BUDDHA IN THE STOREHOUSE: MI BSKYOD RDO RJE ON THE RELATIONSHIP BETWEEN TATHĀGATAGARBHA AND ĀLAYAVIJÑĀNA1
BUDDHA IN THE STOREHOUSE: MI BSKYOD RDO RJE ON THE RELATIONSHIP BETWEEN TATHĀGATAGARBHA AND ĀLAYAVIJÑĀNA
The question of how buddha nature (tathāgatagarbha) relates to Yogācāra psychology, and this tradition’s idea of a substratum consciousness (ālayavijñāna) in particular, has long been a focus of intense discussion and debate among Buddhist scholars, both within and beyond the borders of India. Looking back on such exchanges, one is hard-pressed to find a scholar who has delved more deeply into the complexities of this issue than the Eighth Karma pa Mi bskyod rdo rje (1507–1554). In reviewing his many and varied treatments of the matter, it becomes clear that the question was a focal point for several overlapping problems that were integral to his philosophical project. Among these were the problems of whether and how the Yogācāra ālayavijñāna-vāsanā model could be reconciled with  buddha nature theory,  tantric buddha nature proxies such as the unconditioned ground (gzhi) and causal continuum (rgyu rgyud),  Indian and Chinese Buddhist conceptions of an immaculate consciousness (amalavijñāna), and  the Karma pa’s own tradition’s Mahāmudrā-based *Yuganaddha-Apratiṣṭhāna-Madhyamaka (a Middle
1 This article is the result of research that was generously funded by the Austrian Science Fund (FWF Project number P23826-G15). Many friends and colleagues helped me to rethink and refine the material presented herein. I am especially grateful to the two anonymous peer reviewers of this article for their invaluable editorial comments and suggestions.
Way consisting in Nonfoundational Unity). The author’s repeated forays into these contested subject areas reveal time and again his commitment to reconcile two divergent lines of Buddhist thought and praxis:  the affirmative appraisal of the nature of mind and reality emphasized in Yogācāra and tathāgatagarbha classics, the tantras, and the songs and writings of the Buddhist mahāsiddhas and  the anti-essential Middle Way (Madhyamaka) traditions that sought to avoid extremes of affirmation and denial, existence and nonexistence. His explorations reveal a thinker who was as confident about the mind’s ability to discover an unconditioned mode of being and awareness that is not conceptually determined as he was skeptical about its ability to discover any final foundation. To adequately appreciate his contribution to clarifying the relationship between buddha nature and the substratum consciousness, it is necessary to first trace in broad outline the historical evolution of the ālayavijñāna idea and its complex confrontations with tathāgatagarbha doctrine in India. A knowledge of this Indian background is, indeed, indispensable to understanding, more broadly, how different schools of Tibetan Buddhism sought to integrate buddha nature concepts and Yogācāra psychological concepts such as ālayavijñāna into their key doctrines and practices.
§ 1. The basic problem
In the centuries following the first appearance of tathāgatagarbha doctrines in India (at least as early as 2nd c. CE), opinions became divided (2010). See also Higgins and Draszczyk 2016: 37 et passim. On the uses of the terms zung ’jug rab tu mi gnas pa (and zung ’jug rab tu mi gnas pa’i dbu ma) by Sgam po pa and many of his successors, see Higgins and Draszczyk 2016: 33–35, and Higgins and Draszczyk 2019: vol. 1, 31 et passim. over whether buddha nature should be identified with or distinguished from the Yogācāra idea of a substratum consciousness (ālayavijñāna). Likely the earliest text to equate tathāgatagarbha and ālayavijñāna is the Laṅkāvatārasūtra which deems tathāgatagarbha to be a mere epithet for ālayavijñāna, employed as a provisional locution to dispel the fear of no-self doctrine among the ignorant.
That the question was in large part a semantic one, was not lost on several of the Tibetan scholars who would later struggle to clarify this relationship. Indeed, to determine whether tathāgatagarbha is the same as or different from the ālayavijñāna (or both or neither for that matter) requires that one first determines the conditions necessary for applying these terms in shifting semantic contexts. To make matters more complex, both these concepts were increasingly the target of anti-foundationalist critique, especially by philosophers of the Madhyamaka tradition, who were ever suspicious of the ontological commitments behind their usage and ready to call into question the scope and validity of the concepts themselves.
It is little surprise, then, that in the development of Buddhist thought in India and beyond, each of these two terms came to be interpreted in widely different ways according to changing sectarian and doctrinal climates. Each new generation of scholars was newly confronted with the task of clarifying the relationship between tathāgatagarbha and ālayavijñāna in light of these ever-mutable contexts of interpretation. Before assessing the Eighth Karma pa’s contributions to the issue, it may be useful to examine the development and explanatory role of the ālayavijñāna concept in Indian Buddhism, giving special attention to the divergent attempts to either identify it with or distinguish it from tathāgatagarbha and the nature of mind along the lines of what I have elsewhere termed unity and differentiation models. The fact that Mi bskyod rdo rje’s own attempts to elucidate this complex relationship were informed that its composition and initial circulation may have occurred between the late first and mid-second century CE. See Radich 2015b: 85. by an unusually extensive knowledge of this background makes such an overview a useful point of departure.
The origins and development of the ālayavijñāna concept in Yogācāra literature have been documented and debated elsewhere and need not be reprised in detail here. For present purposes, our attention will be confined to the role the concept played in Yogācāra attempts to describe and explain the conditions of human bondage and liberation. In addressing this issue, leading Yogācāra scholars expanded the traditional Buddhist sixfold model of mind into an eightfold model to better account for the genesis, continuity, and possible transcendence of dualistic cognition. A useful summary of this development is offered by Mchims ston Blo bzang grags pa (1299‒1375) in his commentary on the Abhidharmakośa:
The two Śrāvaka schools [[[Vaibhāṣika]] and Sautrāntika] claim that consciousness is sixfold. The two Ācārya brothers [[[Vasubandhu]] and Asaṅga], however, assert it is eightfold [by including]:  a substratum consciousness (ālayavijñāna) that indistinctly yet incessantly grasps all outer and inner [referents], the world and inhabitants, by objectifying [them]; and  an afflicted mind (kliṣṭaṃ manas) that has the aspect of grasping that [[[substratum]] consciousness] itself as an “I” by objectifying [it].
The substratum and afflicted modes of consciousness are added to the traditional sixfold scheme to better account for the onset and latent structuring of experience in terms of self and other, “I” and “mine.” To understand the role this new model played in Buddhist soteriology, we need to look more closely at its explanatory force and limitations. What range of phenomena did it seek to describe and explain? Recent scholarship has cast light on several problems of continuity that the ālayavijñāna idea attempted to resolve, and that were thought to be inadequately explained in the Abhidharma account of mind. Primary among these were the continuities of various elements of saṃsāric existence including consciousness (vijñāna), feelings (vedanā), vitality (āyus), personal identity (ahaṃkāra), the mind-body complex (nāmarūpa), latent tendencies (anuśaya, vāsanā), and the relation between actions and results (karmaphala). Most vexing was the problem of accounting for the continuity of consciousness, personal identity, and karmic maturation (positive and negative) after periods of unconsciousness or during the transition from one rebirth to the next.
In early Abhidharma exegesis, the conception of a “sub-threshold” mode of consciousness, largely inaccessible to direct reflection, gradually took shape to account for how these continuities play a constitutive role in saṃsāric existence. To better explain the genesis and perdurance of karmic and afflictive conditioning of mind both within and beyond this life, the Abhidharmic sixfold analysis of consciousness was broadened to include a mechanism for the sedimentation of latent tendencies from previous experience that condition the mind and structure perception in terms of subject and object. One subsequently encounters a number of more or less ad hoc attempts in the Abhidharma system to explain the influence of past experience on the present. These included inter alia the realist Sarvāstivādin theory of possession (prāpti) that posited a dharma called prāpti (“obtaining” or “acquisition”) that acts as a kind of metaphysical glue, binding karmic inheritance to a particular mental stream. Another model was the nonrealist “seed theories” of the Sautrāntika school that introduced the “explicitly metaphorical notion (prajñapti-dharma) of seeds (bīja) to represent both the latent afflictions and accumulation of karmic potential within the mental stream.”
It is within Yogācāra-Vijñānavāda thought that one meets with the first systematic attempt to account for this ongoing sedimentation of experience. This tradition’s analysis of latent tendencies, literally “perfuming” (vāsanā : bag chags), sought to explicate in a more methodical and comprehensive fashion those unconscious constitutive processes that remain largely inaccessible to direct apprehension, but that nonetheless influence consciousness at every moment. On this view, consciousness is never wholly accessible to direct reflection, since it is subliminally influenced at every turn by latent traces of previous experience. Stated otherwise, consciousness lives in the medium of its own history, which, however, remains largely unavailable to it. It is karmically-affected insofar as it operates in the light of the past and in anticipation of the future, and does so, by and large, under the influence of its own sedimented habits, presuppositions, and inclinations. We have seen that the ālayavijñāna-vāsanā model allowed Yogācāra scholars to account for the largely unconscious constitutive processes that condition and structure the dualistic mind (citta) at each moment. But it left unanswered an important religious and soteriological problem. Although the ālayavijñāna-vāsanā model certainly helped explain the sources of dualistic consciousness, it could not, on its own terms, explain the sources of spiritual awakening. A concomitant soteriological model was needed to specify why mind’s cessation should result in anything other than cognitive oblivion. The real issue was how a soteriological model premised solely on the cessation of the mind (cittanirodha) could account for the genesis of the nondual wisdom of a buddha, the goal of the Buddhist path.
It was in light of this concern that there arose alongside speculations about a substratum consciousness a number of doctrinal innovations, some internal to and others external to the Yogācāra-Vijñānavāda system, that sought to expand the Yogācāra picture of mind to include a more fundamental, nondual mode of being and awareness. Just as Yogācāra doctrine required a conception of karmically-affected mind to account for the genesis and continuity of conditioned existence, so various scholars in India, China, and Tibet sought to make room for a deeper layer of consciousness that remained primordially unaffected and unafflicted by the ālayavijñāna. We can discern in such developments a persistent effort to articulate the conditions of possibility of a freedom from the conditioning of the afflicted and afflictive mind that did not consist in the mere absence of cognitive activity. We may also observe that this endeavour often drew inspiration from a long-standing disclosive model of liberation that viewed goal-realization not as an altered state of mind, but rather as a discovery of latent modes of being and awareness that are progressively revealed to the extent that what obscures them has been dispelled. Among other things, this model provided an ingenious framework for describing and explaining how the cessation of dualistic mind results not in cognitive extinction, but in the Buddhist goal of spiritual awakening (bodhi), the ultimate awakened mind (bodhicitta).
Amongst Tibetan scholars, the idea of a deeper ground or source – a pure ninth consciousness, or uncorrupted wisdom, which was deemed to be ever-present beneath or beyond the threshold of the ālayavijñāna – provoked a great deal of discussion and controversy. To get a sense of how different, and even divergent, such ground conceptions could be, consider the following passage by the 14th century Upper ’Brug pa (stod ’brug) Bka’ brgyud master ’Ba’ ra ba Rgyal mtshan dpal bzang (1310– 1391). In this illuminating extract from his commentary on Yang dgon pa Rgyal mtshan dpal’s (1213–1258) Ngo sprod bdun ma’i mgur, ’Ba’ ra ba assesses the relative scope and significance of developmental and disclosive models of the ground:
Now, the term “ground” (gzhi) is according to some systems a ground which is something like a field. This ground is held to be that [locus] where things ripen individually in accordance with what has been planted, like barley, wheat, lentils, and so forth. But this entails the fallacy of [taking] ground and results [lit. “fruits”] as different things, because if this productive ground is [taken as] a field, then the resultant barley and lentils and so forth are different from the field’s soil.
In this regard, Chos rje Rin po che [[[Yang dgon pa]]] declared that what is termed “ground” is spontaneously present as the actual basis of all phenomena subsumed under saṃsāra, nirvāṇa and the path, and this ground abides naturally. Yet, it assumes specific guises when it encounters particular conditions and [may therefore] manifest as anything whatsoever. As an example, it is said to be similar to a crystal ball. When this crystal comes into contact with a condition such as something painted [[[red]]], it turns red, or, when it comes in contact with indigo, it turns blue. But even if it appears to turn red, the crystal has not changed in essence. And though it seems to turn blue, the crystal remains unchanged. So, the crystal may turn various colors, but it does not in essence turn into something else. Likewise, mind may go astray into the painful experiences of the hot and cold hells, but it has not for an instant changed in essence and turned into something evil. Even when buddhahood occurs as a result of realization, the essence of mind has not for a moment changed into something good. It is not that mind in itself realizes or fails to realize [anything]. In mind, there is neither good and evil nor anything that becomes differentiated.
