made out for it if it were made clear that it is a metaphor;
Guenther never does this. But even then, its appeal seems to be mainly
a matter of its vague suggestion of a combination of mysticism and
logic; and since we now have good reasons for rejecting each of these
as irrelevant, we may feel able to reject the combination.
Earlier we noted that Wayman's translations of yuganaddha97 are all
based on the idea of combining two things together. Though to some
degree this is intended to reflect the view of Tsong-kha-pa,98 it does not
reflect any difference in the Tibetan phrases zung-du 'jug-pa and zung-
du 'brel-ba used to translate yuganaddha. Phrases like "pair combined"
and "pair united" might qualify as literal translations of zung-du 'brel-ba
but hardly of zung-du 'jug-pa. Neither the Sanskrit yuganaddha nor the
Tibetan zung-du 'jug-pa is a colloquial expression with an obvious natural sense, and the analysis into yuga + naddha and into zung-du + 'jug-
pa yields nothing literal because in both cases the second component
has no "natural" sense related to these translations (viz. "pair combined," &c). (Just this fact is reflected in Padma Dkar-po's own ac-
count of the analysis, given in II. 1 above.) So really the suggestion that
Wayman's translations are literal will not stand up to serious analysis
and can itself be no more than metaphorical. Let us ignore for the moment the possibility that these translations are themselves metaphorical. If we do so, their value must be seen as resting not on any analysis
of the words yuganaddha and zung-'jug, but rather on the degree to
which these words as wholes are as a matter of fact used in the sense of
"pair combined", &c. Now on the negative interpretation of the
verses, such a claim would be simply ridiculous; it can rest only on the
positive interpretation; and this involves us with all the difficulties that
arise from the easy shift between the two styles of interpretation, to
which attention was drawn in section II.2.1. But with this (serious) reservation, the claim is not without merit; it works moderately well (as
far as the verses are concerned) for most of the path section (see Table
1) and also for the important case of the two satya.99 To say this is not to
say that it gives the complete sense of the word yuganaddha as used
here (for instance it omits the samadhi element) but it does at least give a
central element. It turns out, however, that this element is more central
for Tsong-kha-pa than it is for Padma Dkar-po.
So far we have not exploited the striking fact (see Table 1) that most
of the important Gzhung verses on yuganaddha come from the Pahcakrama. Now in the Gzhung-'grel and his other works on bsre-'pho, Padma
Dkar-po makes much use of notions from the Guhyasmaja cycle, especially in relation to the structuring of the various practices. He also
makes much of the idea that the five krama, the four mudra, and the
sadahgayoga give rise to alternative structures for what are essentially
the same repertoires of sampannakrama practices in the Guhyasamaja,
Hevajra, Cakrasamvara, Kalacakra and other tantras. So it is perhaps surprising that he interprets the yuganaddha verses not in the style of the
father class of tantras (of which the Guhyasamaja is the chief) but in the
style of the mother class. Before going on to the details of this, it is important to be clear that the difference has nothing to do with
yuganaddha being samadhi or not; for instance the description of it as
samadhi which opens the 'bras-bu skye-ba'i rim-pa section of the Gzhung-
'grel is in a verse from the Vajramald, a father-tantra (indeed, an akhya-
tantra of Guhyasamaja). Even so, says Padma Dkar-po, the full sense of
yuganaddha is not revealed in the father-tantras. More generally, the
Gzhung receives the name Dakini-upadesa (Mkha-'gro-ma'i man-ngag)
because it is the dakinis who, in the mother-tantra, establish fully yet
briefly what is hidden in the father-tantras.100 This remark is supported
by a verse from the Ddkdrnavatantra (belonging to the mother class) running, "Confidence in the father-tantras comes from understanding the
mother-tantras." Accordingly Padma Dkar-po treats the yuganaddha
verses as mother-tantra, whereas Tsong-kha-pa, taking them from the
Pahcakrama, of course treats them as father-tantra.
Thus we may expect a difference between the two authors in their
styles of interpretation. In order to understand what to expect, we
need to know that the differences do not stem merely from general disagreement about the distinction between the father and mother tantras;
this point needs careful consideration because prima facie their views do
seem to be rather different and because Tsong-kha-pa's views have not
been well-handled in the literature.
Bu-ston and Padma Dkar-po concentrate mainly on listing the various distinctions that are found in the tantras themselves. Bu-ston
thought that there are three kinds of Anuttarayogatantras (father,
mother, and non-dual), a view he inherited from the Sa-skya school.
Padma Dkar-po follows Tsong-kha-pa in holding that there are only
two (father and mother). Tsong-kha-pa's attack on the earlier authors
is directed mainly at this view of three classes, but at the same time he
sought to replace the rather unclear collection of distinctions given by
Bu-ston with a single general rule which would apply even to the
doubtful cases101 and would provide a criterion for the clearer ones.
Tsong-kha-pa sought this rule at the point of greatest interest, which
also happens to be the point which concerns us, viz. the structure of
the sampannakrama. (He is amusingly sarcastic on the uselessness of a
rule based, say, on differences in the number of heads and arms of the
dieties.) He states the rule and summarizes the argument behind it as
"The explanation of the well-established opinion on the distinction of the
anuttarayogatantras as prajna and upaya, as given in the Vajrapanjaratantra
"How indeed should one distinguish them as father and mother
tantras? Though it is generally held that they are to be distinguished
through their utpattikrama cycle, really the difference is mainly one of
the sampannakrama. If we take bliss and emptiness (bde-stong) as means
and insight (thabs-shes, upaya and prajna) respectively, individual
tantras cannot be distinguished; all must be called non-dual as above.
In the present context this distinction cannot be maintained, either
generally or in detail. For in what are accepted by everybody as
mother-tantras, such as Hevajra, mainly bliss is taught, but not in the
Guhyasamaja &c, so there would be the absurdity that the Hevajra
would be a father- and the Guhyasamaja a mother-tantra.
"Thus as far as the distinction of updya-tantra and prajnd-tantra by
means of the sampannakrama goes, prajna must mean paramartha-
mahdsukha-jnand, while upaya must mean samvrti-mdyddeha. Of these
points, the first is about the mother-tantras and is found in the 13th.
chapter of the Vajrapanjara:
'Prajndparamitd as means is called yogini;
Entry into tattva by union (with) mahamudra
Is called yoginttantra.' "
Tsong-kha-pa then explains in more detail how this verse is to be understood in terms of the descent and ascent of bodhicitta, &c. This brilliant and thoroughly convincing piece of analysis is ruined by Mkhas-
grub-rje103 by the omission of most of the important points and the general vagueness and equivocation in what remains. (Given this unpromising material, it is not surprising that Lessing and Wayman's translation is incomprehensible; but they make matters worse by their indifference to the meanings of the technical terms.) Now Tsong-kha-pa's
analysis is brilliant; but it leads straight to the conclusion desired by
Padma Dkar-po, namely that the rule is intelligible only if the terms
prajhd and upaya are taken in the mother-tantra sense, standing respectively for bliss and emptiness. For Padma Dkar-po this last point is very
natural, since he regards the basic emptiness not as svabhavasunyata
but as sarvakdravaropetasunyata which is samvrti-satya (not paramdrtha as
with the Jo-nang-pas) and is means (updya) and which describes the diety,
just as in the first line of the verse. Paramdrtha here is great bliss. Padma
Dkar-po follows Bu-ston in his statement of the sampannakrama distinc-
"The mother-tantras show entry into tattva by the union (sbyor-bas) of
prajndpdramita-updya in mahamudra. The father-tantras mainly show the
svddhisthdnakrama and abhisambodhikrama."