’Ba’ ra ba here draws attention to a long-standing issue that had by his time (14th century) become a subject of intense debate. Can the disclosive 15 Ngo sprod bdun ma’i ’grel pa man ngag rin po che’i sgron me, 2112–2121: de yang gzhi zhes pa ’ga’ re’i lugs kyis zhing lta bu cig gzhi yin la / nas dang gro sran la sogs pa gang btab pa bzhin so sor smin pa cig la gzhir bzhed de / gzhi ’bras tha dad du gyur pa’i skyon yod ste / skyed byed kyi gzhi zhing yin kyang // ’bras bu nas dang sran la sogs pa zhing sa dang tha dad du ’gyur ba’i phyir ro // ’dir chos rje rin po che’i bzhed pas / gzhi zhes pa ’khor ’das lam gyis bsdusa pa’i chos thams cad kyi dngos gzhir lhun gyis grub cing gzhi gzhag tu gnas te / rkyen gang dang phrad pa’i rang gzugs ston cing cirb yang ’char ba ste / dpe shel sgong lta bu cig la bzhed de / shel de nyid tshos la sogs pa’i rkyen dang phrad na dmar por ’gro zhing / rams dang phrad na sngon por ’gro yang / dmar por song yang ngo bo shel las ’gyur ba med / sngon por yong yang shes las ’gyur ba med / de bzhin du kha dog sna tshogs su ’gyur yang ngo bo shes las ’gyur ba med pa bzhin du / sems ’di ’khrul pas dmyal ba tsha grang gi sduga bsngal myong yang sems kyi ngo bo las ’gyur ba’i ngan du skad cig kyang ma yongs / rtogs te ’bras bu sangs rgyas pa’i dus na’ang / sems kyi ngo bo las ’gyur ba’i skad cig kyang bzang du song ba med cing / sems kho rang rtogs ma rtogs min pa sems la bzang ngan nam tha dad du song ba med de / … atext: bsdug; btext: spyir.
idea of goal-realization as the discovery or re-cognition of a basic ground, identified as the unfabricated nature of mind and reality, be reconciled with those (Sautrāntika and early Yogācāra) models that construe goal-realization as a process of maturation or ripening that results from specific causes and conditions? The metaphor of a productive ground (skyed byed kyi gzhi) likened to a field is here deemed inadequate to capture the unchanging nature of mind itself, an innate mode of being and awareness that, like a crystal ball, remains invariant through the myriad transformations it appears to undergo. The organic ground-as-field metaphor works with the idea of a developmental ground wherein causes (hetu) of bondage or liberation ripen into their respective results (phala) and wherein causes and results are distinct from the ground itself. The disclosive ground-as-crystal ball metaphor works with the idea of an invariant disclosive ground, i.e., the incorruptible nature of mind and reality, that remains just as it is even while being (mis)taken for saṃsāra or nirvāṇa. Interestingly, the same tension between developmental and disclosive models animated discussions and debates over buddha nature and the status of buddha qualities. Are buddha qualities a matter of nurture or nature? Are they acquired and developed or are they uncovered and disclosed? Are they “ripened” in a person through a particular combination of causes and conditions or are they disclosed when whatever obscures them is dispelled? In certain Yogācāra, tathāgatagarbha, and tantric discourses, the coordination of these developmental and disclosive models was integral to varying attempts to reconcile the ālayavijñāna and tathāgatagarbha, whether through underscoring their identity or their difference. In any event, the question of what to do with the ālayavijñāna paradigm in the face of a broad range of influential Yogācāra, tathāgatagarbha, and tantric doctrines emphasizing an unconditioned nondual mode of awareness (some of which will be examined in the pages to follow) led to diverse, and often strikingly divergent, systems of reconciliation in India, China, and Tibet.
§ 4. Systems of identification and differentiation
These can be roughly divided into:  systems of identification in which ālayavijñāna is elevated into a monistic principle, a common cognitive source of all phenomena, saṃsāric and nirvāṇic phenomena alike, that is at times equated with buddha nature (taken in this universal sense), and  systems of differentiation which emphasize a basic distinction between the ālayavijñāna and an unconditioned absolute variously described in terms of buddha nature, the nature of mind, and the nature of reality. The differentiation model was typically aligned with a strongly innatist view of the ultimate (buddha nature, the nature of mind, or the nature of reality) which underscored its “sublime otherness” (gzhan mchog) from all that is conventional and adventitious. By contrast, the identification model, predicated on the acceptance of a common ground uniting all conditioned and unconditioned phenomena, emphasized the pervasiveness of the ultimate within the conventional in order to indicate how the ultimate permeates the mind-streams of individuals in bondage. Each system attempted in its own way to specify the relationship (identity or difference) between conditioned and unconditioned modes of consciousness and to chart the transition (path) between them.
Indian authors tended to mention (occasionally) the Śrīmālādevīsiṃhanādasūtra (ŚDS) and Laṅkāvatārasūtra (LAS) (e.g. in Śāntideva’s Śikṣasamuccaya). References by Bhāviveka, Candrakīrti, and Kamalaśīla suggest that tathāgatagarbha was understood by them to be closely related to teaching about ālayavijñāna, and their opinion seems to have been shaped in particular by the Laṅkāvatārasūtra. See Kano 2016: 34–36. The innatist strain of Buddhist thought looks back upon a long history of Indian ideas concerning the luminous and stainless nature of mind.18 Let us briefly review this strain of thought. The idea that mind is originally and naturally luminous, but temporarily obscured by adventitious defilements, has been a recurrent, though by no means homogeneously formulated, preoccupation of Buddhist thought since the time of the Pāli Canon. Its earliest known expression occurs in the Aṅguttaranikāya (AN):
O bhikkhus, this mind is luminous, but it is indeed defiled by adventitious defilements. The issue of whether and in what sense mind can be considered naturally luminous had already become a focal point of deep controversy within early Sarvāstivāda schools, as Eli Franco has shown in his analysis of portions of the so-called Spitzer Manuscript, believed to be the oldest philosophical manuscript in Sanskrit (dated to the Kuṣāna period 3rd c. CE). The concept of luminous mind was in any case by this time quite widely accepted amongst early Buddhist sects, and one finds the metaphor of a crystal which only appears to change colors against different backgrounds occasionally used to illustrate the idea that mind’s nature remains unmodified despite its temporary “colorations” by adventitious (āgantuka) defilements. On this interpretation, soteriology is a matter 18 See Aṅguttaranikāya, vol. 1, p. 10. For a detailed survey of the luminous mind (prabhāsvaracitta), with many examples of its occurrence in Indian Buddhist literature, see Seyfort Ruegg 1969: 412–437. See also Radich 2008 and 2016, which explore Paramārtha’s amalavijñāna and related ideas in light of this Indian background. For a recent discussion of the early Buddhist notion of the “luminous mind,” see Anālayo 2017.
of clearing away adventitious defilements so that the originally pure mind can reveal itself, as it really is. All this points toward tathāgatagarbha doctrine, especially as presented in the RGVV, reflecting a train of thought that could at times diverge from the Yogācāra, and also tantric, view that mind is thoroughly contaminated by conditioning factors and therefore needs to be fundamentally transformed to be liberated.23 Much depended on whether the doctrine of transformation was thought to describe a process consisting in the transmutation of these defilements or, rather, in their complete elimination. On this note, it is worth mentioning that the Buddhist doctrine of transformation, as Sakuma has observed in his study of āśrayaparāvṛtti, was employed within two different models: replacement and elimination. Within the replacement model, as presented in the Śrāvakabhūmi, an old basis of badness or malaise (dauṣṭhulya) is replaced by a new basis of Sautrāntikas and Vaibhāṣikas rejected this view, claiming that mind is not originally pure but is, on the contrary, originally sullied by karma and kleśas (Lamotte 1962: 238).
23 According to Franco (1997: 87), the fundamental transformation (āśraya-parāvṛtti/o- parivṛtti) and luminous mind (prabhāsvaracitta) models are both found in Yogācāra texts but are seldom associated with each other. On the other hand, the two models are closely associated within tathāgatagarbha exegesis. For his arguments and discussion of relevant sources, see Franco 1997: 87–89. On the association of āśraya-parāvṛtti/o-parivṛtti and prabhāsvaracitta in the Ratnagotravibhāga and its Vyākhyā, see Seyfort Ruegg 1969: 419–424. On their relationship, the author states: “C’est en relation avec la luminosité naturelle de la Pensée et de la pureté du tathāgatadhātu au point de vue de sa connexion avec le plan du Fruit que la RGVV fait état de la transmutation de la Base psychique, cette āśrayaparāvṛtti, correspondant ainsi à l’épuisement des impuretés.” In this connection, Franco (1997: 88) pertinently poses the question of “whether the doctrine of prabhāsvaracitta in Yogācāra appears only in Maitreyanātha texts (and of course in commentaries thereon), and if so, whether this could be explained by the fact that Maitreyanātha and his tradition were strongly influenced by tathāgatagarbha ideas.” That the five texts traditionally ascribed to Maitreyanātha in Tibet were indeed produced by the same hand has been contested in modern scholarship, especially with regard to RGVV. See Kano 2016: 20–31.
ease (praśrabdhi). In the elimination model, as presented in the Bodhisattvabhūmi, the basis of badness is eliminated without replacement. It is clear that an elimination model underlies the thesis, advanced in a wide range of tantric and buddha nature works, that goal-realization depends not on modifying the conditioned basis of cognition (e.g. ālayavijñāna) from “worse” to “better,” but rather of clearing it away entirely – on the assumption that it is not, in any case, constitutive but thoroughly adventitious – so that a primordial mode of being (tathatā), which it has temporarily obscured, can reveal itself. All this goes to show that the tension between what we have termed developmental and disclosive models of awakening already has a long and complex history in Indian Buddhism. It is therefore unsurprising that it so often surfaces in Bka’ brgyud and Rnying ma discussions concerning the nature of ground, path, and result in relation to the process of awakening. At the heart of these contrasting models and root metaphors lies the soteriological problem of how to integrate a view of karmically affected cognition into a disclosive view, which gives primacy to an originally pure mode of cognition that remains unaffected by karmic conditioning or causal production.
To better understand this tension, it is necessary to look more closely at these conflicting identification and differentiation models in view of the problems of reconciliation that their confluence in Tibet provoked. Our focus will be limited to specifying  the range of phenomena (within differing views of mind) that each model was intended to characterize and  some of the systemic problems these elicited. The assessment of these problems also requires a brief consideration of buddha nature views that came to prominence in India during the later stage of
an extent that it is perceived and felt most fundamentally as a situation of affliction, suffering, degradation, malaise and powerlessness. It has the effect of hindering, physically and mentally, a yogin’s ability to attain his goal (Davidson 1985: 177). Connotations of existential unease, badness, and self-recrimination are combined with moral notions of fault, failing, recrimination, and hindrance in the Tibetan rendering gnas ngan len, lit. “identifying with (len) a situation (gnas) of baseness/badness (ngan).” See Sgra sbyor bam po gnyis (Ishikawa 1990) s.v. dauṣṭhulya: dauṣṭhulya zhes bya ba du ni smad pa’am ngan pa / ṣṭhā gatinivṛttau zhes bya ste gnas pa la bya / la ni ādāna ste len ba’am ’dzin pa’o // gcig tu na duṣṭu ni nyes pa ’am skyon gyi ming la ni gong du bshad pa dang ’dra ste / spyir na ltung ba dang sgrib pa’i ming ste gnas ngan len du btags //
Identification: Identification strategies typically involved doctrinal transformations whereby the ālayavijñāna of Yogācāra thought is conceived primarily as the source of all saṃsāric phenomena was reinterpreted as a common substratum (ālaya) of saṃsāric and nirvāṇic phenomena, a ground of pollution (saṃkleśa : kun nas nyon mongs) as well as purification (vyavadāna : rnam par byang ba). The most striking and controversial instance of this monistic trend was the Laṅkāvatārasūtra’s identification of the ālayavijñāna with tathāgatagarbha.26 In a similar vein, a much-quoted passage from the now-lost Mahāyānābhidharmasūtra was also at times used as scriptural support for an absolutized conception of the ālayavijñāna as the ultimate source of all phenomena: The beginningless element (dhātu) Is the basis of all phenomena. Because it exists, [it allows for] all forms of life As well as the attainment of nirvāṇa.
The semantic fecundity of the term dhātu employed in this passage has allowed for a wide range of interpretations in Buddhist thought, not least of all in tathāgatagarbha and ālayavijñāna contexts. Several modern 26 On this interpretation and some of its Tibetan proponents such as the Bka’ brgyud masters ’Gos Lo tsā ba Gzhon nu dpal and ’Ba’ ra ba Rgyal mtshan dpal bzang, see Mathes 2008: 18, 117 and 464, n. 612. On the basis of this identification of the ālayavijñāna with the tathāgatagarbha, the Laṅkāvatārasūtra interprets āśrayaparāvṛtti as the transformation or purification of the seventh consciousness (manas) which liberates the pure ālayavijñāna. See Lai 1977: 67f. On some of the critical responses to this identification in the Rnying ma tradition, see Higgins 2013: 151‒154. scholars have commented on the term’s polysemy, noting an abundance of significations which include “natural element” (e.g., the five mahābhūta), “corporeal element” (such as the bodily humors), “psychophysical element” (e.g., eighteen dhātu), “mental element,” “constituent part,” “ingredient,” “raw material,” “disposition,” “mineral,” “principle,”
“nature” (svabhāva), “cause” (hetu), “sphere,” “realm,” “expanse,” “relic,” and “ashes of the cremated body.” The semantic polyvalence of dhātu meant, in effect, that it could be used to support either tathāgatagarbha or ālayavijñāna doctrines as the context demanded. Thus, in RGVV, the “beginningless dhātu” of the passage in question is taken to refer to tathāgatagarbha, the transmundane and naturally pure garbha, which is conceived as the “matrix of all phenomena” in the sense that “it is the basis, ground, and foundation of all dharmas [i.e., buddha qualities] that are indivisible [from it], not recognized as being separate [from it], and unconditioned.” However, in the Mahāyānasaṃgrahabhāṣya
“disposition”) or dbyings (“expanse,” “sphere”). Seyfort Ruegg (1969: 494–496) has noted the semantic affiliations between dhātu and the nascent ālayavijñāna and tathāgatagarbha concepts. See the Abhidharmasamuccaya (AS, p. 15), for example, where dhātu is characterized as the “seed of all phenomena” (sarvadharmabīja), an identification that connects the Sautrāntika bīja theory with tathāgatagarbha concepts such as dhātu and gotra. Seyfort Ruegg also observes that the Bodhisattvabhūmi establishes semantic equivalences between dhātu and the concepts bīja, gotra, ādhāra, niśraya, hetu, and so forth. In RGVV, the term dhātu is used more often than garbha in reference to buddha nature. See Seyfort Ruegg 1969: 261–264. The author does note, however, “…qu’il existe un certain flottement dans l’emploi du mot dhātu, et que ce mot n’est pas un synonyme exact de tathāgatagarbha, encore que les deux termes s’emploient souvent comme des équivalents” (ibid. 261, n. 1). He elsewhere comments that “…while the tathāgatagarbha is said [in RGV] to exist in all sentient beings without exception, the tathāgatadhātu on the other hand is present not only on the level of ordinary beings but also, evidently, on the level of buddhahood itself” (Seyfort Ruegg 1989: 19). Takasaki (1966: 91) has noted that the term in the compound buddhadhātu signifies both the nature (svabhāva) of a buddha and cause (hetu) for becoming a buddha. On further polysemic significations of the term dhātu, see Radich 2008, 2016, and Jones 2015. See also Monier-Williams 1899 and Böhtlingk 1998 [1879–1889] s.v. dhātu.
and Triṃśikābhāṣya, the dhātu of this passage is taken to refer to the ālayavijñāna, in keeping with the Yogācāra standpoint of these works.32 The irony here, as Ronald Davidson has pointed out, is that the author of the passage was likely partisan to neither of these theories but “merely wished to delineate a rudimentary form of an imperishable element which was soteriological in nature, yet acted as the basis for the stream of consciousness of an individual in bondage.”