The first of these remarks reflects the language of the Vajrapahjara verse
(which they and everybody else quote) while the second comes from a
verse from the Dakdrnavatantra which is also quoted by everybody. Of
course Bu-ston and Padma Dkar-po list many other modes of distinction which are not accepted by Tsong-kha-pa, but these are irrelevant
here; it is obvious from Padma Dkar-po's discussion in the Phyag-chen
gan-mdzod that it was this particular point about the sampannakrama
that he had in mind when describing Tsong-kha-pa's treatment as pha'i
lugs, and not the others. Padma Dkar-po also appears to follow Bu-ston
in describing the distinctions between father- and mother-tantras as
provisional (drang-don, neyartha) whereas ultimately or really (nges-don-
du) all the anuttarayogatantras are non-dual as between prajhd and
updya}05 Tsong-kha-pa does not make this point. But this appearance
of difference is an illusion, since Padma Dkar-po's use of these terms
differs from that of Bu-ston and Tsong-kha-pa somewhat.106 All Padma
Dkar-po means is that the more advanced person will understand the
ultimate purpose (mthar-thug-gi don) without all this explanation, while
the somewhat less advanced person needs to supplement the father-
tantra with the mother-tantra, as explained above. (The least advanced
person cannot go further than what the father-tantras strictly and literally say, and part of Padma Dkar-po's complaint is that Tsong-kha-pa
does not appear to offer more than this.)107
Roughly speaking, then, the two authors agreed that the father-
tantras concentrate more on the separate stages of the sampannakrama
practices, while the mother-tantras tend to prefer a more unified or
holistic approach. Hence the phrase "fully yet briefly" (tshig nyung-
ngus zhib-mor) mentioned a few paragraphs ago in connection with the
phrase Ddkini-upadesa. And now, taking the last few paragraphs together, we can see part of what, in Padma Dkar-po's view, lies behind
his differences with Tsong-kha-pa over yuganaddha. For Padma Dkar-
po, the notion of yuganaddha is to be understood in a unified or holistic
way, the analysis into pairs and their modes of combination being secondary and relatively unimportant. For Tsong-kha-pa the notion is one
of combination, the constituents being as it were primary and their
combination, as a product of those constituents, secondary. More generally, for Padma Dkar-po what is primary is the buddha-knowledge
(sangs-rgyas-kyi ye-shes) or sahajajhdna (lhan-cig skyes-pa'i ye-shes). This
corresponds roughly to the anuloma (lugs 'byung-ba) approach. For
Tsong-kha-pa (as Padma Dkar-po sees him) what is primary are the
components, say the two satya or the two krama,108 and it is only when
these are known separately (cf. YK13) that they can then be combined
in yuganaddha. This corresponds roughly to the pratiloma (lugs Idog-pa)
Now we can see why the translations of yuganaddha proposed by
Wayman are so much less unsatisfactory for Tsong-kha-pa than for
Padma Dkar-po. They reflect precisely the emphasis on analysis as
against a holistic view, on the pratiloma as against the anuloma, on the
father against the mother tantra, of which it might be excessive to say
that they run right through Tsong-kha-pa's way of thinking, but which
certainly seem to inform his style of exposition.109
Two aspects of Padma Dkar-po's view of yuganaddha should be seen
in relation to his discussion of Tsong-kha-pa. First, there is the primacy
of the combination over its terms (or rather the claim that yuganaddha is
not just a combination of terms). Second, there are proper and improper ways of combining the terms. Really these two cannot be
cleanly separated. When one still thinks mainly in terms of combining
two things, there is saiksayuganaddha. When the question of combining
two terms no longer arises because they are seen and known as aspects
of a single situation, there is aiaiksayuganaddha. But these two also cannot be cleanly separated, for two reasons. First, as already mentioned,
saiksayuganaddha already is the real thing, as far as it goes; and second, if the mode of combination is wrong we do not even have
saiksayuganaddha, we just have a mistake. To put it another way,
saiksa is asaiksa, only seen in the father-tantra way, while asaiksa is
saiksa, seen in the mother-tantra way. As so often, the mistake is to
treat a relational difference or a difference in points of view as a difference of two distinct items. When one reads Padma Dkar-po's discus-
sion of Tsong-kha-pa, one tends to be more impressed by his points
about the modes of combination because there are more of them and
they are more detailed and seem easier to grasp; but the point about
the father-tantra is there explicitly110 and indeed without it the details
do not make sense.
Let us consider some of these details for the case of svddhisthdnal
prabhasvara. Padma Dkar-po's own view, given in the Gzhung-'grel, is
found above, under path-yuganaddha. He uses the description in terms
of depth and clarity in a similar way in at least five other places111
(twice quoting Naropa, in fact from the Sekoddesatika, and once in the
Vimalaprabhd; apparently this line of thought comes from the Kdlacakra-
tantra). In all these materials Padma Dkar-po is unrelenting in his insistence that the two elements cannot be separated.112 But he says that
Tsong-kha-pa, after quoting the verse, says that the illusory body
which is samvrti and svddhisthdna, and the radiant light which is
paramdrtha are two halves, and when these are inseparably merged
('dus-pa), that level is called zung-'jug.113 Padma Dkar-po's own phrase
for what happens is "not two in (its) nature" (gnyis-su med-pa'i rang-
bzhin111). These examples illustrate what I called the difference in point
of view, the difference between the father- and mother-tantra styles of
Now for some of the specific differences on the mode of combination. Tsong-kha-pa continues by saying that since zung-'jug cannot be
understood unless the two halves have been understood, he will go on
to explain them (I omit this). Now in the abhisambodhikrama there is no
illusory body, while in the svddhisthdnakrama there is no radiant light.115
It is because these appear alternately that there is no yuganaddha116 (and
because of fizz's it would be irrelevant to take the previous remark as being about the practices rather than what appears in them). But when
the forward (anuloma) and backward (pratiloma) processes of going in
and out of the radiant light go on at the same time,117 so that the essence (ngo-bo; rupal) of the two is (are?) inseparable118 there is
We know by now that Padma Dkar-po will see this whole approach
as misconceived: far from zung-'jug not being understood until the two
halves are understood, they cannot be understood except in relation to
it. But now I want to concentrate on what he has to say about the mode
of combination. Zung-'jug, Padma Dkar-po says,119 is not simply the
merging, according to the method of the father-tantra, of two different
items like the two horns of an ox. In any case, these two items are too
dissimilar to merge in the way suggested by the analogy. They are
even mutually repugnant; for since samdhita (mnyam-bzhag) is, on this
view, without appearances, there can be no samvrti in it,120 and vice-
versa. Similarly, if asaiksayuganaddha arose through complete purification by entering the radiant light by means of pindagraha and anubheda,
then samvrti-satya, being impure, could not belong to it (and there is a
contradiction).121 So for Padma Dkar-po, the mode of combination (in
this class of cases) is not the merging of two quite different (even repugnant) items; the two items have to suit one another, as it were. In
this case he ensures this by demanding that samvrti-satya (of which the
svddhisthanakrama is a particular case) contains purified appearances,122
rather than impure ones, as with Tsong-kha-pa. (In the language of
Part I, samvrti-satya is part of dngos-po'i gnas-lugs). Again, for Padma
Dkar-po, paramdrthasatya contains purified appearances123 rather than
being merely empty of content. We see here Padma Dkar-po combining
something like the three levels of satya used by Bhavaviveka124 (i.e.
samvrti and parydya- and nisparydya-paramarthd) with the three levels
used by Candraklrti (samvrti-matra, and samvrti- and paramdrthasatya).125
The point that the "integrated items have to suit one another in the
first place illustrates again the overriding importance for Padma Dkar-
po of the combination over its constituents. Also the notion of
yuganaddha does not seem to contain within itself any single or unified
notion of a mode of combination; the different pairs are combined together in different ways (to give the right kind of result, as it were).
(This point tends to be masked by the concentration of attention on the
two satya and their instances.) Each of these points shows separately
that for Padma Dkar-po there is no such thing as yuganaddha considered mainly as a relational concept (in abstraction from its other elements). The various pairs, considered merely as related, do not in their
being related have anything more in common than their merely being related; and of course this is not enough to form even the basis of the notion of yuganaddha as the goal of Vajrayana. This point is apt to be obscured by the importance of the constituents, in the cases usually discussed (samsara/nirvana, the two satya, the illusory body and the radiant light). One goes along with the vague idea, encouraged by the current translations and perhaps by Tsong-kha-pa's treatment, that these
things only have to be seen as related in the right way, and the task is
completed. But a glance at some of the other cases (rdgalaraga,
pindagrahalanubheda &c.) is enough to see that this is an illusion, that
the different "relationships" have nothing in common and that the
nub of the idea must be somewhere quite different. I suggested that
Tsong-kha-pa's treatment encourages the illusion, but I am far from
certain that he himself succumbed to it; the encouragement comes
from his reliance on the father-tantras. This is of course Padma Dkar-
po's very first point (cf. note 119); it should not be allowed to detract
from the sharpness of his criticisms about the mode of combination.