Now, the term dhātu, as noted previously, can signify, among many other things “natural element.” Taken in this sense, it was, in RGVV, associated specifically with the “element of space” (ākaśadhātu), the stable, unchangeable element that is the foundation of the other three mutable elements (mahābhūta) earth, water, and wind, though it itself has no foundation in any of them. Thus it is compared with the pure mind (cittaśuddhi) that is considered to be “the foundation of all incorrect mental engagements (ayoniśomanaskāra) [though this] nature of mind (cittasya prakṛti) itself has no foundation in any [such] phenomena.” In some earlier tathagatagarbha texts, dhātu was already employed as a shorthand for buddhadhātu or tathāgatadhātu, two of the most common epithets of buddha nature. Seyfort Ruegg has drawn attention to a number of doctrinal contexts wherein dhātu was used to bridge nascent ālayavijñāna (gotra, bīja) and tathāgatagarbha theories. It does not require much conjecture to see how this idea of a fundamental element (dhātu) or seed of all phenomena (sarvadharmabīja) could at times be identified with the idea of a basic ground or foundation that is the source not only of saṃsāric phenomena, but nirvāṇic phenomena as well. On balance, however, this monistic trend seemed to find more detractors than supporters amongst Buddhist scholars in India. vinirbhāgadharmāṇāṃ muktajñānānāṃ saṃskṛtānāṃ dharmāṇāṃ niśraya ādhāraḥ pratiṣṭhā tathāgatagarbha iti / 32 See Takasaki 1966: 290.
In China, such an identification was endorsed by certain Chinese Yogācāra scholars such as Hui-yüan, who drew scriptural support from Guṇabhadra’s recensions of the Śrīmālādevī and Laṅkāvatāra sūtras,38 even though Hui-yüan’s own teacher Paramārtha (499‒569) explicitly rejected such an identification. The rapprochement between these lines of thought in China has much to do with their close historical association and, more precisely, with the fact that the principal texts of both systems were translated at around the same time and by the same Buddhist teachers.39 In Tibet, the trend toward the identification of ālayavijñāna and tathāgatagarbha seems to have garnered more criticism than support, whether it was explicitly rejected as bad theorizing or explained away as a rhetorical ruse to lure the spiritually immature. However, one does find an important, and all but overlooked, strain of early Rdzogs chen exegesis (8th to 12th c.) that equates buddha nature (or rather “bodhi quintessence,” byang chub snying po; the Sanskrit *bodhigarbha is not attested) with the substratum (kun gzhi) itself, based on an understanding of both terms as virtual synonyms of ultimate bodhicitta. However, when the Yogācāra ālayavijñāna enters the picture, as
and also the Vijñānavādin’s ādānavijñāna (= ālayavijñāna) has been taught for the sake of certain persons who have not freed themselves from the dogmatic postulation of a self (ātmagrāha)” (Seyfort Ruegg 1989: 40). See also Eckel 2008.
38 See Paul 1984: 51.
39 According to Paul (1984: 6–7), “[s]ince tathāgatagarbha literature was translated at the same time as Yogācāra and by the same masters, these two types of thought became closely linked in the minds of their Chinese audience … Paramārtha’s ideas, particularly his concept of amalavijñāna or ‘pure consciousness,’ have often been regarded as an amalgam of Yogācāra and tathāgatagarbha, because of the philosophical interfusion begun in India and the historical association of the two doctrines from the outset in China.” On the life and teachings of Paramārtha, see also Frauwallner 1951; Seyfort Ruegg 1969: 439f., 109f.; Radich 2008 and 2016; and Paul 1984.
it does increasingly from the 9th century onward, it is invariably contrasted with the absolute kun gzhi (along the lines of the above quotation of ’Ba’ ra ba) and relegated to the conventional level of transient, conditioned phenomena that disappear at the time of realization. Differentiation: Another line of response to the encounter between tathāgatagarbha and Yogācāra currents of thought was to sharpen and radicalize the difference between the ālayavijñāna and the unconditioned ultimate. According to certain Yogācāra works such as the Dharmadharmatāvibhāga, the ālayavijñāna is identified as the basis of all defilement that needs to be fundamentally transformed (āśraya-parāvṛtti, o-parivṛtti) or purified away for the realization of suchness to occur. The RGVV also discusses this transformation of basis but without any mention of the term ālayavijñāna. Rather, the transformation of basis describes the disclosive process whereby the tathāgatagarbha, which is said to be a term for suchness possessing defilements (samalā tathatā) or the element not yet liberated from the sheath of afflictions, gives way to suchness without defilements (nirmalā tathatā), which is equated with the dharmakāya of the Tathāgata and characterized as the perfectly transformed basis on the level of buddhahood. The acceptance of a mode of consciousness more fundamental than ālayavijñāna is implicit in the distinction between ālayavijñāna and supramundane mind (lokottaracitta : ’jigs rten las ’das pa’i sems) drawn with bodhimaṇḍa (“seat [or supreme place] of awakening”) which was rendered as byang chub snying po in Tibetan.
in Mahāyānasaṃgraha 1.45‒48. In Sthiramati’s commentary on Triṃśikā 29‒30, a similar distinction is drawn between ālayavijñāna and the supramundane wisdom (lokkottarajñāna : jigs rten las ’das pa’i ye shes) that overturns or replaces it (parāvṛtti). We may note also that the distinction between dual consciousness (vijñāna) and wisdom (jñāna), and the transformation of the former into the latter, was central to the Highest Yoga tantras (rnal ’byor bla na med pa’i rgyud) and to the Indian and Tibetan tantric works based on these. The distinction is, for example, one of the central topics of the Profound Inner Meaning (Zab mo nang gi don) of the third Karma pa Rang byung rdo rje (1284‒1339) – a doctrinal summary of the Yoganiruttara tantras – and its many commentaries.
§ 5. Models of Transformation
As a general observation concerning such differentiation models, we may point out that the Tibetan exoteric and esoteric tantric models of transformation tended to either  distinguish between two modes of the substratum – viz. a defiled mode that is the basis of saṃsāric existence and an undefiled mode that is the basis of awakening, or  introduce a ninth consciousness or ninth ground beyond the ālayavijñāna.
 Bivalent substratum conceptions became widespread in the ascendant Tibetan Buddhist orders during the later diffusion (phyi dar) of Buddhism in Tibet. Examples are the various Bka’ brgyud distinctions between pure and impure substrata (e.g. dag pa’i kun gzhi versus ma dag pa’i kun gzhi), the Jo nang distinction between substratum wisdom and substratum consciousness (kun gzhi ye shes versus kun gzhi rnam shes), and the complex variety of Rnying ma differentiations between genuine and conditioned substrata (e.g. don gyi kun gzhi versus rkyen gyi kun gzhi), as well as this tradition’s distinctive Rdzogs chen snying thig distinction between an unconditioned primordial ground (gdod ma’i gzhi, ye gzhi) and the various conditioned substrata (typically, four kinds of kun gzhi).47
 Systems of transformation positing a factor beyond the ālayavijñāna48 were elaborated in the above-mentioned works attributed to Maitreya-Asaṅga which specified the luminous mind (prabhāsvaracitta) and dharmakaya or dharmatā to be modes of being or awareness distinct from, but also a precondition of, the ālayavijñāna. This line of thought had a considerable influence in Tibet and China as we see reflected in the works of early figures such as Paramārtha (499–569) in China and Ye shes sde (8th c.) in Tibet. The point emphasized in these systems is not that the unconditioned absolute is simply the result (phala) of the transformation of ālayavijñāna but is rather that pre-existing ground (ālaya) which remains when this conditioned and conditioning overlay has been purified away. It is worth noting, for example, that thinkers of the Tibetan Rdzogs chen Snying thig tradition at times considered the idea of fundamental transformation – literally, “a transformation of the basis” (āśraya-parāvṛtti, o-parivṛtti : gnas ’gyur) – to be of merely provisional meaning since it was employed with the hidden intention (ldem dgongs) of guiding beings in accordance with their varying interests and degrees of understanding. Their reasoning can be summarized
48 It is of interest to note that the term ālaya without vijñāna is already being used in the Ghanavyūha to denote the different Bodhi sattva levels. See Seyfort Ruegg 1973: 35 and Mathes 2008: 442, n. 297.
in this way: if human reality is, in its most ontologically primary condition, spontaneously present and unconditioned, then its realization requires no production or modification by means of causes and conditions.
In China, the idea of an originally and naturally stainless mode of consciousness beyond the ālayavijñāna gained popularity in the sixth century due to the influence of the Indian Yogācāra monk and translator Paramārtha (499–569). This scholar controversially posited a ninth, immaculate consciousness (amalavijñāna), which is unaffected by the conditioning influences of the ālayavijñāna (the karmic “seeds and fruits” of attachments and aversions), and which is closely associated with the perfect nature (pariniṣpannasvabhāva) and suchness (tathatā). For Paramārtha, this amalavijñāna is invariant and undefiled (anāsrava) in contrast to the ālayavijñāna which is transient and defiled (sāsrava). While the ālaya is the source of afflictive emotions and badness (dauṣṭhulya), the amala is the abiding source of nonconceptual wisdom (nirvikalpajñāna) and saintly activity. According to Paramārtha, the fundamental transformation of ālayavijñāna entails its complete elimination, resulting in the realization of pure consciousness (amalavijñāna). According to Paul Demiéville, the issue of whether the ālayavijñāna or amalavijñāna should be regarded as the basis of consciousness and the world itself was already the subject of heated doctrinal controversy in China before Paramārtha’s arrival and had resulted in two distinct schools of thought. Bodhiruci’s (6th c. CE) school maintained that the foundation of all cognition is the ālayavijñāna, a view presented in the Mahāyānasaṃgraha. Ratnamati’s (6th c. CE) school, on the other hand, made the same claim for suchness (tathatā), thus betraying its allegiance to the tradition of the Mahāyānasūtrālaṃkāra. The critical point of divergence was whether the ālaya[[[vijñāna]]] was considered to be the ultimate neutral basis of human reality (à la Bodhiruci) or to be a conditioned entity that must be fundamentally transformed (i.e., eliminated) if the goal of buddhahood is to be realized (à la Ratnamati). Ratnamati’s school and the late Yogācāra exegesis of Asaṅga provided doctrinal support for Paramārtha’s controversial claim that the foundation of all cognition is not the ālayavijñāna but the amalavijñāna. By the seventh century, the controversy gradually died down under the authority of Xuanzang (ca. 603–664) who came down on the side of Bodhiruci in positing the ālayavijñāna as fundamental.
Erich Frauwallner has additionally pointed to numerous Indian antecedents of this dispute that are symptomatic of an underlying tension that could not be so neatly divided along sectarian lines. Despite attempts by the Chinese schools to trace their views to Indian antecedents in the schools of Dharmapāla (530–561) in the case of ālayavijñāna, and Sthiramati (475–555) in the case of amalavijñāna, an analysis of their works does not render support for such clear affiliations. Rather, it points to deep dialectical tensions of a more systemic and perennial nature. As a case in point, Frauwallner cites the following summary of a tension between developmental and disclosive soteriological models by Sthiramati himself in his Madhyāntavibhāgaṭīkā (my translation): The dharmakāya of the buddhas consists in the transformation of the basis in that all obscurations are eliminated and the seeds of the uncontaminated dharmas [i.e., buddha qualities] that function as their counteragents are accumulated; it has power over all phenomena and is without the ālaya[[[vijñāna]]]… Others, on the other hand, say that it is only the dharmadhātu, completely purified through the removal of all adventitious defilements, calling the embodiment (kāya) of the nature of the phenomena (dharmatā) dharmakāya.
Though both accounts construe the transformation of the basis as a process whereby the ālayavijñāna is eliminated, they interpret this process of goal-realization quite differently. The first interprets it developmentally, namely as the accumulation and ripening of “seeds” of uncontaminated buddha qualities, which in turn counteract obscurations (i.e., seeds of contaminated phenomena) and thus lead to their final elimination. The second interprets goal-realization disclosively, namely, as the revelation of dharmadhātu – that which embodies the very nature of things (dharma[tā]kāya) – which occurs as the adventitious defilements that shroud it are purified away.57
Paramārtha’s view of mind seemed to have gained little traction in Tibet and was not much discussed until the 14th century. This despite the fact that his views were available to Tibetan scholars early on through its critique by the Korean monk Wŏnch’ŭk (613–696) in his extensive commentary on the Saṃdhinirmocana. During the Tang dynasty, the commentary was translated in Dunhuang from Chinese into Tibetan (under the title Dgongs ’grel gyi ’grel chen) by one Chos grub (Chinese: Facheng). When Paramārtha’s analysis of mind and his controversial concept of an immaculate consciousness (amalavijñāna) are eventually taken up in Tibet, they appear to have met mainly with a critical reception, first by Mchims ’Jam dbyangs (d. 1267) and Red mda’ ba Gzhon nu blo gros (1349–1412), and thereafter by Dge lugs pa scholars such as Tsong kha pa Blo bzang grags pa (1357–1419), as well as a number of his later commentators such as ’Jam dbyangs Bzhad pa’i rdo rje (1648–1721/22), Gung thang Dkon mchog bstan pa’i sgron me 57 This view, as Frauwallner notes, interprets the expression dharmakāya as deriving from dharmatākāya (“embodiment of the nature of phenomena”) by dropping the suffix tā (Frauwallner 1951: 159, n. 3).
(1762–1823), Blo bzang ’Jam dbyangs smon lam (18th c.), and Blo bzang Dam chos rgya mtsho (1865–1917). While Tsong kha pa, in his early Yid dang kun gzhi’i dka’ ba’i gnas rgya cher ’grel pa, defends the ālayavijñāna as a valid doctrine (a position he would later abandon), he explicitly rejects Paramārtha’s introduction of a ninth consciousness on the rationale that “if there were a fundamental (gtso bo) consciousness other than the ālayavijñāna, it would be a permanent entity (rtag pa’i dngos po : nityabhāva).” But given that entities are by nature impermanent, he argues, the concept of amalavijñāna is self-contradictory and in any case unverifiable. Thus, the Dge lugs pa repudiate amalavijñāna on the same grounds that they reject positive conceptions of tathāgatagarbha: both are adjudged to be metaphysical postulates, reified abstractions, that cannot withstand critical assessment. To recapitulate, we have seen that the Tibetan reception of Indian Buddhism was marked from the outset by the kinds of deep doctrinal tensions between developmental and disclosive soteriological paradigms whose lines of influence in India and China we have been tracing. It was also marked by parallel tensions between differentiation and identification models as scholars sought to reconcile a complex variety of Buddhist ideas concerning conditioned and unconditioned modes of cognition and reality.
We are finally prepared to consider the Eighth Karma pa’s substantial efforts to clarify and explain the relationship between tathāgatagarbha and ālayavijñāna. In a nutshell: the author adopts the differentiation model to the extent that he insists upon drawing an unequivocal distinction between buddha nature and the substratum consciousness. Further, he considers scriptural passages equating the two to have a merely provisional meaning, as further interpretation is deemed necessary. At the same time, however, he advocates, with certain qualifications, the identification model of goal-realization when it comes to clarifying how the ultimate, buddha nature, permeates the mind streams of beings in bondage. Let us now consider how he coordinates these viewpoints. Mi bskyod rdo rje’s distinction between the ālayavijñāna and buddha nature, or the nature of mind, may be seen as part of a broader attempt by his tradition to integrate Yogācāra psychology into a unified buddha nature theory that can accommodate, but also contextualize, the differentiation and unity models. The result is a synthesis that accords primacy to buddha nature and the nature of mind, while at the same time allowing for provisional accounts of the organic, teleological maturation in sentient beings of the qualities characteristic of a buddha.
It must be reiterated at the outset that the Eighth Karma pa considered the identification of ālayavijñāna and tathāgatagarbha to be of merely provisional meaning (neyārtha), geared as it was toward certain Cittamātra followers who, on account of their idealist bias, were inclined to reify dynamic psychological processes. In his Single Intent Commentary (Dgongs gcig kar tīg) on ’Jig rten gsum mgon’s Single Intent, vajra precept 8.36, the Karma pa declares that ālayavijñāna theory “was postulated by Cittamātra proponents as a basis for karma and [its] results.” He adds that “[the ālayavijñāna] was posited in a provisional sense in order to ward off the danger of a view of nihilism, though it does not exist, even in the context of correct conventional reality (tha snyad bden pa).” If it did exist, he argues, “one would have to experience it independently of the cognitions of the six senses. But it is precisely because it is not so established that Candra[kīrti] noted that anyone who says that the ālaya[[[vijñāna]]] exists is not fit to be taught emptiness [and] explained it as being ‘incorrect.’”68 In his Nerve Tonic for the Elderly (hereafter Tonic), the Karma pa had similarly observed that “there were some instances where the Bhagavān indicated the ālayavijñāna by means of the term garbha in order to graciously take on board Mind Only proponents.”
teachings.” MA (ed. Xuezhu Li 2012) evaṃ hi gambhīratarān padārthan / na vetti yas taṃ prati deśyam // asty ālayaḥ pudgala eva cāsti / skandhā ime vā khalu kevalāś ca // Tib. D 3861, 206a5‒6: kun gzhi yod cing gang zag nyid yod la // phung po ’di dag ’ba’ zhig nyid yod ces // bstan pa ’di ni de ltar chos zab don // rig par mi ’gyur gang yin de la’o // See also Mi bskyod rdo rje’s remark in KNVV vol. 1, 212‒213 that “Candrakīrti and others say that there is no substratum cognition that is commonly established as an empirical experience (dmigs myong), even conventionally.” slob dpon chen po zla ba grags pa la sogs kyis kun gzhi’i shes pa tha snyad du’ang ’jig rten gyi dmigs myong gi grags grub tu med par gsungs pa’i phyir /
68 The relevant passage in Dgongs gcig kar ṭīg V.2, MDSB vol. 6, 9026–9032, reads: “[’Jigs rten gsum mgon] did not teach a theory like that of the Cittamātra proponents [wherein] the ālayavijñāna is postulated as a basis for karma and [its] results. Rather, the substratum consciousness (ālayavijñāna) was posited in a provisional sense in order to avert the danger of a view of nihilism, though it does not exist, even in the context of correct conventional reality (tha snyad bden pa). If it did exist, then one would have to experience it independently of the cognitions of the six senses. But it is precisely because it is not so established that Candra[kīrti] noted that anyone who says that the ālaya[[[vijñāna]]] exists is not fit to be taught emptiness [and] explained it as being ‘incorrect.’” kun (gzhi rnam para) shes pa ni // sems tsam pa yi las ’bras rten // ’jog byed rnam gzhag de ’dra ba // ma gsungs chad par lta ba’i nyenb // spangc phyir drang ba’i don du bzhag // yang dag tha snyad bden par yang // kun gzhi’i rnam shes yod min te // yod na tshogs drug shes pa las // gzhan du nyams su myong dgos na // ma grub de phyir zla bas kyang // gang dag kun gzhi yod ces pa // stong nyid bstan par mi ’os la // a GCMD gyi rnal ’byor; em. as per GCKL and GCBC; b GCMD om. // and addit. yang dag kun gzhi yod ces ’di dpe gnyis gang dag; this appears to be an annotation (though the note is not distinguished in any way from the main text); c GCMD spong em. as per GCKL and GCBC.
Despite his reservations about the hypostatization of ālayavijñāna, the Karma pa did acknowledge its heuristic value in accounting for problems of mental causality and continuity on the level of conventional appearances. Furthermore, he did not deny the validity of inferring the operation of largely unconscious constitutive processes (latent tendencies) that continually condition and structure thought and behaviour, even if one could not purport to know anything whatsoever about their assumed location and nature. It is clear, then, that Mi bskyod rdo rje viewed the ālayavijñāna as a useful explanatory concept and model, but also as a hypothetical construct to which, strictly speaking, no thing corresponds, since one searches in vain for any autonomously existent entity corresponding to the term.
This latter strain in his thinking helps to explain why the Karma pa could at times heartily endorse Candrakīrti’s thesis that the ālayavijñāna is an untenable postulate. A notable instance is a passage from the author’s Treasury Containing the Wealth of Profound Mahāmudrā (Zab mo phyag chen gyi mdzod sna tshogs ’dus pa’i gter) in which he goes so far as to characterize the “no ālayavijñāna” thesis as being “more intellectually refined” in “the context of deeply investigating the ultimate” than accounts accepting its existence, which were endorsed by no less an authority than the Third Karma pa: (garbha) in order to graciously take on board Mind Only proponents.” bcom ldan ’das kyis sems tsam pa dag rjes su bzung ba’i phyir / kun gzhi rnam shes la snying po’i sgras bstan pa zhig yod pa la dgongs pa …
When the Bodhicittavivaraṇa, Madhyamakāvatāra, and other texts explain that the ālayavijñāna is untenable, they proceed to explain that mere appearances are [due to] latent tendencies alone. As for the exegesis of both the noble Ācārya [[[Nāgārjuna]]] and Candrakīrti, the reason they did not accept the ālayavijñāna is that all phenomena are entirely devoid of any factor that is self-sufficient in terms of function and essence. That being so, since [this ālayavijñāna] would have to be an independently existing consciousness capable of serving as the basis of all phenomena, [despite its being] an obscuration that shrouds dharmadhātu [while itself being] indeterminate, it was rightly rejected. Nonetheless, according to some other Ācāryas, the Victorious [[[Buddha]]] taught the classifications of skandhas, dhātus, and āyatanas in order to invalidate non-Buddhists’ beliefs in a self, a creator, and a consumer. In particular, in the case of explanations according special status to the ālayavijñāna as expounded in the Laṅkāvatāra and so on, it is evident that [these texts] explained very eloquently the criteria of cause and effect in the context of establishing appearances as mind. [This account] was also extolled by the illustrious Rang byung who followed this later tradition. But for me, in the context of deeply investigating the ultimate, the former tradition appears to be [more] intellectually refined.
To understand the import of this analysis, it is helpful to bear in mind the contrasting hermeneutical paradigms Mi bskyod rdo rje was working with. To start with, the Third Karma pa Rang byung rdo rje’s alleged espousal of the ālayavijñāna is only comprehensible in light of his adherence to the Yogācāra differentiation model – a model that strongly shaped the Eighth Karma pa’s views of mind and buddha nature as well.
Let us now briefly review some precedents for Mi bskyod rdo rje’s distinction between tathāgatagarbha and ālayavijñāna in the work of his Karma Bka’ brgyud predecessors. In his Profound Inner Meaning (Zab mo nang gi don) auto-commentary, the Third Karma pa Rang byung rdo rje had influentially equated the ālayavijñāna with impure mind (sems ma dag pa) and sharply distinguished it from pure mind (sems dag pa), which he equated with buddha nature. He added that “the general discourses of all vehicles refer to mind as such (sems nyid) but this should be known to be two-fold: possessing purity and being impure.” He proceeds to equate the mind possessing purity with  mind as such (sems nyid), as extolled in Saraha’s Dohākoṣagīti (DKG) 43 as the seed of all of saṃsāra and nirvāṇa,  the (innately) pure mind (cittaśuddhi) as described in Ratnagotravibhāga I.55–57 by analogy with space, which supports the other elements but is itself unsupported, and  mind’s luminous nature as it is defined in Aṣṭasāhasrikāprajñāpāramitā 5b1–2.
In a similar vein, the Third Karma pa further distinguished the mundane mind – the ālayavijñāna with its eightfold consciousness (kun gzhi tshogs brgyad) – from the supramundane mind (’jigs rten las ’das pa’i sems : lokottaracitta), buddha nature. This distinction is made both in his Profound Inner Meaning with reference to Mahāyānasaṃgraha 1.45‒48 and in his Dharmadhātustava commentary on stanza 46ab, where he states that mind is observed as having two aspects, mundane and transmundane.79 We previously mentioned Sthiramati’s similar distinction, by way of commentary to Triṃśikā 29‒30, between the substratum consciousness (ālayavijñāna) and the supramundane wisdom (lokottarajñāna) which overturns or replaces (parāvṛtti) it. The Third Karma pa’s distinction between pure and impure minds is further developed by his successors, most notably in the extensive Profound Inner Meaning commentarial literature. For example, Mi bskyod rdo rje’s teacher Karma phrin las suggests in his Profound Inner Meaning commentary (dated 1509 in the colophon) that although buddha nature, which he equates with the nature of mind (sems kyi rang bzhin) and substratum wisdom (kun gzhi ye shes) – a term originally coined by Dol po pa and widely adopted by Tibetan scholars – appears to be blended with the ālayavijñāna like milk in water, it may (as the Indian analogy suggests) be separated from it by the wise, just as milk is extracted from water by the mythical goose:
The substratum wisdom is buddha nature as explained above. This is precisely what is meant by “the nature of mind” in the Prajñāpāramitā and the Uttaratantra (RGV), “the mind that is like a wish fulfilling gem” in the Dohā, and “the beginningless element as the basis of all phenomena” in the Abhidharmasūtra.83 Here, we describe it as wisdom. It abides in the substratum consciousness in a blended manner, like water and milk. 79 Chos dbyings bstod pa’i ’grel pa, 611ff. which comments on Dharmadhātustava (DDhS) 46ab: Skt. cittam evaa dvidhā dṛṣtaṃb loki lokottaraṃ ca yat / Tib. sems nyid rnam pa gnyis su mthong / ci ltar ’jig rten ’jig rten ’das // a Ms eta em. as per D: nyid; b Ms dṛṣta em. as per D: mthong. Therefore, those being ignorant regarding the definitive meaning have not recognized the substratum wisdom.85 In one of his “Question and Answer” (dris lan) texts, Karma phrin las makes the additional observation that the term “substratum wisdom” does not imply that the substratum and wisdom are the same but rather that wisdom itself resides within the substratum. With this interpretation he can claim that the wisdom present in the ground, which is equated with natural luminosity, the purity of mind, and the indestructible nucleus (mi shigs pa’i thig le), is the substratum simpliciter (kun gzhi tsam), serving as the ground for both saṃsāra and nirvāṇa without itself being identifiable with either. This substratum simpliciter is distinct from the substratum consciousness, which “functions as a ground for the unfoldment of worldliness” but “is unable to serve as a ground for nirvāṇa.” a mixture of water and milk in one and the same vessel; geese drink but the milk while the water remains as it is. Likewise in the case of a mixture of afflictions and wisdom in one and the same vessel, the yogins drink the wisdom, leaving ignorance behind.” Ed. Liu Zhen 2015: 20: yathodakena sammiśraṃ kṣīram ekatra bhājane / kṣīram pibanti haṃsā hi udakaṃ ca tathā sthitam // [52/Tib. 62] evaṃ hi kleśasammiśram jñānam ekatra bhājane / pibanti yogino jñānam ajñānam sphorayanti te // [53/Tib. 63].