Perhaps the time has come to say briefly why I prefer Seyfort
Ruegg's choice of "integration" over all the other suggested translations of yuganaddha, zung-'jug. "Integration" seems to cover well the
"merging" ('dus-pa) attributed to Tsong-kha-pa; and for Padma Dkar-
po, "Integration-samad/zz" seems to be the right kind of metaphor to
convey the idea behind zung-'jug-gi ting-nge-'dzin, as he used this
phrase. Integration is a kind of action (in the broadest possible sense);
"integration-samad/zz" suggests a samadhi in which various different
kinds of viewpoint are, well, integrated; and I think this metaphor is
about the best we can expect. (Because it z's a metaphor, and not one
naturally suggested by our own cultural context, I like to keep the asterisk: "integration.) This paper has not dealt with the Sa-skya view of
zung-'jug; but I suspect it will be covered by the proposal, as being
somewhere in between the two views just mentioned.126
Earlier I mentioned that "pair-combined" &c. might perhaps also be
intended to be read as metaphors. Well, perhaps they might. But I
think it is clear by now that the suggestive power of this and similar
phrases will not accomplish anything presently relevant which is not
accomplished by their literal meaning.127
At the end of section II.2.1, I suggested that the application to the
notion of yuganaddha of the ideas of Part I might be expected to take
two forms. First, yuganaddha is itself the goal; and second, being divided into ground, path and goal, it might be expected to illustrate this
division as discussed in general in Part I.
There are plenty of illustrations of the first form of application. One
example: the radiant light is the climax of the path, and its path-secfz'on
describes the meditation techniques of pindagraha and anubheda (ril-
'dzin and rjes-gzhig). In the path-division of yuganaddha we again find
these two techniques, about which the verse says that when they, together with the three phases of entering, staying in and rising from the
radiant light, become identical, there is yuganaddha; and we saw Padma
Dkar-po explain this by saying that they become identical inasmuch as
there is no movement away from the radiant light. Thus the specific
techniques described as part of the radiant light doctrine have as their
specific goal or culmination what is described under the same heading
in yuganaddha. Similar remarks can be made about other aspects of
yuganaddha, as correlated with the culmination or the goal of other
parts of the pathi.
The example mentioned also illustrates the second form of application. For pindagraha I anubheda is (part of) yuganaddha as the path, and
this is the culmination of what is described under the radiant light as
the path (in both senses of this phrase). Similarly, when all the other
things which fall under the two satya have been "integrated, nothing
more remains to be done and the goal has been attained, and so the
yuganaddha of the two satya as the goal in general coincides with the goal
of the various path-parts of yuganaddha (falling themselves, of course,
under the satya). It is easy to give more illustrations of this theme.
These two different ways of looking at pindagraha/anubheda can be
summarized in a rather slogan-like way thus: The culmination of the
path is the path-aspect of the culmination. And along these lines we
can see the two applications to yuganaddha of the ground/path/goal distinction as reflecting a single vision in two different ways.
When stripped of the accretions imposed upon them by Western authors, the doctrines discussed in this paper emerge as rational (though
not rationalistic) and sensible (though not commonsensical). I think
this becomes clearer when we look at the sources of the remaining obscurities. One very fundamental thing not explained here is the notion
of a non-discursive cognition or knowledge (nirvikalpajhana, mam-par
mi-rtog-pa'i ye-shes), upon which depend Padma Dkar-po's conception
of paramarthasatya and of dngos-po'i gnas-lugs and also that part of the
radiant light doctrine needed for our present purposes.128 Another
thing which remains obscure is how it comes about that a person who
has attained yuganaddha then goes on to all the siddhi of mahamudra, including all the buddha's powers and qualities, especially the ethically
important power of effortless (Ihun-grub) action &c. However only the
connection of yuganaddha with buddhahood depends on these siddhi,
not the notion of yuganaddha itself. In my view the complete notion of
buddahood does have something transcendental about it (cf. tathagata)
but whatever it is, it is to be explained in terms of yuganaddha and not
vice-versa, at least as far as Padma Dkar-po is concerned. More important, both nirvikalpajhana and the buddha's powers and qualities form
an integral part of the Mahayana as a whole; neither they nor any obscurity or mysteriousness in them have anything specific to do with Tibetan thought or with the tantras. Apart from whatever support he
draws from these two general features of the Mahayana, then, Padma
Dkar-po's accounts of tantra as continuity, of the three tantras of
ground, path and goal, and of yuganaddha, seem rational enough. And
their difficulties, such as they are, can be considered independently of
the obscurities, much greater in my view, of the six topics of Naropa,
the cosmological aspects of the doctrine,129 the (perhaps psychological)
doctrine of manomayakdya, and many other matters.
Because of these further obscurities, it may be too early to suggest
that Tibetan thought in general and Vajrayana thought in general may
not have quite that special degree of obscurity and mysteriousness
which they often seem to have in Western accounts. Be that as it may,
here I have shown how to understand some of the leading conceptions
of one of Tibet's most important Vajrayana writers as little more obscure
or mysterious than certain well-known features of the whole
*»*    ***   ***
[For Appendix A, see Part I]
***   ***   ***
gang 'jug-pa rten-gyi gang-zag, 8a4::5a5
gang-du 'jug-pa'i lam-la gnyis
cig-car-du 'jug-pa'i lam-la gnyis
bsdus-pa'i don, 10b5::6a2 (given in Appendix A)
mngon-par rtogs-pa gtan-la dbab-pa-la gsum
dngos-po'i gnas-lugs-la gnyis
lus dngos-po'i gnas-lugs-la gnyis
rags-rim-nas ngos-bzung*, 14bl.:7b5
phra-rim gtan-la dbab-pa-la bzhi
lus ji-ltar grub-pa'i tshul, 24bl::c.8a3
rtsa dngos-po'i gnas-lugs, 45a6.:10bl
rlung dngos-po'i gnas-lugs, 74a3:.Tlb5
byang-sems dngos-po'i gnas-lugs, 97b4:.T3a4
sems dngos-po'i gnas-lugs-la gsum
gzhi phyag-rgya chen-po gtan-la dbab, 115a2::14bl
lam phyag-rgya chen-po nyams-su blang, 120a4:.T5a3
'bras-bu phyag-rgya chen-po mngon-du gyur-ba, 128b3:.T5b3
lam-ia gnyis
smin-byed-kyi dbang, 129b6:.T6a6
grol-bar byed-pa'i lam-la gnyis
rim gnyis bsgom-tshul spyir-bshad*. 142a6.:17a6
mngon-par rtogs-pa so-sor gtan-ia dbab-pa-la gnyis
bskyed-rim, 147b4:18a4
mdor-bstan*, 164a2:.T9a3
rgyas-bshad-la brgyad
gtum-mo lam-gyi gzhi-ma, 179b6::20a6
las-rgya lam-gyi 'bogs-don, 278b2.:x
sgyu-lus lam-gyi srog-zhing, 310a6::77a5
rmi-lam lam-gyi drod-tshad, 324bl::87b4
'od-gsal iam-gyi snying-po, 331al::94b6
bar-do lam-gyi blo-chod, 345b6:.T08a4
'pho-ba lam-gyi bsun-ma, 360b6.:119b3
grong-'jug lam-gyi 'thud-ma, 368a4::133b3
bras-bu skye-ba'i rim-pa-la gsum
Ita-ba mam-dag zung-'jug-gi ting-nge-'dzin bslab-tshul-la gsum
gzhi zung-'jug rtogs-tshul, 370a5:.T36a4
lam zung-'jug bsgom-tshul, 372a3.:136b3
'bras-bu zung-'jug 'char-tshui, 374a3::137al
sgom-pa mam-dag zung-'jug-gi ting-nge-'dzin bslab-tshul, x:.T37b4
spyod-pa mam-dag zung-'jug-gi ting-nge-'dzin bslab-tshul-ia gnyis
chags-can-gyi spyod-pa, x:.T38bl
chags-bral-gyi spyod-pa, x.:139b5
rim-gyis 'jug-pa'i lam, 378b3::x
References: 8a4::5a5 means gzhung-'grel 8a4, khrid-yig 5a5; x = absent
(The sections marked * contain especially useful summaries of the
reasons why the sections following them are organized as shown here.)
The verses are headed by the Sanskrit words used in Table 1 and in
sec. 11.2.2. Each heading is followed by the verse itself (as Padma Dkar-
po quotes it from the Gzhung), and then, following the ligature j , an
extract from Padma Dkar-po's commentary (cf. II.2.2). I
Bibliographic information may be found in Table 1.