85 Zab mo nang don gyi rnam bshad snying po, in RDSB vol. 14, 601‒4: de la kun gzhi’i ye shes ni / gong du bshad pa’i bde bar gshegs pa’i snying po ste / de yang sher phyin dang rgyud bla ma las / sems kyi rang bzhin du gsungs pa dang / do hā las / yid bzhin nor ’dra’i sems su gsungs la / chos mngon pa’i mdo las / thog ma med pa’i dus kyi dbyings / chos rnams kun gyi gnas yin te / zhes gsungs pa yang don ’di nyid do / de la ye shes su bshad de / de yang kun gzhi’i rnam shes la chu dang ’o ma bzhin ’dres pa’i tshul du gnas pas / nges don la rmongs pa rnams kyis kun gzhi’i ye shes ngos ma zin par / In a slightly later Profound Inner Meaning commentary (dated 1514), Dwags ram pa Chos rgyal bstan pa (1449‒1524) equates Rang byung rdo rje’s pure mind (dag pa’i sems) with the pure substratum wisdom (dag pa’i kun gzhi ye shes), as well as the causal continuum (rgyu rgyud) of tantrism. He goes on to distinguish it from the ālayavijñāna qua impure mind (sems ma dag pa’i kun gzhi rnam shes). Citing Mahāyānasaṃgraha (MS) I.45‒48 in support of this view, he further notes that “this Mahāyānasaṃgraha text specifically characterizes the ālayavijñāna as the basis of sentient being (sems can gyi gnas), but specifies that is not the cause of nirvāṇa.” But if this is the case, his interlocutor asks, then what generates the qualities of purification (vyavadānadharma)? To this Dwags ram pa replies “the entire range of buddha qualities of purification depends on the substratum wisdom (kun gzhi ye shes), the aforementioned pure mind.” He concludes the discussion by criticizing “certain Sa skya Lam ’bras followers who, having neither seen nor heard the
having defilements, they are buddhas, but not perfectly realized buddhas. Although the substratum and wisdom are not the same, there is not the slightest fallacy of contradiction in explaining that the indestructible nucleus is the ground of saṃsāra and nirvāṇa.” de dag sbyin sogs dkar chos dang ’brel bas // rnam par shes pa’i chos nyid las ’das shing // ye shes gnyis su med pa’i tshul ’chang ba // dri med chos kyi dbyings kyi rgyu mthun pa’i // sems kyi dag pa rang bzhin ’od gsal nges // bla med rdzogs byang chub pa’i gzhi byed phyir // kun gzhi ye shes zhes gsung de yis ni // ’khor bar ’khor ba’i gzhi mi byed do // bag chags kun dang ldan pa’i kun gzhi la // kun gzhi’i rnam par shes pa zhes bya ste // des ni srid pa ’phel ba’i gzhi byed kyang // mya ngan ’das pa’i gzhi ru mi rung ngo // kun gzhi tsam ni gzhi la bzhugs pa yi // ye shes nyid yin de phyir sems can rnams // dri mar bcas pa’i sangs rgyas yin par ’dod // ’khor dang myang ’das kun gi gzhi yin kyang // ’khor ’das gnyis ka yin par mi ’gyur ba // ’di ni bsam gyis mi khyab pa yi gnas // kun gzhi ye shes yod par khas len kyang // kun gzhi ye shes gcig par mi smra mod // dri bcas sangs rgyas yin phyir sems can rnams // sangs rgyas yin kyang rdzogs pa’i sangs rgyas min // kun gzhi dang ni ye shes mi gcig kyang // mi shigs thig le ’khor ’das kun gyi gzhir // bshad la ’gal ba’i nyes pa rdul tsam med // The Ngo gro bla ma in the title likely refers to Ngo khro rab ’byams pa Dbang phyug dpal, see Higgins and Draszczyk 2016: vol. 1, 194, n. 565. above-cited Mahāyānasaṃgraha passages, assert that the ālayavijñāna (kun gzhi rnam shes) is the ‘causal continuum substratum’ (kun gzhi rgyu rgyud), thus putting on display all of their hidden flaws.”90
Such developments certainly helped shape the Eighth Karma pa’s own attempts to clarify the relationship between buddha nature and ālayavijñāna. In his One or Two Minds? A Reply to Bla ma Khams pa (Bla ma khams pa’i dris lan mi gcig sems gnyis), for example, he draws a parallel distinction between innate mind (gnyug ma’i sems) and adventitious mind (glo bur gyi sems), equating the former with buddha nature and the latter with ālayavijñāna.91 As scriptural support, he cites Rang byung rdo rje’s statement in the Profound Inner Meaning auto-commentary that mind has been explained both in terms of pure and impure modes. The Eighth Karma pa then specifies that the pure mode is underscored in the classification in Ratnagotravibhāga I.47 between three phases of the buddha element: impure, pure-impure, and completely pure. The pure mode, he continues, refers to the unobscured self-aware wisdom (sgrib bral rang rig pa’i ye shes), whereas the impure mode refers to the obscured consciousness consisting in deluded ignorance (sgrib bcas rmongs pa ma rig pa’i rnam par shes pa). Like Rang byung rdo rje and several of his commentators, Mi bskyod rdo rje recognized that the term “substratum” had been used with notably different, and at times divergent, connotations in Buddhist sources, and 90 Zab mo nang don sems kyi rnam par thar pa’i gsal ba’i rgyan, in RDSB vol. 12, 1143‒4: drangs ma thag pa’i theg bsdus kyi lung snga phyi gnyis po ma mthong zhing ma thos pa’i lam ’bras pa kha gcig kun gzhi’i rnam par shes pa nyid kun gzhi rgyu rgyud du khas len pas ni rang gi nang mtshang thams cad ngom par byed pa ste … 91 See Bla ma khams pa’i dris lan mi gcig sems gnyis, MDSB vol. 3, 2215–2222.
therefore required careful analysis and clarification. Rang byung rdo rje had observed that “the term kun gzhi (ālaya), when it is used independently of the expression rnam par shes pa (vijñāna), is not necessarily [a shorthand] for kun gzhi rnam par shes pa (ālayavijñāna) but can also refer to suchness (tathatā : de bzhin nyid).”94 This point is later reiterated by Kong sprul Blo gros mtha’ yas (1813‒1899) when he notes that ālaya is combined with vijñāna in the compound ālayavijñāna in order to distinguish it from instances where buddha nature and suchness (tathatā) are described as substrata (kun gzhi). Adding his own clarification, Mi bskyod rdo rje contends in his Direct Introduction to the Embodiments Commentary (hereafter Embodiments) that it is necessary to posit the naturally luminous buddha nature, which has been termed an ālaya (Tib. kun gzhi, lit. “all-ground”), as the common ground of buddhas and sentient beings. “On the one hand,” he explains, it is due to [[[beings]]] having this cause, buddha nature, that its result [[[buddhahood]]] is brought forth. On the other hand, it is due to the influence of all the adventitious defilements which obscure or obstruct [[[buddha nature]]] that all phenomena of saṃsāra and nirvāṇa occur by way of dependent arising. If this [[[buddha]]] nature did not exist, then the conventional arising of saṃsāra and nirvāṇa, bondage and liberation, and so on would not exist. 94 Zab mo nang don gyi ’grel pa, in RDSB vol. 7, 3832: ’di yang kun gzhi zhes bya ba la rnam par shes pa’i sgra ma smos na de bzhin nyid la yang kun gzhis brjod du rung ba’i phyir rnam par shes pa smos so // The Eighth Karma pa adds that one must nonetheless distinguish this abiding buddha nature from the discontinuous substratum consciousness. “The substratum [[[consciousness]]] is not perpetually continuous (rgyun brtan pa) since it comes to an end once the karmic seeds aspect [ceases on] the eighth level and the karmic maturation aspect [ceases on] the ninth level.” By contrast, buddha nature “is perpetually continuous since it neither waxes nor wanes from sentient beings up to buddha. Thus, it was posited as the ground of all phenomena comprising bondage and freedom, saṃsāra and nirvāṇa, the innate and adventitious, and the two truths.”
Mi bskyod rdo rje concludes that this ever-present and all-pervading buddha nature is available as a “cause” for realizing the goal of buddhahood precisely because it is a condition of possibility for all phenomena subsumed under the two realities. Prima facie it seems difficult to square the Eighth Karma pa’s conception of buddha nature as a universal substratum comprising saṃsāra, nirvāṇa, and the path with his repeated admonitions not to conflate unconditioned buddha nature with adventitious saṃsāric phenomena. He was certainly not the first to face the problem of reconciling these two quite different Buddhist theses concerning the relationship between unconditioned buddha nature and conditioned phenomena:  an independence thesis specifying how unconditioned buddha nature is independent of all conditioned phenomena comprising saṃsāra and der bzhag pa yin te / sangs rgyas kyi snying po’i rgyu ’di yod pa’i dbang las / ’bras bu de bskyed pa la sgrib byed dam gegs byed kyi glo bur gyi dri ma thams cad kyang ltos pa’i dbang gis ’khor ’das kyi chos thams cad rten cing ’brel par ’byung ba’i tshul gyis ’byung ba’i phyir / snying po ’di med na ’khor ’das ’ching grol sogs tha snyad du’byung ba med par ’gyur la / ’di yod pas de lta’ang ma yin te … In other words, soteriological conventions such as saṃsāra and nirvāṇa, or bondage and liberation, are predicated on the possibility of freedom from the shackles of ignorance and delusion; and buddha nature is precisely the condition of this possibility.
nirvāṇa and the path;  a dependence thesis specifying how all such conditioned phenomena depend for their existence upon the unconditioned, because the latter is the very condition of their possibility. While the first takes buddha nature as a soteriological substratum – the condition of possibility of liberation and nirvāṇa but not of saṃsāra, the second more broadly construes it as a phenomenal substratum – the condition of possibility of all phenomena, saṃsāra and nirvāṇa and the path. The notion of a basic substratum that unifies both the processes of cyclic existence (saṃsāra) and of liberation from it (nirvāṇa) fulfills the demand for some principle of continuity in a system that otherwise rejects the existence of any patient of suffering or agent of liberation that are identifiable as a self. A recurrent tension between the independence and dependence accounts is discernable in Mi bskyod rdo rje’s treatments of the relationship between buddha nature and the substratum. A variant of the independence thesis is defended in a general outline of the Single Intent (dgongs gcig) philosophy within the Karma pa’s Single Intent Commentary wherein he explicitly rejects the view that mahāmudrā and tathāgatagarbha can be identified as the source of both saṃsāra and nirvāṇa. In doing so, he provides a valuable overview of differing conceptions of the phenomenal and soteriological substrata (kun gzhi : ālaya) that figure in Bka’ brgyud exegesis:
The way in which mahāmudrā does not function as a basis for all of saṃsāra and nirvāṇa requires explanation. In the Mahāyāna tradition, the presentation of the substratum (ālaya) is explained as having three [features]:  [[[karmic]]] ripening,  [[[karmic]]] seeds and  the uncorrupted element (zag med khams). The first two are said to function as the foundation of saṃsāra. The third is the cause of nirvāṇa; being the extraordinary distinct set of six cognitive domains which functions as the basis of the unfolded potential and the like, it is described as the natural outflow of dharmakāya. The way in which the ālaya is a basis of saṃsāra [comprises both]:  what is based on it by way of [[[karmic]]] seeds for any of the [three] realms, as the predispositions for their emergence, and  what is based on it by way of [[[karmic]]] ripening as the three sufferings of the three realms and so forth.  The progressive awakening of latent tendencies of studying, thinking, and meditating and so on is described in terms of the uncorrupted element that is precisely the producer and produced of nirvāṇa. Hence, there are some for whom this ālaya presented as the basis for saṃsāra and nirvāṇa functions [also] as the foundation of mahāmudrā. [But] this was not the intent of the Lord of Sages. The mahāmudrā of Mantra[[[yāna]]], the tathāgatagarbha of the final turning, and the prajñāpāramitā of the middle turning and so forth are special methods of revealing the single intent. Among these, the nature of the perfection of wisdom (prajñāpāramitā) is not a foundation for either saṃsāra or nirvāṇa because [this] nature has always been beyond the whole tangle of conceptual elaborations, such as saṃsāra and nirvāṇa. With this passage, the Eighth Karma pa sharply distinguishes the substratum of karmically-conditioned processes of karmic seeds and their maturation (saṃsāra) from the substratum of the uncorrupted element (anāsravadhātu), which is equated with tathāgatagarbha, mahāmudrā, and prajñāpāramitā. Stated succinctly, tathāgatagarbha (qua mahāmudrā) cannot be regarded as a source of saṃsāra and nirvāṇa, or of any other phenomena designated by reciprocally determined constructs of this kind, because its nature eludes appropriation by conceptual thought.
The dependence thesis, emphasizing the pervasion of the conditioned by the unconditioned, is outlined in the above-mentioned passage from Embodiments in which buddha nature is identified as the substratum of saṃsāra, nirvāṇa, and the path. Mi bskyod rdo rje proceeds to address the question “If that buddha nature is not the substratum consciousness, which is of the essence of the mind of adventitious defilements, then why has it been described in that way?” He replies that “since that [[[buddha nature]]] is the root of all phenomena comprising pure and impure substrata etc., it is not inconsistent to explain it in that way.” To support this point, he quotes the Ghanavyūhasūtra: The substratum of the various levels, That is also the goodness that is buddha nature. The tathāgatas have indicated this nature (garbha) By means of the term “substratum” (ālaya). Although the garbha has been declared to be the ālaya,
Commenting on this same passage in his Single Intent Commentary, Mi bskyod rdo rje explains that “the term ‘substratum of various levels’ was described as a substratum with reference to all seeds and causes of saṃsāra and nirvāṇa.” He adds, however, that “when the cause of nirvāṇa is referred to as a ‘substratum’ (ālaya), it is not possible that this could [signify anything] other than buddha nature. This is because the ālaya[[[vijñāna]]] is the malaise (dauṣṭhulya) of defilement, whereas this [[[buddha nature]]] is precisely the natural luminosity, which is not the malaise of defilement.” This brings us full circle from the phenomenal substratum back to the soteriological substratum. And once again it is evident that the independence and dependence theses can only be reconciled within a broader contextualist perspective of the Buddhist path. In short, although buddha nature may indeed be regarded in an inclusive sense as a substratum or precondition of all causes and all phenomena, conditioned and unconditioned, the aspirant must nonetheless distinguish within its scope between pure and impure substrata – the unconditioned buddha nature and conditioned substratum consciousness – in order to avoid conflating the two. In any case, when viewed in light of one another, the independence and dependence accounts of how the tathāgatagarbha relates to the ālayavijñāna equally accentuate the abiding and fundamental nature of the former and contingent and superfluous nature of the latter. In this regard, they may be seen as two aspects of the differentiation model, the former stressing the sublime otherness (gzhan mchog) of buddha nature, the latter stressing its all-pervasiveness in sentient beings. Their common focus on the unreality of the ālayavijñāna underscores the disclosive standpoint: it is only with the dissolution of the conditioned and conditioning substratum consciousness that buddha nature or substratum wisdom can fully reveal itself.
At various points, the Karma pa explores the deleterious consequences, both exegetical and practical, that may result from confusing these pure and impure substrata or, more specifically, buddha nature and the substratum. In the Tonic, Mi bskyod rdo rje criticizes Shākya mchog ldan’s identification of buddhagarbha with the ālayavijñāna on the grounds that the roles these two ideas play in Buddhist soteriology are diametrically opposed. Traditionally, the ālayavijñāna was posited as a repository (“storehouse”) of karmic tendencies and their maturation; buddha nature was posited as the potential for spiritual awakening as the result of dispelling karmic tendencies. In this regard, the identification of buddha nature and the substratum consciousness leads to the absurd consequence that the former, even in the phase of goal-realization – where it is traditionally held to be purified of all karmic conditioning – continues to store karmic tendencies and gives rise to their results. In response to Shākya mchog ldan’s claim that “saṃsāric latent tendencies deposited by conventional karma arise in this buddha nature,” Mi bskyod rdo rje states:
Since you assert that buddha nature is a functional surrogate103 of naturally pure awakening, it absurdly follows [from your claim] that this surrogate of naturally pure awakening in oneself and others is a warehouse manager who stores a whole range of latent tendencies of virtuous and nonvirtuous deeds of sentient beings as well as the warehouse [itself]. And, were it possible that [[[buddha nature]]] could be contaminated at all, then even in its state of fruition – i.e., buddhahood free from defilements which you yourself accept – it would still store the latent tendencies of the karma of all sentient beings. This is because, in general, the dharmadhātu and buddha nature would be what stores the latent tendencies of that karma. If these [[[Wikipedia:tendencies|tendencies]]] are stored by it, it stands to reason that even buddha nature in the phase of fruition would store and give rise to these karmic latent tendencies. Elsewhere in the text, the Karma pa takes particular issue with Shākya mchog ldan and his students for not properly distinguishing within the universal substratum or ground proper between its pure and impure substrata. He begins by noting that
if the dharmadhātu is taken as the basis of adventitious defilements, then you need to clearly distinguish between the substratum wisdom (kun gzhi ye shes) and [[[substratum]]] consciousness ([[[kun gzhi]]] rnam shes). If you don’t distinguish them, then it is inappropriate if you explain the 103 The term go chod po here combines two related senses:  facilitator or functionary: something able to perform any function, or fulfil any task, required of it, for example a machine used in manufacturing certain products, and  surrogate or proxy: something able to perform the function of, and thus stand in for, something else.