Samsara /nirvana
^■(^•q^E^| al^'^"^^"^^	
^W\«T«1| ]0|c:-Sjcj|gzj|^g^-g^^]| i^-Sj^	
<Y<^'^'ljl ... K^'^'^'f'^	
Grahya/grahaka [and sakara/nirakara]
^rq^-q^tyifq^q ^•q-^-q-^-|L:"
q^^qc^-^q^^^-q^l 15'^r^-qS^'
q-^r^-q^r jlj^^-^|-q^n]^^n]-^| ijar
C  C C C Cv
q|^q^^^q-^-q]50]-^^| pi^'UK-gapr
N3 <^ /
^•q^^l^^q-l^-^zji^r;-!  |Q^n|^r;^	
aj'^-q^aj-^l^i    J^-q§^^*r^-q*r
zj|^-q-|^]-^zi|    |<|aj-^-|c:-q^Q^zi|-q^r;
^•^•^•q^^-^-^| p^g^^c:'
jg^^^^-^Q^'j^'^^^Tg^-^^'^! ^
qs^^*rcrj*cq|     |^a]-u\g^^-q-z3|r;^	
v     ev      ev      -^s.      ev ev
<^<*f]*rq-5^|  f |^'q^-q^0|^|  £|^q;\"
^ -SO ^
^<*rq^| | |^a]"q^^,q^q^q,^-q-n]§n]--
qo^«r$|q«i| ^■^'^""'■^T^'V	
ADK: Abhidharmakosa
BCA: Bodhicaryavatara
BCAP: Panjika on BCA by Prajnakaramati, ed. Vaidya
GST: Guhyasamajatantra, ed. Bagchi
HT: Hevajra-tantra, ed. and trans. Snellgrove
MMV: Madhyamakavatara
MSL: Mahayanasutralahkara
PK: Pancakrama, ed. Poussin
PPD: Pradipoddyotana, sde-dge
Siks.: Siksasamuccaya
(For Tilopa's Aha-pramdna samyag-ndma ddkini-upadesa, see Gzhung
under Padma Dkar-po in the Tibetan section.)
(The works are grouped by author. The full title, where given, is prefaced by a sobriquet or short title in italics, by which the work is identified in the footnotes. The different Rgyud-sde spyi'i rnam-gzhag
are also identified there by author.)
Dol-po Shes-rab Rgyal-mtshan
Ri-chos nges-don rgya-mtsho
Rgyud bla-ma'i grel-pa Legs-bshad nyi-ma'i od-zer
Gsang-'dus bshad-thabs: Gsang-ba 'dus-pa'i rgyud-'grel-gyi
bshad-thabs-kyi yan-lag gsang-ba'i sgo-'byed, gsung-
'bum vol. 8
Pradipoddyotana-tikd, ibid.
Rim-lnga'i dmar-khrid, vol. 9
Rgyud-sde spyi'i mam-bzhag rgyud-sde Gsang-ba gsal-byed,
Tsong-kha-pa (from the 18-vol. bka-'bum)
Sngags-rim chen-mo, vol. ga
Sgron-gsal mchan-'grel, vol. nga
Sgron-gsal dka-gnas-kyi Mtha'-gcod rin-chen myu-gu, vol. ca
Rim-lnga rab-gsal sgron-me, vol. ja
Rgyud-sde spyi'i rnam-gzhag, Tibetan text in LW
Mi-bskyod Rdo-rje
Dwags-brgyud grub-pa'i shing-rta, dbu-ma-la 'jug-pa'i rnam-
bshad dpal-ldan Dus-gsum mkhyen-pa'i zhal-lung
Sgam-po-pa Bkra-shis Rnam-rgyal
Nor-bu'i od-zer: Gsang-sngags rdo-rje theg-pa'i spyi-don mdor-
Nyi-ma'i od-zer: Dpal Kye'i rdo-rje rgyud-kyi rgyal-po legs-par
Padma Dkar-po (except where otherwise stated references are to
the 24-vol. reproduction of the Gnam-'brug Par-ma)
Tshad-ma 'Jam-pa'i dgongs-rgyan, vol. 1
Tshad-ma rigs-pa'i snying-po, vol. 1
Gsan-yig: Bka'-brgyud-kyi bka-'bum gsil-bu-rnams-kyi gsan-
yig, vol. 4
Mngon-mdzod: Chos-mngon mdzod-kyi bshad-pa 'grel-pa'i
lugs, vol. 8
Nges-don grub-pa'i shing-rta: Dbu-ma'i gzhung-lugs gsum gsal-
bar byed-pa, vol. 8
Rgyud-sde spyi'i mam-bzhag Mkhas-pa'i yid-'phrog, vol. 11
Dus-'khor gsang-mdzod: Mchog-gi dang-po'i sangs-rgyas mam-
par phye-ba gsang-ba thams-cad bshad-pa'i mdzod, vol.
'Khor-lo sdom-pa'i rnam-bshad: Dpal 'Khor-lo sdom-pa'i rgyud-
kyi mam-par bshad-pa mkha-'gro dga-ba rgyud-sde'i
snying-po, vol. 14
Yid-'phrog: Dpal Kye rdo-rje'i spyi-don grub-pa'i yid-'phrog,
vol. 15
Gsang-'dus rgyan: Gsang-ba 'dus-pa'i rgyan zhes-bya-ba Mar-
lugs thun-mong ma-yin-pa'i bshad-pa, vol. 16
Phyag-chen gan-mdzod: Phyag-rgya chen-po man-ngag-gi
bshad-sbyar rgyal-ba'i gan-mdzod, vol. 21
Rnal-'byor bzhi'i mdzub-tshugs: Rnal-'byor bzhi'i bshad-pa don-
dam mdzub-tshugs-su bstan-pa, vol. 21
Gzhung: Bka yang-dag-pa'i tshad-ma mkha-'gro-ma'i
man-ngag by Tilopa (cf. Indian section); references
are to the version in Rare Bka'-brgyud-pa texts from
Himachal Pradesh. (Also in Gdams-ngag mdzod, vol. 7.)
Gzhung-'grel: Jo-bo Naropa'i khyad-chos bsre-'pho'i
gzhung-'grel rdo-rje 'chang-gi dgongs-pa gsal-bar
byed-pa, Rtsib-ri ed.
Khrid-yig: Jo-bo Naropa'i khyad-chos bsre-'pho'i khrid
rdo-rje'i theg-par bgrod-pa'i shing-rta chen-po,
Rtsib-ri ed.
Lam-bsdu: collection of short works entitled after its first
member: Bsre-'pho lam-dbye-bsdu, Rtsib-ri ed.
Rim-lnga 'khrid-pa: Rim-pa lngar 'khrid-pa rnal-'byor pha'i
rgyud-kyi snying-po bsdus-pa, found in Lam-bsdu
(Apart from the Gzhung all these Bsre-'pho works are found
also in the Gnam-'brug Par-ma, vols. 22-3.)
Kong-sprul Blo-gros Mtha'-yas
Rgyud-bla'i rnam-'grel: Rgyud bla-ma'i snying-po'i don-gyi bshad-
srol dang sbyar-ba'i rnam-'grel phyir mi-ldog-pa'i sengge'i
M. V. Aris (1981) and A. S. Kui (eds.): Tibetan Studies in honour of
Hugh Richardson: Proceedings of the 1979 Oxford
Conference (Warminster: Aris & Phillips, 1981)
M. M. Broido (1979): The term dngos-po'i gnas-lugs as used in Padma
Dkar-po's Gzhung-'grel (in Aris (1981))
M. M. Broido (1983a): Padma Dkar-po on the two satyas (JIABS, 1985
no. 2 pp. 7-56)
M. M. Broido (1983b): Abhipraya and implication in Tibetan linguistics
(J. Ind. Phil., vol 12 (1984) pp. 1-33)
(N) H. V. Guenther: The Life and Teaching of Naropa (Oxford:
Clarendon, 1963)
(Y) H. V. Guenther: Yuganaddha (Varanasi: Chowkhamba Skt. Ser.,
(TVL) H. V. Guenther: The tantric view of life (Berkeley and London:
Shambala, 1972)
(Pers) H. V. Guenther: Tibetan buddhism in Western perspective
(Emeryville: Dharma publishing, 1977)
(LW) F. D. Lessing and A. Wayman: Introduction to the Buddhist
Tantric systems (Second ed.; Delhi: Motilal Banarsidass,
(BU) D. Seyfort Ruegg: The life of Bu Ston Rin Po Che (Rome: ISMEO,
J. Takasaki: A Study on the Ratnagotravibhaga (Uttaratantra) (Rome:
ISMEO, 1966)
(W) A. Wayman: The Buddhist Tantras (New York: Samuel Weiser, 1973)
(YG) A. Wayman: Yoga of the Guhyasamajatantra (Delhi, Motilal
Banarsidass, 1977)
P. M. Williams (1979): Tsong-kha-pa on kun-rdozb bden-pa (in Aris
(1981) ).
1. Though Guenther (Pers. 101) draws attention to the different
views of Tsong-kha-pa and the Bka'-brgyud-pas, he mentions only
Padma Dkar-po among the latter (and this only elsewhere).
2. On the particular issue of bden-gnyis zung-'jug (yuganaddha of the
two satya) and its special cases svddhisthdnaiprabhdsvara &c. Padma
Dkar-po has reviewed Tsong-kha-pa's views at some length at Phyag-
chen gan-mdzod 155a6 ff. See below and Broido (1983a). This passage
seems to form the basis for Guenther's remarks mentioned in note 1.
See below also for Padma Dkar-po's view of the difference between
zung-'jug in the sutras and the tantras. (On Padma Dkar-po and other
Bka'-brgyud-pas on the sutra/tantra distinction in general, see again
my paper (1983a)).
3. RS 29, Y 206-7 (many times), Pers. 55, 72, 73
4. Y 138
5. Y 161
6. TVL 109
7. TVL 17
8. Y 135
9. Y 206
10. Pers 101
11. Pers 75
12. Pers 72, 78, 98, 101, 109
13. Phyag-chen gan-mdzod, 155a6 ff. The passage is introduced by
Padma Dkar-po with the words Btsong-kha-pa chen-pos, which may
mean either a question from, or the attribution of an opinion to, Tsong-
kha-pa. I have not been able to find this long passage in Tsong-kha-
pa's works on the Guhyasamaja cycle, and on stylistic grounds too it
seems likely to be Padma Dkar-po's own summary of what he saw as
Tsong-kha-pa's view. I also think that Guenther is mistaken if he does
indeed think that the view expressed in this passage can be adequately
represented by "harmonious juxtaposition"; see below.
14. Pers 101 note 42
15. W 40, 129; YG 152, 153, 167, 228, 284, 312
16. LW 266-7
17. YG 172
18. LW 199
19. LW 320, 325; YG 179
20. BU62-4
21. I try to observe the uniformity rule and to retain the structure of
Tibetan sentences, but without regarding these as matters of principle.
22. See P. F. Strawson, "Categories", in Freedom and Resentment and
other essays, (London: Methuen 1974)
23. Gzhung-'grel 370a5
24. Ibid.: yuga zung/ naddha ni gnyis ma-yin-pa ste/ de'i phyir, zung-'jug
zhes btagssol
25. This nges-tshig ("etymology", nirukta) is a good example of the
kind of sgra ji-bzhin ma-yin-pa (non-literal, na-yatharuta) interpretation
which Padma Dkar-po calls yi-ge and which is typically used for forming such "etymologies" for awkward Sanskrit words. For the yi-ge style
of interpretation and its three modes, see Broido (1983b) and references given there.
26. Lam-bsdu 161a5
27. The phrase zung-du 'jug-pa and the schemata (1), (2) are syntactically ambiguous: is the phrase zung-du 'jug-pa to be taken as irreducible
or as (nom. +postp.) + verbal adj.? I have resolved this question ad
hoc by reducing the phrase, but only because this course leads to a
clearer exposition. My general logico-grammatical intuitions are in
favour of reduction, while my intuitions about this particular phrase are
that is is irreducible; but these latter intuitions are surely semantic
rather than syntactic. Its syntactic structure does seem to be the reduced
one, but argument on this point would have to rest on syntactic analysis
of its components in other contexts, such as is hardly possible here.
"Use" arguments point towards irreducibility (partly because of the
prevalence of the short form zung-'jug and the absence of significant
variations of the long form zung-du 'jug-pa, suggesting that the syntactic details suppressed in the short form are not worked hard when the
word is used). Unfortunately "use" is not very sensitive to the syntax/
semantics distinction.
28. Lam-bsdu 161a5, see (3), which is followed there by zung-'jug rim-
pa shes-par-'gyur zhes bstan-tel bden-pa gnyis gnyissu-med-pa shin-tu yang-
dag-pa'i bdag-nyid-can-gyi ting-nge-'dzin yin-pas zung-'jug-go I. The Skt. in
(5) is mine. Three equally clear examples of the phrase zung-'jug-gi ting-
nge-'dzin can be found in the headings of the lta-ba/sgom-pa/spyod-pa divisions of the 'bras-bu skye-ba'i rim-pa in the khrid-yig, as given in Appendix B.
29. The six are the utpattikrama (bskyed-rim, Lam-bsdu 153bl) together
with the usual five stages of the sampannakrama (rdzogs-rim), Lam-bsdu
155a6, 157a2, 158b6, 160a5, 161a5. For Padma Dkar-po's views on the
names, number and numbering of the krama, see Gsang-'dus-rgyan,
30. Gzhung-'grel 370a5
31. Skt. yuganaddhakdya, translated by Guenther as "unity of
opposites" (sic), Y 134. The passage comes from Naropa's Sekoddesa-
tikd. Padma Dkar-po gives the Tibetan at Gzhung-'grel 345al and Phyag-
chen gan-mdzod 38b6.
32. Especially the word so-sor rang-gis rig-par-bya, Skt. pratydtmad-
higamya &c./ bden-pa gnyis gnyis-su med-pa'i rang-bzhin 'di-nyid zung-du
'jug-pa zhes brjod-del .... so-sor rang-gis rig-par-bya'ol (ibid., cf. also (3)
in the text, and Gzhung-'grel 344b5).
33. On the Mahayana /Vajrayana or sutraltantra distinction, cf. note 2.
34. See Part I, where the word thig-le (bindu) presents a similar problem.
35. See e.g. Phyag-chen gan-mdzod 49b4 ff., Rnal-'byor bzhi'i mdzub-
tshugs 13b. These passages are translated and discussed at length, and
the first quoted, in Broido (1983a), especially from the point of elucidating how the term lhan-skyes applies to the two satya.
36. Skt. akrtrima, akrtaka. In the case of the two satya Padma Dkar-po
and Mi-bskyod Rdo-rje seem to have taken both zung-'jug and lhan-
skyes to imply that it is nonsense to speak of either one appearing and
functioning separately. They criticize Tsong-kha-pa, Bo-dong-pa and
the Jo-nang-pas for establishing relationships which are merely contingent (all this is dealt with at length in Broido (1983a)). However there
does seem to be one case where Padma Dkar-po does what he complains of in these other authors. Gzhung-'grel 375a3: "When the sight of
things as they are is obscured because accompanied by vikalpa, one
speaks of samvrti-satya or of reality obscured; this is the time when purification by the radiant light has not occurred." The Tibetan runs:
rnam-par-rtog-pa-dang-bcas-pa dngos-po'i de-kho-na-nyid mthong-ba-la
sgrib-pas-na kun-rdzob-kyi bden-pa-am yang-dag sgrib zhes kyang bya-lal
'od-gsal-gyis sbyang-ba ma-byas-pa'i skabs-sol
Padma Dkar-po seems to have slipped here, since for him samvrti-satya
normally z's purified. Perhaps the remark refers to the cittavisuddhikrama
(cf. point II in sec. 1.3). An impure samvrti cannot take part in
yuganaddha (see note 121).
38. Lhan-skyes may also be applied to single nouns (e.g. ye-shes, dga'-
ba &c., also names of deities) which describe the combinations of the
things born together.
39. Rigs Skt. kula, varna &c. See Part I.
40. Slob-pa'i zung-'jug
41. Slob-pa'i zung-'jug ni 'od-gsal rtogs-nas mthar phyin-gyi bar-gyi skad-
del bden-pa gsar-du mthong-nas de-la goms-par-bya-ba sgom-lam phye-ba Ita-
bu (Gan-mdzod 157b4)
42. Ibid.
43. Ibid. 31a2: see Part I, note 33
44. The uses of zung-'brel noted in (a)-(g) are of course non-tantric,
but I can see no reason why this word might not find uses in the
tantras also. The zung-'brel/zung-'jug difference appears on the face of it
to cut across the sutra/mantra difference.
45. On the basis on his Bka'-gdams experience, Sgam-po-pa introduced into the Bka'-brgyud tradition the view that mahamudra can be
attained in non-tantric mahayana. This view was criticized by Sa-skya
Pandita, but these criticisms have been rejected by 'Gos Gzhon-nu-
dpal. The controversy forms an important strand in the Phyag-chen gan-
mdzod, see also Broido (1983a).