Mi bskyod rdo rje proceeds to underscore the need to respect the semantic ranges and functional roles of context-specific terms such as “substratum consciousness” and “substratum wisdom” in order to avoid serious confusions in thought and meditation:
Consequently, when anyone contends that it is necessary to accept [this] ālayavijñāna, which is the basis of adventitious defilements, it follows that it is inadmissible to then introduce within that ālaya a distinction between the pure and impure. This is because were it possible of that which is termed buddha nature (*sugatagarbha) or dharmadhātu or substratum wisdom (kun gzhi’i ye shes) to function as the basis for the arising of adventitious defilements, then there would not be any role left (mgo bde ma byung ba) for the ālayavijñāna to be the basis of such [[[defilements]]]. Moreover, among you and the teachers in your lineage, there is not even one who has penetrated this matter deeply. Some assert that the clarity aspect in the context of the substratum consciousness is the substratum wisdom. Some assert that the clarity aspect that is the intrinsic essence of the substratum consciousness is not conducive to nirvāṇa since it does not transcend saṃsāra. Some claim that saṃsāra manifests in that clarity aspect which is the substratum wisdom or [[[buddha]]] nature. Therefore, you masters and disciples – is nirvāṇa the clarity aspect of the substratum consciousness or is the substratum consciousness the clarity aspect of the substratum wisdom? Masters and disciples, you must give up this ignoble talk! 108 Rgan po’i rlung sman, MDSB vol. 15, 10205–10211: chos dbyings kyi glo bur gyi dri ma’i rten byed na / khyod cag kun gzhi ye shes dang rnam shes gnyis ’byed dgos la / mi ’byed na glo bur dri ma’i rten du gyur pa’i kun gzhi de ye shes dang bde gshegs snying po la ’chad na mi rung bas / glo bur dri ma’i rten kun gzhi’i rnam shes zhig cis kyang khas len dgos zer nas kun gzhi la dag ma dag gnyis kyi dbye ’byed byed pa de mi ’thad par thal / Just as Mi bskyod rdo rje had, in an earlier part of this text, warned about the “collapse of all terminological conventions” that results from conflating buddha nature with its adventitious defilements, he here warns against the confusions that may arise from conflating aspects of the substratum wisdom, i.e., buddha nature, with aspects of the substratum consciousness.
This analysis forms part of a broader critique of the epistemological foundations of the tantric buddha nature theory outlined in Shākya mchog ldan’s Cakrasaṃvara Commentary (bde mchog rnam bshad). There, Mi bskyod rdo rje contends that the Sa skya scholar’s tendency to blur the lines between consciousness (rnam shes) and wisdom (ye shes) weakens the entire edifice of his buddha nature theory. Specifically, Shākya mchog ldan is accused of equating the clear and knowing cognition – the subjective, inward-looking part of consciousness – with nondual wisdom, and of thus aligning the outward-looking (objective) and inward-looking (subjective) poles of consciousness with the two truths, the conventional and ultimate respectively. For the Karma pa, this model of consciousness reflects his opponent’s allegiance to an alīkākāravāda (False Aspectarian) Cittamātra view, which equates the apprehending aspect of cognition ’byed byed pa de mi ’thad par thal / bde gshegs snying po’am chos dbyings sam kun gzhi’i ye shes kyi ming can de nyid kyi glo bur gyi dri ma ’char ba’i rten du rung ba gang zhig de rung na de’i rten la kun gzhi’i rnam shes kyi mgo bde ma byung ba’i phyir / gzhan yang khyod dpon slob brgyud pa dang bcas pa la rnam rtog gting tshugs pa gcig kyang med par / res kun gzhi rnam shes kyi steng gi gsal cha de kun gzhi ye shes su khas len / res kun gzhi rnam shes kyi rang ngo’i gsal cha ’khor ba las mi ’da’ bas myang ’das su mi rung bar khas len / res kun gzhi ye shes sam snying po’i gsal cha de la ’khor ba ’char zer / des na khyed rang dpon slob rnams kun gzhi rnam shes kyi gsal cha myang ’das yin nam / kun gzhi ye shes kyi gsal cha kun gzhi rnam shes yin dpon slob kha ngan pa gyis la byon zhig / An interlinear note to this passage explains that Shākya mchog ldan maintained in his Bde mchog rnam bshad that consciousness (rnam shes) arises as the clarity aspect (dwangs cha) of wisdom whereas his student Paṇ chen Rdo rgyal ba (a.k.a. Rdo rje rgyal mtshan, b. 15th c.) proclaimed that wisdom is the clarity aspect of consciousness. “Thus the positions subscribed to by these two, master and disciple, are [as] opposed as East and West.” Little is known about Rdo rje rgyal mtshan but Mi bskyod rdo rje composed a response (in meter) to questions of Rdor rgyal ba. In the colophon, the Karma pa attributes numerous works to Rdor rgyal (none of which are extant) which encompassed the fields of epistemology, Madhyamaka, Abhidharma, Tantra, and “most notably [Rdo rgyal’s] Gzhan stong commentarial text on the Kālacakra.” See Paṇ chen rdo rgyal ba’i legs bshad rnam par dkar ba’i shel gyi glegs bu la drang po’i thig vaiḍūrya’i ri mo btab pa, in MDSB vol. 3, 2573‒4.
with nondual wisdom. Now, as Mi bskyod rdo rje and much of the Indian Buddhist tradition maintain, mundane consciousness (vijñāna : rnam shes) is dualistic precisely on account of its subjectivizing and objectifying activities, whereas wisdom (jñāna : ye shes) is characterized precisely by the absence of such a dualism. Consequently, both the sense and explanatory power of this crucial distinction, which is a cornerstone in Shākya mchog ldan’s own doctrinal system as well, are forsaken when he links the subject pole of consciousness with wisdom and erects an entire soteriology on this shaky foundation. The same line of criticism is applied to Shākya mchog ldan’s attempt to bring the ālayavijñāna into line with buddha nature.
Some of Mi bskyod rdo rje’s most cogent reflections on the relationship between buddha nature and ālayavijñāna are to be found in his attempts to coordinate and clarify the tantric interpretations of these concepts. A lucid summary is given in a section of the author’s Single Intent Commentary in which he comments on ’Jigs rten gsum mgon’s Single Intent, vajra precept 8.36: “Through the power of blessing, the substratum (kun gzhi : ālaya) is actualized in a short time.” In clarifying this precept, the Karma pa offers a valuable explanation of sūtric and tantric views of the substratum and their complex relationship with buddha nature doctrine. Confining our attention to the parts of this section which pertain to the relationship between buddha nature and the substratum, we may begin with the Karma pa’s initial reframing of precept 8.36: If this vajra precept is restated very clearly, it says this: “Through the instructions of one who has perceived that buddha nature of the three continua which is the final intent of the sūtras and tantras [and] which has been given the name ‘substratum’ (ālaya), one is able to actualize it in a short time via the key points. If one is able to do this, then by directly recognizing the subtlest root of saṃsāra [i.e., the ālayavijñāna], which is to be abandoned via the Mantrayāna, one engages in relinquishing it. And when one engages in that, one cannot help but attain the buddha[hood] of the sūtra and mantra traditions.”
In this rather elaborate reworking of the precept, we can pick out several key points that are central to Mi bskyod rdo rje’s own viewpoint. Buddha nature, the final intent of sūtras and tantras, is equated with the tantric continua (rgyud) of ground, path, and fruition, and is also said to be known as a “substratum.” It is this substratum which is actualized in a short time by way of Mantrayāna pith-instructions. These help one to recognize and finally relinquish the subtlest root of saṃsāra, which he later identifies as the substratum consciousness, and thus attain buddhahood of the Mantra tradition. Central to this interpretation is the distinction between the unconditioned substratum (or threefold tantric continuum) and the conditioned and conditioning ālayavijñāna. This distinction underlies Mi bskyod rdo rje’s explanation of the four perfections of buddha nature. [This] nature is  “pure” because [it] does not serve as a basis for latent tendencies;  “true selfhood” because in [its] selflessness, even the conceptual elaborations regarding “no self” have completely subsided;  “bliss” because it is free from body-mind produced by the subtle movement of ignorance; and  “permanent” because the undefiled spiritual element of this kind is the uninterrupted continuity of buddha activities. After quoting Ratnagotravibhāga I.35ab, the Karma pa proceeds to explain that such actualization goes hand in hand with understanding the increasingly subtle roots of saṃsāra that obscure it: “When one actualizes this [[[buddha]]] nature, [one] is able to understand the chaff which obscures it – saṃsāra – and [to understand] not only its coarse root, but also its subtle and subtlest roots.”
mngon du bya nus la / de nus pa na sngags kyi theg pas spang bya’i ’khor ba’i rtsa ba ches cher phra ba de ngo shes nas de spong bar byed pa la ’jug cing / de ’jug pa na mdo sngags lugs kyi sangs rgyas mi thob ka med ’byung bya ba ’di bzhugs / a GCMD mi; b GCMD om. par; c GCMD om. pa; all these are em. as per GCKL. § 10. The ālayavijñāna as the most subtle root of saṃsāra (tantric view) According to the foregoing analysis, ālayavijñāna is the most subtle root of saṃsāra and therefore the most obdurate obstacle to liberation. Mi bskyod rdo rje explains in his Single Intent commentary that for the Buddhist practitioner, this deeply rooted substratum consciousness represents the final barrier standing in the way of awakening once the other roots of saṃsāra – the beliefs in an inner self (of the Vaibhāṣika and Sautrāntika), an outer world (of the pratyekabuddha), and reification in general (Madhyamaka) – have been systematically eradicated. One is then able to confront, through third dharmacakra (i.e., buddha nature) teachings, the subtlest root of saṃsāra, ālayavijñāna, conceived as the repository of latent tendencies known as the “defiled purity of mind.” One can then proceed, through Mantrayāna, to deal with the final roots of saṃsāra, the latent tendencies for rebirth. The root of saṃsāra for Śrāvaka Vaibhāṣikas and Sautrāntikas consists in the personalistic false views (satkāyadṛṣṭi), while for pratyekabuddhas, it is the belief in the reality of objects. Commonly among Mādhyamikas, it consists in elaborations based on reifications of signs. And in the final wheel, it is taken to be the indeterminate ālayavijñāna, construed as the repository of latent tendencies, which is called the “defiled purity of mind.” Although [this conception of a] buddhahood in which all these roots of saṃsāra have been relinquished is discovered on the sūtric path, the roots of saṃsāra explained in the Mantra[[[yāna]]] concern the latent tendencies for transference [via rebirth].
In this regard, Mi bskyod rdo rje suggests that the Mantrayāna path of eradicating all these roots of saṃsāra and attaining buddhahood supersedes the so-called ‘sūtric’ path precisely insofar as the former eradicates the most deeply-rooted latent tendencies – those leading to rebirth, which are associated with the ālayavijñāna. On this point, he distinguishes tantric from sūtric conceptions of the ālayavijñāna: This kind of root of saṃsāra is also designated by the term ālayavijñāna. In this regard, this [[[tantric]]] ālayavijñāna is subtler than the ālaya[[[vijñāna]]] explained in the context of the sūtras. From the Kālacakra [[[Vimalaprabhāṭīkā]]]:
In sum, the complete eradication of roots of saṃsāra, which Mi bskyod rdo rje correlates with the threefold purification of obscurations as outlined in Ratnagotravibhāga I.47, depends on eliminating two successive substrata: “ the aspect which serves as the ground of all afflictions 117 Vimalaprabhāṭīkā (VPṬ), Tib. D 1347, 267a2.
118 Dgongs gcig kar ṭīg V.2, MDSB vol. 6, 8985–7. ’di lta bu’i ’khor rtsa la kun gzhi’i rnam shes zhes kyang brjod de / de lta na’ang ’di’i kun gzhi’i (rnam shesa) ni mdo phyogs nas bshad pa’i kun gzhi’i las ches phra ba’i kun gzhi ste / (dus ’khorb) las [[[’khor]] ba ’dir] c ’pho ba’i skad cig (gang yin pad) des sems can rnams kyi skye bar byed de… zhes dang / mngal du kun gzhi rnam par shes pa rdul dang khu ba dang yang dag par ldan pa’i (chos can noe) // zhes dang / [[[nyi ma]] rdul dang zla ba khu ba] dus kyi ’khor lo med pa ste kun gzhi’i rnam par shes pa med pas mi byed do // a GCMD rnams; b GCMD dur khrod; c addit. as per D 1347; d GCKL, GCMD om.; addit. as per D 1347; e GCKL do // All em. as per GCKL. See Vimalaprabhāṭīkā, Tib. D 1347, 267a2‒3. According to the Kālacakra account of embryogenesis, when the ālayavijñāna combines with the uterine blood of the female and semen of the male due to movements of vital life forces (prāṇa), conception takes place. As V. Wallace (2001: 6) explains: “At the time of conception, the father’s semen and mother’s uterine blood, which are made of the five elements, are ‘devoured’ by the consciousness which, accompanied by subtle prāṇas, enters the mother’s womb. When conception takes place due to the power of time, the semen and uterine blood within the womb slowly develop into the body of the individual. This occurs due to the spreading of prāṇas. The growing fetus consumes food comprised of six flavors – bitter, sour, salty, pungent, sweet, and astringent – and these six flavors originate from the six elements, the sixth being gnosis. Consequently, the body of a fetus becomes a gross physical body, composed of the agglomerates of the atomic particles. The elements of the father’s semen give rise to the marrow, bones, nāḍīs, and sinews of the fetus; the elements of the mother’s uterine blood give rise to the skin, blood, and flesh of the fetus. Thus, all the elements and psychophysical aggregates that constitute the human being come into existence due to the union of the atomic agglomerates of the father’s semen and mother’s uterine blood.” This process is described in detail in Vimalaprabhāṭīkā, Tib. D 1347, 115a6‒115b6.
and  that which is the ground of all latent tendencies even [when] the afflictions have disappeared.”119 He notes that the ground of afflictions is overcome once the state of arhatship [is attained by] śrāvakas and pratyekabuddhas, whereas the so-called “ground of latent tendencies of ignorance” is overcome at the end of traversing the ten spiritual levels of a bodhisattva. The Eighth Karma pa proceeds to defend the Yogācāra theory of ālayavijñāna over rival non-Buddhist ātmavādin theories on the grounds that it provides a model of the genesis and possible cessation not only of afflictions (kleśa), but also of the latent tendencies which continually give rise to them. It is only the complete eradication of the latter that will bring liberation from rebirth. Thus, if one claims that the succession of births in saṃsāra is terminated by merely having relinquished afflictions without knowing how the substratum (ālaya) serves as a basis for coarse and subtle afflictions, the latent tendencies and the like, this would be like the [view of the] nihilist heretics. For this reason, the substratum theory is superior to [the ātman doctrine of] heretics. Mi bskyod rdo rje acknowledges the explanatory value of the ālayavijñāna model when it comes to identifying the most deep-seated obstacles to liberation, the source of latent tendencies of ignorance.