46. Phyag-chen gan-mdzod 32b3 (discussion of this point continues until 35b6).
47. Rgyal-dbang-rje is the name usually used by Padma Dkar-po for
(Mi-pham) Kun-dga' Dpal-'byor (1423-78), the second of the line of
'Brug-pa incarnations of which Padma Dkar-po was the fourth. He collected together the doctrines of the 'Brug-pa tradition, which had become scattered among the lineages descending from the first 'Brug-pa,
Gtsang-pa Rgya-ras Ye-shes Rdo-rje (1161-1211). The biographies of
these personalities may be found in the Blue Annals. Rgyal-dbang-rje
had a gift for aphorism, and Padma Dkar-po often quotes his pithy formulations of key points of 'Brug-pa doctrine.
48. For instance, the three main headings of the sa-bcad of the Nges-
don grub-pa'i shing-rta say (see note 52) that in this method of setting up
madhyamaka, the ground is the yuganaddha of the two satya, the path is
the yuganaddha of prajhd and upaya, and the goal is the yuganaddha of the
two buddhakayas (dharma- and rupa-). For the first two of these see Table
1. Now it seems obvious enough that all three pairs are concerned with
something which continues to exist as "integration, but the contrast
with such negative-looking pairs such as rdgalaraga & grahyal grahaka
which have merely to be given up (Table 1) is not as clear as it may
seem. More analysis of the structure of the Nges-don grub-pa'i shing-rta
is found in Broido (1983a, esp. Appendix C). The bden-gnyis zung-'jug
section of this work uses the same Vajrayana terms (gshis, gdangs, gnas-
lugs &c.) and ideas (e.g. inseparability of ground, path and goal, 66b2)
as other, clearer treatments discussed in this article. Its use of the
madhyamaka notion of the equivalence of siinyata and pratiyasamutpada is
worth noting and is pursued further in the thabs-shes zung-'jug section,
49. I mean a concept of something non-dual, not the incoherent
"non-dual concept" of something. Concepts make distinctions, it is
what they are for. Tsong-kha-pa has recognised this in relation to the
similar non-dual category tattva (de-kho-na-nyid) in his Rigs-pa'i rgya-
mtsho, 244a4, where he points out that one has to know what a word
means before one can reflect on whether one understands that for
which is stands. As Guenther has remarked, Padma Dkar-po too
(Phyag-chen gan-mdzod 32a2, 33b4) stressed the importance of words
and letters for conveying tattva: Pers. 38-9, note 4. At the end of this
note, Guenther suggests that these Indian and Tibetan writers modelled knowing on seeing, holding that language "draws attention".
(But I am unhappy about the comparison of Saraha with Wittgenstein.)
50. See Appendix B under sems dngos-po'i gnas-lugs
51. Gzhung-'grel 331a4, Khrid-yig 94b6. Padma Dkar-po's very pithy
summary (mdor-bstan) of the stages (skabs) of ground, path and goal in
this case (Gzhung-'grel ibid.) has been translated by Guenther (N 90)
and apart from some oddities with the technical terms he conveys well
the gist of Padma Dkar-po's remarks. I hope soon to publish a detailed
account of this (radiant light) section of the Gzhung-'grel.
52. See note 48. The headings from the Nges-don grub-pa'i shing-rta
gzhi dbu-ma bden-gnyis zung-'jug-tu thag-bcad
lam dbu-ma thabs-shes zung-'jug-tu nyams-su blang-ba
'bras-bu dbu-ma sku-gnyis zung-'jug-tu mngon-du bya-ba;
they are discussed briefly in Broido (1983a) (which is mainly about the
gzhi section).
53. Mahamudra is probably the most comprehensive and historically
resonant of these terms; madhyamaka is the one which has most roots in
the kind of analysis with which our own culture makes us familiar,
while the radiant light is perhaps the most culturally unfamiliar. At
least for Padma Dkar-po, yuganaddha is the most fundamental of them,
and this is my reason for treating it, and not the others, in detail here.
54. See the section of Broido (1983a) on the distinction between gnas-
lugs phyag-chen and 'khrul-lugs phyag-chen The line of argument given
here works most easily for the samsara /nirvana case.
55. For a somewhat more systematic reflection on tensions of this
kind, see the concluding remarks of this article.
56. The Heart (hrdaya) surra and the Prajhdpdramitd sutras in general
are full of this kind of negation, of course.
57. A possible starting-point in this. Our verses appear in the Gzhung
as well as in YK, and so they receive commentary in the Gzhung-'grel,
the exact places being given in Table 1. In each case, Padma Dkar-po
sketches the point of view from which (in his view) the verse, with the
listed description, is to be taken. But even though this material is relatively short (c. 7 foil.), only very detailed study, such as cannot be recorded here, would enable us to draw conclusions of the required generality; a mere impression is of little interest.
58. Cf. note 51.
59. Gzhung-'grel 343a3 ff; see note 62.
60. Both these points are expressed in the famous verse from the
Gzhung, quoted in the second paragraph of the section "Reconstruction
of Padma Dkar-po's argument" in Part I. The goal section of the
Gzhung also opens with the line I'bras-bu skye-ba'i rim-pa nil (Gzhung
8al, Gzhung-'grel 370a5). Here the word ni ("as for . . .") does not distinguish between a title and some other kind of term referring to a section of subject-matter, but Padma Dkar-po uses the phrase 'bras-bu
skye-ba'i rim-pa explicitly as his heading several times, e.g. Gzhung-'grel
14bl, 370a4.
61. As noted earlier, other arguments of this kind are available based
on the relations of the subject-matter of the Gzhung-'grel with other
writings of Padma Dkar-po on topics divided by the scheme ground/
path/goal (rgyud-gsum). See notes 50-53.
62. Compare Gzhung-'grel 340a3 ff. with Lam-bsdu 161a5ff,; cf. similar
parallels between the path division in the radiant light section of the
Gzhung-'grel and the mngon-byang section of the Rim-lngar 'khrid-pa
(Lam-bsdu 160a5).
63.Gzhung-'grel 340a6, Lam-bsdu 161bl
64. I'Od-gsal-las Idang-ba'i tshul/. Gzhung-'grel ibid. (The Rim-lngar
contains fewer headings, sources of quotations &c.) In the Gzhung
(6a4) we find 'phar-ba for Idang-ba in the verse as quoted in the Gzhung-
'grel; these words are no doubt alternatives for Skt. utthdna.
65. Gzhung-'grel 340b3, Lam-bsdu 161b3
66. Ibid. b4, b6
67. Lam-bsdu 162a4. These three points are dealt with also in the
Gzhung-'grel, but this does not support the present argument.
68. Gzhung-'grel 341a3
69. Especially the very final part of the radiant light treatment (mthar-
thug-pa'i 'bras-bu, 344bl) is similar to the very final part of the 'bras-bu
skye-ba'i rim-pa treatment (377b5) and also to the goal section of the
Nges-don grub-pa'i shing-rta (cf. notes 50, 54).
70. Gzhung-'grel 9bl ff. (comment on the verse quoted in Part I, note
71. Some of the technical terms in this summary are explained in
Broido (1983a)
72. These two zung-'jug are still to be distinguished in some way not
clear to me. Padma Dkar-po does not seem to be identifying blta-bya
with Ita-byed, concentrating on the error of taking them as gzung-ba and
'dzin-pa, i.e. (presumably) as ontologically separate.
73. Bskyed-rim, the stage of generating the deities and the rest of the
visualization. Non-dual awareness is normally the province of the
stage of completion (utpannakrama in the verse, but the form sampannakrama is perhaps more common; Tib. always rdzogs-rim).
74. Gzhig-pa: the process by which the visualization is made smaller
and smaller until it dissolves into the void or the radiant light.
75. Sgyu-lus, mdyadeha; here the visualization of oneself as the deity
and the world as a mandala, regarded as a whole.
76. Snod-bcud; here snod (vessel) stands for the visualization of the
world as a mandala, and bcud (contents) for that of oneself as the deity.
77. For gshis see e.g. Pers 56-8; Broido (1983a). The notion of gshis
contains several tensions; really it is the capacity for paramarthasatya
but Padma Dkar-po (and Guenther, following him) often identify the
two. Further it is not clear whether gshis is subjective or inheres in objects. There is a similar difficulty over the word ngo-bo (rupa) as used,
e.g. of paramarthasatya, in such important madhyamaka loci as MMV
VI.23 and its bhdsya. If as often seems the case gshis and ngo-bo inhere in
objects, we have the incoherent notion of an objective correlate for
pure subjectivity. In any case Guenther is certainly right about the
close parallel between gshis and ngo-bo.
78. Sems-kyi gnas-lugs, short for sems dngos-po'i gnas-lugs. For the contrast between this feature-universal and the sortal universal yid (ma-
nas), see Broido (1979). It is precisely this feature-aspect which is exploited whenever the depth-clarity metaphor is brought into play (cf.
note 112 and Part I, note 51).
79. The radiant light is self-perceived (cf. Part I, note 53; also n. 77
80. I have not been able to follow the pattern (if there is one) behind
Padma Dkar-po's references to snag-stong zung-'jug, gsalstong zung-'jug
&.C. in these passages of commentary to verses in which the terms do
not occur. Padma Dkar-po differed from what he regarded as the orthodox Sa-skya and Bka-brgyud view of these four kinds of yuganaddha
(Phyag-chen gan-mdzod, 52a2, 52b2, 54a5, 55a4 gives his view of the
81. Prajhopayayuganaddha is the path of madhyamaka, according to the
Nges-don grub-pa'i shing-rta (see notes 48 and 52). This very short treatment in the Gzhung-'grel is consistent with the use of the term Prajhopayayuganaddha in either surra or mantra contexts. This consistency is a
typical sign of the use (in the latter contexts) of a father-tantra terminology; cf. notes 101-107 below.
82. Zung-'jug de-nyid. It cannot mean this particular yuganaddha (as distinct from others explained in other verses) on pain of circularity. The
remark is an inference from the general character of the good of others
to a particular aspect of this good as connected with action and agent.
Needless to say this inseparability of action and agent is one of the
most common themes both of ethical discussion (e.g. in the sutras) and
of analysis (e.g. in madhyamaka works). Padma Dkar-po takes it for
granted that his readers are familiar with all this material, supporting
the quoted remark merely with a verse from BCA.
83. The central importance of the notion of dngos-po'i gnas-lugs for
Padma Dkar-po's thought is argued on its own merits in Broido (1979).
84. Gzhung-'grel and Khrid-yig: see Appendix B; Nges-don grub-pa'i
shing-rta; notes 48 and 52.
85. E.g. Lam-bsdu 161a5, reviewed above; Phyag-chen gan-mdzod 155a6
ff, for which see below and Broido (1983a).
86. It may be worth trying to deflect some impatient reactions to this
proposal. Why do I not simply give Padma Dkar-po's own account?
But what can this mean, other than printing the Tibetan texts? Any suggestion that one can translate these without extensive discussion of the
technical terms can only rest on a confusion. Again, why do I not just
say what Padma Dkar-po's conception of yuganaddha was? If read very
informally, this is just what I will do. But if read more formally, as a de-
mand to isolate an identifiable "conception" attributable to Padma
Dkar-po on the basis of some kind of evidence, then it suffers from the
long-standing confusion which Quine has christened "the 'idea' idea".
87. The word "mystical" might itself be thought to be in need of explanation. I mean for instance the idea of an idea which contains logical contradictions, but is not thereby rendered empty; or of an experience of something to which can be strictly and literally ascribed contradictory attributes, &c.
88. Cf. notes 4,5
89. Cf. notes 11, 12. Similarly for his "identity" (e.g. N 116).
90. Since the question of metaphors in translation will come up
again, it may be worthwhile trying to say something general about it.
In word-by-word translation, the criterion of accuracy is the linguistic
function of each word in its context (what Grice has called its utterer's
meaning). Metaphors function by suggestion. If the foreign word is a
metaphor, we may try to find an English one with the same suggestive
power. But if there are great cultural differences, no such word may exist. In any case a literal translation may be irrelevant, as lacking the
right suggestive power as regards the audience for whom the translation is intended. (Because of this argument I support Guenther's complaints about the literal translation of words like vdyu (rlung) and bindu
(thig-le) by "wind" and "drop" in their technical uses, and his attempts
to replace these English words by words which would literally (in English) convey something relevant (e.g. motility, creative potentiality).) If
the foreign word is not a metaphor, we may still need one in English if
there exists no literal equivalent. So it seems there can be no general argument against the use of metaphors in translation. But there may be
particular arguments. A particular metaphor may simply have the
wrong suggestive power. Or the cultural context in English may simply not make it clear to us that a metaphorical use is intended (or which
metaphorical use). All my specific complaints about metaphors as
translations in this paper are examples of one of these two sorts of objection.
91. This is my main objection to Guenther's translation of yuganaddha
by such phrases as "unitive Being" (cf. note 6). It is the same as the objection to the translation of dngos-po'i gnas-lugs by "the concrete fact of
Being" which is discussed in Part I, note 50.
92. It is well-known that Kant's intelligible intuition is similar to
yogipratyaksa (rnal-'byor mngon-sum). Padma Dkar-po seems to have had
little use for the latter. Of course he talks a great deal about yoga (rnal-
'byor) (but cf. Part I note 76), about pratyaksa and about what Guenther,
rightly in my view, has called "intuitive understanding" (mngonsum-
du rtogs-pa). But I see no reason to think that this mngon-sum is what is
discussed in pramana texts (similar to Kant's "intuition"). We must remember that the normal meaning of the Sanskrit word pratyaksa is
"perception", and that Dignaga's claim that this is without concepts
(kabpanapodha) was a great departure from this normal meaning. In his
own pramana works, Padma Dkar-po held that the formalized pramana
of Dignaga and Dharmaklrti is purely conventional (kun-tu thasnyad-
pa'i tshad-ma); this he contrasts with "real" or "proper" pramana which
is much more like what is discussed in this paper (don-dam tshad-ma; see
his tshad-ma 'jam-pa'i dgongs-rgyan, 4bl ff., and tshad-ma rigs-pa'i snying-
po, 4b5 ff.).
93. In Part I we saw Padma Dkar-po saying that dngos-po'i gnas-lugs is
the ground of samsara and nirvana and the ground upon which the path
rests. These claims are soteriological, and have nothing to do with
Kantian a priori or metaphysical claims to ground experience.
94. Neither yuganaddha nor dngos-po'i gnas-lugs have anything to do
with sense-data or sense-datum theories. Confusions in this area may
arise from the conflation of mngon-sum (as Padma Dkar-po uses the
term) with the pratyaksa of Indian pramana works (see note 92).
95. An important technical term connected with this line of thought is
tha-mal-gyi shes-pa, lit. "ordinary cognition"; as I show (1983a), Padma
Dkar-po's own view of this term comes to the similar "natural cognition". Guenther's "primordial knowledge" (Pers 77) perhaps reflects
confusion of tha-mal (common) with tha-ma (poor, inferior; last), especially in the phrase thog-ma dang tha-ma.
96. As is well-known, confusion between unity of experience and experience of unity has a long and disastrous history in Western thought
(e.g. P. F. Strawson, The Bounds of Sense, p. 162).
97. See notes 15-19.
98. Tsong-kha-pa's view of yuganaddha will be considered below
(mainly through Padma Dkar-po's eyes).
99. Even in cases such as the two satya and svadhisthdna/prabhdsvara,
the treatment in the Pahcakrama itself is not adequately conveyed by
"pair combined" &c. for reasons argued in Broido (1983a); but for present purposes these points may be regarded as refinements.
100. Gzhung-'grel 7b2.
101. Cf. Rim-lnga rab-tu sgron-gsal, 14a5; Mtha-gcod, 26b3: Ikhyad-par de
gnyis ni pha-rgyud dang ma-rgyud-kyi gtso-bo re-la tshad-ldan-gyis bkral-ba'i
tshul snang-nal thabsshes-kyi rgyud gzhan-rnams de'i sder gtogs-kyi yan-lag
yin-pasl tshul de dngos-su mi-gsal-ba-rnams kyang phyogs gnyis-po der
drangs-nas bshad dgos-pas 'jog-tshul des rgyud gnyis-po kun-la mi-khyab-pa
yang min-nol
102. Cf. Rim-lnga rab-tu sgron-gsal 12b3; Mtha-gcod 25a3:
thabsshes so-so'i rgyud-kyi 'jog-tshul-la. . . . gnyis-pa Rdo-rje-gur-lasogs-
pa-las gsungs-pa-la. . . . legs-par gnas-pa'i phyogs bzhag-pa (ni):
o-na pha-rgyud ma-rgyud gnyis-su 'jog-tshul ji-ltar byed snyam-nal rgyud 'di
gnyis-la bskyed-rim-gyi skor-nas kyang sosor 'jog-pa'i phal-pa'i khyad-par
yang 'dod mod kyang gtso-bor rdzogs-rim-gyi sgo-nas khyad-par gzhag-pa
/de yang bdestong-gi thabs-shes-la Itos-nas thabs-shes re-re-ba'i rgyud-du mi-
'jog-par gnyis-med-kyi rgyud-du 'jog-pa sngar bshad-pa Itar yin-lal de-la
bltos-nas ni shas-che-chung-gi sgo-nas kyang gzhag-tu mi-rungstel Kye-rdor-
sogs ma'i rgyud-las bde-ba shas-cher bstan-pa Itar 'Dus-pa-las ma-gsungs-pasl
Kye-rdor pha-rgyud dang 'Dus-pa ma-rgyud-du gzhag dgos-pa-i skyon-du
'gyur-ba'i phyir-rol
Ides-na rdzogs-rim-gyi sgo-nas thabs-shes so-so'i rgyud-du gsungs-pa'i thabs-
shes ni shes-rab don-dam bde-ba-chen-po'i ye-shes dang/ thabs kun rdzob sgyu-
ma'i sku'ol
/de la dang-po'i sgo-nas rnal-'byor-ma'i rgyud-du 'jog-pa ni Gur-gyi le'u bcu-
gsum-pa-lasl. . . .