119 Dgongs gcig kar ṭīg V.2, MDSB vol. 6, 8904–5: de la kun gzhi la nyon mongs thams cad kyi gzhi byed pa’i cha dang / nyon mongs log kyang bag chags thams cad kyi gzhir gyur pa gnyis yod pa … In particular, to attain great awakening, it is necessary to identify that obscuration which is the “ground (sa) of latent tendencies of ignorance” – [i.e.,] the substratum (ālaya) – that serves as an impediment to such [[[attainment]]]. But to identify that, it is necessary to engage in the vast spectrum of renunciation and realization pertaining to the level of buddhahood. Hence, with the exception of buddhas and bodhisattvas, this doctrinal approach to the substratum (ālaya) is not known by commoners and ordinary yogins. As the Laṅkāvatāra [II, re: v. 98]122 states: Were the endogenous form [of consciousness] to cease, then the ālayavijñāna would also cease.123 [However,] Mahāmati, if the ālayavijñāna ceased, then this doctrine would be no different from the nihilistic doctrine of the tīrthikas. Still, despite his recognition of the superiority of the Yogācāra account of the genesis of affliction and cyclic existence over selfhood theories, the Karma pa does not go along with the equation of ālayavijñāna and
122 LAS II, prose re: v. 98 (ed. Nanjio 1923: 3818‒392). svajātilakṣaṇe punarnirudhyamāne ālayavijñāna-nirodhaḥ syāt / ālayavijñāne punarnirudhyamāne nirviśiṣṭastīrthakarocchedavāde nāyaṃ vādaḥ syāt / Tib. H 110, 109b7‒110a2.
123 In the Laṅkāvatāra, the expression “endogenous form of consciousness” (svajātilakṣaṇa-vijñāna) refers to the deep structure of consciousness – the ālayavijñāna itself – which is thought to underlie the active manifest (or discernable) forms of consciousness (lakṣaṇa-vijñāna) and to survive their destruction. Here, the fundamental transformation or, literally, transformation of the basis (gnas gyur : āśraya-parāvṛtti, o-parivṛtti), whereby consciousness collapses or subsides into the ālayavijñāna (which in the LAS is equivalent to buddha nature), like waves into the ocean, first involves the cessation of these active manifest forms of consciousness (lakṣaṇavijñāna), and then of the continuity (prabandha) aspect. The key-point in the above quotation is that the cessation of dualistic perceptions and conceptions is the cessation of the karmic[ally conditioned] form of consciousness (karmalakṣaṇa) but not of the unconditioned endogenous form of consciousness, which is the ālayavijñāna itself.
tathāgatagarbha. To be more specific, although the Laṅkāvatāra, as a Yogācāra text, had taken ālayavijñāna as a condition of both cyclic existence and awakening and thus identified it with tathāgatagarbha, Mi bskyod rdo rje restricts the scope of ālayavijñāna to cyclic existence and relegates the conditions of awakening to tathāgatagarbha (the substratum proper) alone. This account leaves no room for the equation of buddha nature and the substratum consciousness. For the Eighth Karma pa, it is the tantric analysis of ālayavijñāna which probes the deepest roots of cyclic existence and thus offers the best prospect of eradicating them. In this analysis, ālayavijñāna “is otherwise termed ‘luminous mind as first principle’ (gtso bo : pradhāna, i.e., prakṛti)125 because it is said that the twenty-three transformations (pariṇāma) of saṃsāra evolve by virtue of this first principle.”126 It is further described as “‘mind in the fourth (bzhi pa : turīya) state,’
125 The term gtso bo (pradhāna), an epithet of prakṛti (“nature”) in the Sāṃkhya system, reflects the Kālacakra assimilation of concepts drawn from (or more familiar from) non-Buddhist sources into a Buddhist tantric context. As Mi bskyod rdo rje stresses in his Sku gsum ngo sprod rnam bshad, this assimilation should not be understood as an attempt to establish an equivalence between the two systems since postulates such as prakṛti, puruṣa, and ahaṃkāra are not accepted, even conventionally, by Buddhists. Rather, it should be seen as an attempt to reinterpret Sāṃkhya cosmology in light of Buddhist soteriological and psychological paradigms which reject these postulates. See Higgins and Draszczyk 2019 for details.
‘luminosity of deep sleep,’ and ‘the moment mind takes rebirth.’” Taken together, these largely non-Buddhist tantric terms identify a state of mind which has dispensed with afflictions but not yet with the most deep-seated tendencies that perpetuate saṃsāra. Mi bskyod rdo rje cautions that this type of luminous mind is still adventitious because it appropriates the entire eightfold ālayavijñāna complex and remains something separable (i.e., removable): “Because this kind of luminous mind is separable (’bral rung), it is described as ‘adventitious defilement.’ Further, since this substratum luminous mind, which is the root of saṃsāra, appropriates the entire eightfold consciousness, it is called the ‘appropriating cognition’ (ādānavijñāna).” In short, such altered states of mind may provide a glimpse of buddha nature, the ultimate, but should not be confused with it. As Mi bskyod rdo rje explains,
When these [[[mind states]]] manifest, although the mind of [[[buddha]]] nature remains unclear, it may nonetheless become slightly clearer, even to those who have not yet embarked upon the subtlest path. In that regard, however, the activities of maturation and liberation (smin grol) are not effective for them – a mere glimpse does not become the vajra yoga. Because of this subtle and profound point, my [[[teacher]]] the venerable Ras pa chen po declared: Comparative analysis of the expositions of the fourth state of the mind in the Kālacakratantra and in Śaiva tantras reveals striking similarities, and yet it shows some fundamental differences with regard to the nature of that state. They agree that the fourth state of the mind marks the blissful state of consciousness in which all conceptualizations disappear and any sense of duality vanishes. However, in Śaiva tantras, the fourth state of the mind is also a state of self-realization, a state in which one becomes aware of one’s undivided, essential Self, and consequently becomes free of spiritual ignorance (avidyā). It is a condition by which one rises to the fifth state, or the state of liberation, within one’s lifetime (jīvanmukti). In the Kālacakratantra, on the other hand, the fourth type of awareness, though nondual at the time of the emission of regenerative fluids, is still tainted with the habitual propensities of spiritual ignorance (avidyā-vāsanā) and is thus embedded in the cycle of existence.”
Nowadays, there are some who say, “since we have already integrated with the luminosity of deep sleep, there is no doubt we will awaken [to buddhahood] in the luminosity of death.” Many [of them] harbor [such] confidence [in their] minds. But let us not confuse mind which is the root of saṃsāra with luminosity! In the final analysis, then, the point of distinguishing buddha nature from the substratum consciousness is to clarify, as the Karma pa does in considerable detail, the respective conditions of awakening and delusion. His explication of the ālayavijñāna as the subtlest root of saṃsāra and as the final barrier to awakening is consistent with the Buddhist goal of eliminating all sources of suffering and bondage in order to realize liberation. But far from providing a justification for the exclusion of ālayavijñāna from the arena of Buddhist epistemology, his analysis instead legitimizes it as a worthy object of investigation – if only as an object of refutation (dgag bya) on the conventional level – on the same grounds that the conventional itself is accorded this status. The ultimate is discoverable only in and through the conventional, at which point the conventional is no more. It should be clear from the foregoing analysis that Mi bskyod rdo rje recognizes the superiority of the Yogācāra ālayavijñāna theory over rival non-Buddhist ātmavādin theories in accounting for the continuity of mental afflictions as well as the tendencies that perpetuate them. He does not hesitate to employ the ālayavijñāna model when it comes to delineating the set of conditions necessary for both cyclic existence and awakening. In these and other ways it becomes obvious that he does not reject the Yogācāra model of mind per se but only this tradition’s proclivity to hypostatize the mind and ālayavijñāna, to confuse them with wisdom and buddha nature and construe them as a basis of awakening.
These considerations certainly help explain why distinctions between pure and impure substrata – e.g., the substratum consciousness and substratum wisdom – assume the importance they do in Mi bskyod rdo rje’s philosophy. The reader should bear in mind, however, that such distinctions are viewed by the Karma pa as facets of a groundless ground, or a “substratum simpliciter” (kun gzhi tsam) in the wording of his teacher Karma phrin las. This ground is, paradoxically, found to be groundless in the dual sense of having no essential characteristics that make it what it is and no deeper, shovel-stopping bedrock on which it depends.
1. Abbreviations D GC GL Derge edition of Bka’ ’gyur and Bstan ’gyur. The Tibetan Tripiṭaka. Taipei Edition. Taipei: SMC Publishing, 1991. Dgongs gcig kar ṭīg. 3 editions: GCMD, GCKL, and GCBC; see Mi bskyod rdo rje. Rgan po’i rlung sman. 3 editions: GLMD, GLVV, GLNB; see Mi bskyod rdo rje. H KN KPDL MDSB P Lhasa edition of the Bstan ’gyur. Lhasa: Zhol par khang, 1934. Sku gsum ngo sprod rnam bshad. 3 editions: KNVV, KNMD, and KNRM; see Mi bskyod rdo rje. Karma phrin las dris lan. See Karma Phrin las Phyogs las rnam rgyal. Mi bskyod rdo rje gsung ’bum. 2 editions: MDSB and MDSB2; see Mi bskyod rdo rje. Peking edition of Bka’ ’gyur and Bstan ’gyur. The Tibetan Tripiṭaka. Peking Edition. Tokyo/Kyoto: Tibetan Tripiṭaka Research RDSB Institute, 1957. Karma pa Rang byung rdo rje gsung ’bum. See Rang byung rdo rje. AN AS Aṅguttaranikāya (part I). Ed. R. Morris, 2nd ed. rev. by A.K. Warder. London: Pali Text Society (PTS), 1923. Abhidharmasamuccaya. Ed. Pralhad Pradhan. Santiniketan: DDhS Visvabharati Publishing Press, 1950. Dharmadhātustava. A Critical Edition of the Sanskrit Text with the Tibetan and Chinese Translations, a Diplomatic Transliteration of the Manuscript and Notes. Ed. Liu Zhen. Sanskrit Texts from the Tibetan Autonomous Region no. 17. Beijing and Vienna: China Tibetology Publishing House and Austrian Academy of Sciences
LAS MA Press, 2015. D 1118, bstod tshogs, vol. ka, 63b5‒67b3 (pp. 1265‒1343). P 2010, bstod tshogs, vol. ka, 73a7‒77b8. Dohākoṣagīti, Saraha. See Shahidullah 1928. Dohākoṣahṛdayārthagītiṭīkā, Avadhūtīpa. Tibetan translation. D 2268, rgyud ’grel, vol. zhi, 65b–106b. P 3120, rgyud ’grel, vol. tsi, 97a–138a. Ghanavyūhasūtra. Tibetan translation. H 113, mdo sde, vol. cha, 1b1–86a2. Jñānavajrasamuccaya. Tibetan translation. H 787, rgyud, vol. tsha, 402a7-402b1 Laṅkāvatārasūtra. See Nanjio 1923. Madhyamakāvatāra-kārikā. Ed. Xuezhu Li. China Tibetology 18
(2012): 1–16. (Skt. text of MA chapter 6, verses 1–97.) Madhyamakāvatāra, Candrakīrti. Tibetan translation. Ed. Louis de la Vallée Poussin. Bibliotheca Buddhica 9. Delhi: Motilal Banarsi- MAVṬ dass, 1992 [1907–1912]. Madhyāntavibhāgaṭīkā, Sthiramati. Exposition systématique du Yogācāravijñapti-vāda. Ed. Susumu Yamaguchi. Nagoya: Librairie Hajinkaku, 1934. 2. Primary Sources: Indian Works
MPNS Mahāparinirvāṇamahāsūtra. Tibetan translation.