/shes-rab pha-rol-phyin-pa'i thabs/ /'di ni rnal-'byor-mar brjod-dol
/phyag-rgya chen-po rabsbyor-bas/ Igang-phyir de-nyid-la 'jug-pal
/rnal-'byor ma-yi rgyud ces-byal
103. LW 260-3 (nothing presently relevant is found in the rdzogs-rim
section of Mkhas-grub-rje's discussion of earlier views, p. 254). The account of the rang-lugs is slightly better in Pan-chen Bsod-nams Grags-
pa's Rgyud sde spyi'i mam-bzhag skal-bzang-gi yid-'phrog, but much of the
point still escapes the reader because the gzhan-lugs is almost omitted.
104. Rdzogs-rim-gyi khyad-par nil shes-rab-kyi pha-rol-tu phyin-pa'i thabs
phyag-rgya chen-mo-la sbyor-bas de-kho-na-nyid-la 'jug-pa ston-pa ma-rgyudl
rang-byin-brlab dang mngon-par-byang-chub-pa'i rim-pa gtso-bor ston-pa pha-
This entire remark occurs in Bu-ston at least twice (Rin-po-che mdzes
rgyan, 281a2; Gsang-ba gsal-byed, 127b4), and also identically apart from
the reversal of the two clauses in Bkra-shis Rnam-rgyal's Nor-bu'i od-
zer, 14a5. However there is a variant which replaces phyag-rgya chen-mo
with rgya chen-po (sic); this occurs in Bu-ston's Gsang-'dus bshad-thabs,
22al, and is repeated exactly in Padma Dkar-po's Rgyud-sde spyi'i mam-
bzhag, 34b2. The variant does not seem to make sense; and since the
Vajrapahjara verse (which is quoted in all texts, see note 102) contains
the line /phyag-rgya chen-po rab sbyor-bas/ without variation, I have accepted the quoted version for Padma Dkar-po too.
105. Padma Dkar-po, Rgyud-sde spyi'i mam-bzhag 32b5, 35a6; the latter
(nitartha) quotes the well-known GST verse on this topic which is also
quoted for similar reasons by Mkhas-grub-rje and the others.
106. The use of the terms drang-don (neyartha) and nges-don (nitartha) in
the tantras is quite different from its use in the sutras. Their use by Bu-
ston, Tsong-kha-pa and Padma Dkar-po is discussed, especially in relation to the tantras, in (1983b).
107. For reasons of this kind it would be desirable to give a detailed account of the dispute about the classification of the anuttarayoganatan-
tras in which, among other things, the bshad-thabs terms were treated
with the care their importance deserves (they do not receive this care in
Lessing and Wayman). I hope to present such an account soon elsewhere.
108. utpattikrama/sampannakrama or svadhisthdnakrama/abhisambodhi-
krama, as the case may be.
109. Much criticism of Tsong-kha-pa by Bka-brgyud-pa authors such
as Padma Dkar-po and Karma-pa Mi-bskyod Rdo-rje is related in some
way to this set of issues, often expressed as the claim that various pairs
of notions are (as Tsong-kha-pa explains them) not properly lhan-cig
skyes-pa (sahaja, born together).
110. Phyag-chen gan-mdzod 156b2, 'di pa'i lugs-la/ . . . (comment on the
immediately preceding quotation or explanation attributed to Tsong-
kha-pa (n. 13) ).
111. Gzhung-'grel 115a4, 116a2, 332bl; Yid-'phrog 17b6.
112. On the inseparability of depth and clarity in this context, we
have: tha-dad-du 'byed mishes-pa twice (Gzhung-'grel 116a4, 332b3); tha-
dad phye yang ya-bral-du phye-ba nam-yang misrid-pas (Yid-'phrog 18a2);
gnyis-pa med-pa (from the Vimalaprabhd, see Part I note 51); de gnyis
mtshan-nyid tha-dad-du phye yang ngo-bo-nyid ni gcig yin-no (Gzhung-'grel
119a4). The last passage is especially interesting because of its analysis
in terms of sahaja. The illusory body is present throughout the abhisam-
bodhikrama (Gan-mdzod 157a5). Even in the svddhisthdnakrama, the agent
of purification is the radiant light (Pahcakrama II.5-6, V.26, V.30, quoted
Gan-mdzod ibid.) Similarly, the illusory body is self-purified (criticism of
the Bo-dong-pa position, Gan-mdzod 157b6), and similarly in terms of
sealing (ibid. 39a4). In addition to all this, svddhisthdna and prabhdsvara
are instances of the two satya, on whose inseparability Padma Dkar-po
is just as insistent (see my (1983a)).
113. De gang-la dbyer-med-par 'dus-pa'i go-'phang ni zung-'jug-go (Gan-
mdzod 115b2; an almost identical remark also at 156bl).
114. Used twice of the two satya in contexts related to the present, both
times in the same remark taken from the Sekoddesatikd: bden-pa gnyis
gnyis-su med-pa'i rang-bzhin 'di-nyid zung-'jug-pa zhes brjod-dol: Gan-
mdzod 36a2 and Gzhung-'grel 345al.
115. Phyag-chen gan-mdzod 156a4 (cf. Sngags-rim chen-mo 406b, 410a,
quoted LW 192).
116. Gan-mdzod ibid: res 'jog-tu gyur-ba'i phyir zung-'jug med-dol
117. Ibid. 156a5: gnyis dus mnyam 'byung-zhing
118. Ibid. 156bl; cf. also note 112.
119. Ibid. 156b2: 'di'i pa'i lugs-la/ glang-la rwa-co Ita-bu gnyis zung-'jug-
120. Ibid. 156b4
121. Ibid. 156b5: de yang ril-'dzin rjes-gzhig-gis 'od-gsal-bar yang bcug-pas
mthar dag-pa-na mi-slob-pa'i zung-'jug 'byung-bar gsungs-pasl kun-rdzob kyi
bden-pa de ma-dag de-srid mi-slob-pa'i zung-'jug ma byung;  cf. Mkhas-
grub-rje (LW 326-7): de-nas ma-dag-pa'i sgyu-lus de-nyid ril-'dzin dang rjes-
gzhig-gi bsam-gtan gnyis-kyis. . . .
122. This topic is dealt with in more detail in Broido (1983a).
123. Phyag-chen gan-mdzod 157a5: mnyam-gzhag-tu snang-ba dag-pa
124. Padma Dkar-po identifies paryaya-paramdrtha (rnam-grangs-pa'i
don-dam) with a purified samvrti (see discussion in Broido, 1983a and
1983b), and says here that it is the mere tearing of the veil of ignorance
and contains but a little paramdrtha (Gan-mdzod, 157a2).
125. Candrakirti's distinction between samvrti-satya and samvrti-matra
was developed, somewhat differently in each case, by Tsong-kha-pa
(see Williams (1979) ) and by Padma Dkar-po (Nges-don grub-pa'i shing-
rta 36b6).
126. Cf. for instance the passages quoted by Seyfort Ruegg, BU 62-4
(but cf. also note 80).
127. This suggestive power is not needed for zung-'brel (see section
128. For present purposes all we need to know about the radiant light
is that it is self-cognising and paramarthasatya. Cf. notes 77, 79.
129. See Appendix B under lus ji-ltar grub-pa'i tshul.


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items