RGV versions. Louvain: Bureaux du Muséon, 1938. D 4048, sems tsam, vol. ri, 1b1–43a7. P 5549, sems tsam, vol. li, 1a1–51a2 (vol. 112, pp. 215–236). Ratnagotravibhāga Mahāyānottaratantraśāstra, Maitreya RGVV (ascribed). See Johnston 1950. Ratnagotravibhāgavyākhyā, Asaṅga (ascribed). See Johnston 1950. [The manuscripts A and B on which Johnston’s edition is based are 1b1–43a. Ed. Étienne Lamotte. Vol. 1: The Tibetan and Chinese
described in Johnston 1950: vi‒vii. See also Bandurski et al. 1994: ŚDS
TŚBh VMS VPṬ H 92, 435b5‒436a4. Triṃśikābhāṣya. See Lévi 1925: 15–45. Vijñaptimātratāsiddhi, Vasubandhu. See Lévi 1925: Vol. 1. Vimalaprabhāṭīkā, Puṇḍarīka. Ed. Jagannatha Upadhyaya. Bibliotheca Indo-Tibetica 11. Vol. 1. Sarnath: Central Institute of Higher Tibetan Studies, 1986. D 1347, rgyud, vol. 102, 1–5737. 3. Primary Sources: Tibetan Works ’Ba’ ra ba chen po Rgyal mtshan dpal bzang _____ Ngo sprod bdun ma’i ’grel pa man ngag rin po che’i sgron me. In Rje ’Ba’ ra ba chen po Rgyal mtshan dpal bzang gi bka’ ’bum. 14 vols. Dehradun: Ngawang Gyaltsen and Ngawang Lungtok, 1980: vol. 11, 203–282. Blo bzang Dam chos rgya mtsho _____ Rnam rig pa’i lugs kyi yid dang kun gzhi’i don cung zad bshad pa ngo mtshar gzugs brgya ’char ba’i me long. In Collected Works of Rongga Lozang Damchoe Gyatso gsung ’bum. 6 vols. New Delhi: N. Sopa, 1975: vol. 1, 187–198. Blo bzang ’Jam dbyangs smon lam _____ Yid dang kun gzhi’i rtsa ’grel gyi dka’ gnas gsal byed nyi zla zung ’jug. In Collected Works of Ke’u tshan sprul sku Blo bzang ’jam dbyangs smon lam. 2 vols. Dharamsala: Library of Tibetan Works and Archives, 1984: vol. 1, 187–260. Dwags ram pa Chos rgyal bstan pa _____ Zab mo nang don sems kyi rnam par thar pa’i gsal ba’i rgyan. Full title: Dpal rdo rje’i tshig zab mo nang gi don gyi ’brel bshad sems kyi rnam par thar pa’i gsal ba’i rgyan. In RDSB vol. 12, 1‒684. Gung thang Dkon mchog bstan pa’i sgron me _____ Yid dang kun gzhi’i dka’ gnas rnam par bshad pa mkhas pa’i ’jug ngog. In The Collected Works (gsung ’bum) of Gung thang Dkon mchog bstan pa’i sgron me. 8 vols. Lhasa, 2000: vol. 2, 279–406. ’Jam dbyangs Bzhad pa’i rdo rje _____ Grub mtha’ rnam bshad rang gzhan Grub mtha’ kun dang zab don mchog tu gsal ba kun bzang zhing gi nyi ma lung rigs rgya mtsho skye dgu’i re ba kun skongs. In Collected Works (gsung ’bum) of ’Jam dbyangs Bzhad pa’i rdo rje. 16 vols. Mundgod, Karnataka: Gomang College, 1997: vol. 15, 33–1092. ’Jigs med gling pa Mkhyen brtse ’od zer _____ ’Jigs gling gsung ’bum. The A-’dzom chos-sgar redaction of the Collected Works of Kun-mkhyen ’Jigs-med gling-pa Rang-byung rdo-rje mkhyen-brtse’i ’od-zer. 14 vols. Paro: Lama Ngodrup and Sherab Demy, 1985–. _____ Rdzogs pa chen po Kun tu bzang po ye shes klong gi rgyud. In ’Jigs gling gsung ’bum, vol. 12, 63–82. Karma Phrin las Phyogs las rnam rgyal _____ Dri lan drang ba dang nges pa’i don gyi snang byed ces bya ba ngo gro. In KPDL ca 108‒139. _____ Karma phrin las dris lan (KPDL). Full title: Chos rjes karma phrin las pa’i gsung ’bum las thun mong ba’i dri lan gyi phreng ba rnams. In The Songs of Esoteric Practice (Mgur) and Replies to Doctrinal Questions (Dris-lan) of Karma-’phrin-las-pa. New Delhi: Ngawang Topgay, 1975: 87‒223. Reproduced from prints of the 1539 Rin-chen-ri-bo Blocks. _____ Zab mo nang don gyi rnam bshad snying po. Full title: Zab mo nang don rnam bshad snying po gsal bar byed pa’i nyin byed ’od kyi phreng ba. Karma phrin las. In RDSB vol. 14, 1‒553. Kong sprul Blo gros mtha’ yas _____ Rnam shes ye shes ’byed pa’i bstan bcos ’grel pa. Full title: Rnam shes dang ye shes ’byed pa’i bstan bcos kyi tshig don ’grel pa rang byung dgongs pa’i rgyan. In: Mdo sngags mtshams sbyor. Zi ling: Mtsho sngon mi rigs dpe skrun khang, 2004: 331‒385. Mchims Blo bzang grags pa _____ Chos mngon pa gsal byed legs par bshad paʼi rgya mtsho. Sarnath: Sakya Students’ Union, 1996. Mi bskyod rdo rje, Karma pa VIII _____ Bla ma khams pa’i dris lan mi gcig sems gnyis. In MDSB vol. 3, 219‒223. _____ Dgongs gcig kar ṭīg (GC). Collection of 8 texts on the Single Intent (dgongs gcig) system. Editions used:  GCMD: In MDSB vols. 3–6.  GCKL: ’Bri gung dgongs gcig kar ṭīg chen mo. 5 vols. Kathmandu: Bal yul Karma legs bshad gling bshad grwa, 2012.  GCBC: ’Bri gung bka’ brgyud chos mdzod chen mo. 151 vols. Lhasa: 2004: vols. 77–83. _____ Dgongs gcig kar ṭīg I.2. Full title: Dgongs gcig grub mtha’i spyi ching. Full title: Chos dang chos ma yin par rnam par ’byed pa’i gtam chen po zab mor nang don ’khrul par ngo sprod par byed pa grub mtha’i spyi ching. Editions used: GCMD: In MDSB vol. 4, 19‒68. Also in GCKL vol. 1, 11–42. _____ Dgongs gcig kar ṭīg V.2. Full title: Dgongs gcig gi gsung bzhi bcu pa’i ’grel pa. In MDSB vol. 6, 728‒939. Also in GCKL vol. 5, 235–438. Also in GCBC vol. 83, 1‒289. _____ Mi bskyod rdo rje gsung ’bum (MDSB). Full title: Dpal rgyal ba karma pa sku ’phreng brgyad pa mi bskyod rdo rje gsung ’bum. Editions used:  MDSB: From a computer generated dbu can edition of the Mi bskyod rdo rje gsung ’bum. 26 vols. Lhasa, 2004.  MDSB2: Dpe rnying bris ma edition. From a xylographic copy of a handwritten manuscript dbu med edition of Karmapa Mi bskyod rdo rje gsung ’bum. 14 vols. Varanasi: Vajra Vidya Institute, n.d. _____ Paṇ chen rdo rgyal ba’i legs bshad. Full title: Paṇ chen rdo rgyal ba’i legs bshad rnam par dkar ba’i shel gyi glegs bu la drang po’i thig vaiḍūrya’i ri mo btab pa’o. In MDSB vol. 3, 224‒257. _____ Rgan po’i rlung sman or Bdud rtsi’i dris mchog (GL). Full title: Rje yid bzang rtse ba’i rgyud gsum gsang ba dang / paṇ chen shakya mchog ldan gyi bde mchog rnam bshad gnyis kyi mthar thug gi ’bras bu gzhi dus gnas lugs / lam dus kyi rnal ’byor rnams la dpyad pa bdud rtsi’i dri mchog. Editions used:  GLMD: In MDSB vol. 15, 975‒10245.  GLVV: In MDSB2, vol. 6, 1a–22b2.  GLNB: In Rgyal dbang karma pa sku ’phreng brgyad pa mi bskyod rdo rje’i rnal ’byor rgyud kyi rnam bshad sogs. 4 vols. Thimpu: Kunsang topgyel, 1979: vol. 3, 249‒3356. _____ Sku gsum ngo sprod rnam bshad (KN). Full title: Sku gsum ngo sprod kyi rnam par bshad pa mdo rgyud bstan pa mtha’ dag gi e vaṃ phyag rgya. Editions used:  KNVV: Varanasi: Vajra Vidya Institute Library, 2013, 3 vols.  KNMD: In MDSB vols. 21‒22.  KNRM: Dbu med edition. 4 vols. Gangtok: Rumtek Monastery, 1978. _____ Zab mo phyag chen gyi mdzod sna tshogs ’dus pa’i gter. In MDSB vol. 15, 1025‒38. Rang byung rdo rje, Karma pa III _____ Chos dbyings bstod pa’i ’grel pa. Full title: Dbu ma chos dbyings bstod pa’i rnam par bshad pa. In RDSB vol. 7, 1‒125. _____ RDSB: Karma pa Rang byung rdo rje gsung ’bum. 16 vols. Ziling: mTshur phu mkhan po lo yag bkra shis, 2006. _____ Zab mo nang don gyi ’grel pa. Full title: Gsang sngags zab mo nang don kyi ’grel pa. In RDSB vol. 7, 371‒664. _____ Zab mo nang gi don zhes bya ba’i gzhung. In RDSB vol. 7, 308–360. Shākya mchog ldan _____ Bde mchog rnam bshad. Full title: ’Khor lo sdom pa la rgyun chags kyi sdeb sbyor gyi sgo nas bstod pa dang // Bde mchog rnam bshad dpal dang po’i sangs rgyas rab tu [text: du] grub pa. In: Paṇ chen Shākya mchog ldan gyi gsung ’bum legs bshad gser gyi bdud rtsi. Delhi: Ngagwang Topgyal, 1995: vol. 8, 1–193. Sgra sbyor bam po gnyis. See Ishikawa 1990. Tsong kha pa Blo bzang grags pa _____ Yid dang kun gzhi’i dka’ ba’i gnas rgya cher ’grel pa. Collected Works of Tsong kha pa. 27 vols. New Delhi: Ngawang gelek demo, 1975–: vol. 27, 356–474. Bkra shis lhun po edition. See also Bstan ’gyur P 6149, 173–95. Wŏnch’ŭk _____ Dgongs ’grel gyi ’grel chen. Full title: ’Phags pa dgongs pa zab mo nges par ’grel pa’i mdo rgya cher ’grel pa. D 4016, vol. 220, 1b1‒291a7. 4. Secondary Sources Almogi, Orna. 2010. “Māyopamādvayavāda versus Sarvadharmāpratiṣṭānavāda: A Late Indian Subclassification of Madhyamaka and its Reception in Tibet.” Journal of the International College for Postgraduate Buddhist Studies 14: 135‒212. Anālayo, Bhikkhu. 2017. “The Luminous Mind in Theravāda and Dharmaguptaka Discourses.” Journal of the Oxford Centre for Buddhist Studies 13: 10–51. Bandurski, Frank, Bikkhu Pāsādika, Michael Schmidt, and Bangwei Wang. 1994. Untersuchungen zur buddhistischen Literatur. Sanskrit-Wörterbuch der buddhistischen Texte aus den Turfan-Funden. Beiheft 5. Göttingen:
Vandenhoeck & Ruprecht.
Böhtlingk, Otto. 1998 [1879–1889]. Sanskrit-Wörterbuch in Kürzerer Fassung. 7 vols. Hamburg: Buske. Burton, David. 2004. Buddhism, Knowledge and Liberation: A Philosophical Study. Aldershot: Ashgate Publishing Co. Davidson, Ronald. 1985. “Buddhist Systems of Transformation: Āśraya-parivṛtti/-parāvṛitti Among the Yogācāra.” PhD thesis, University of California. Demiéville, Paul. 1929. “Sur l’authenticité du Ta tch’eng k’i sin louen.” Bulletin de la Maison Franco-Japonaise. Série française, tome II, No. 2: –78. Eckel, David. 2008. Bhāviveka and His Buddhist Opponents. Cambridge, MA: Harvard University Press. Edgerton, Franklin. 1953. Buddhist Hybrid Sanskrit Grammar and Dictionary. 2 vols. Delhi: Motilal Banarsidass. Franco, Eli. 1997. Dharmakīrti on Compassion and Rebirth. Wiener Studien zur Tibetologie und Buddhismuskunde, 38. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Franco, Eli. 2000. “Lost Fragments from the ‘Spitzer Manuscript.’” In R. Tsuchida and A. Wezler, eds., Harānandalaharī. Reinbek: Wezler: 77–110. Frauwallner, Erich. 1951. “Amalavijñānam und Ālayavijñānam.” Beiträge zur indischen Philologie und Altertumskunde: Walter Schubring zum 70. Geburtstag. Hamburg: De Gruyter: 148–159. Griffiths, Paul. 1991. On Being Mindless: Buddhist Meditation and the Mind- Body Problem. La Salle, Illinois: Open Court Publishing Co. Higgins, David. 2013. The Philosophical Foundations of Classical rDzogs chen in Tibet: Investigating the Distinction between Dualistic Mind (sems) and Primordial Knowing (ye shes). Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 78. Wien: Arbeitskreis für Tibetische und Buddhistische Studien der Universität Wien. Higgins, David. Forthcoming. “*Bodhigarbha (byang chub snying po): Preliminary Notes on an Early Rdzogs chen Concept of Buddha Nature.” In Klaus-Dieter Mathes and Michael Sheehy, eds., The Other Emptiness: Perspectives on the Zhentong Buddhist Discourse in India and Tibet. New York: State University of New York (SUNY) Press. Higgins, David and Martina Draszczyk. 2016. Mahāmudrā and the Middle Way. Post-Classical Kagyü Discourses on Mind, Emptiness and Buddha Nature. 2 vols. Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 90.1–2. Wien: Arbeitskreis für Tibetische und Buddhistische Studien der Universität Wien. Higgins, David and Martina Draszczyk. 2019. Buddha Nature Reconsidered: The Eight Karma pa’s Middle Path. 2 vols. Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 95.1–2. Wien: Arbeitskreis für Tibetische und Buddhistische Studien der Universität Wien. Hodge, Stephen. 2012. “The Mahāyāna Mahāparinirvāṇa-sūtra: The Text and its Transmission.” Revised version of a paper presented in July 2010 at the Second International Workshop on the Mahāparinirvāṇasūtra, Munich University. Hamburg: Zentrum für Buddhismuskunde. https://www.buddhismuskunde.uni-hamburg.de/pdf/5-personen/hodge/the-textual-transmisssionof-the-mpns.pdf. Accessed July 2019. Husserl, Edmund. 1989. Ideas Pertaining to a Pure Phenomenology and to a Phenomenological Philosophy. Second Book. Tr. R. Rojcewicz and A. Schuwer. Dordrecht: Kuwer Academic Publishers. Ishikawa, Mie. 1990. A Critical Edition of the sGra sbyor bam po gnyis pa: An Old and Basic Commentary on the Mahāvyutpatti. Studia Tibetica 18. Tokyo: The Toyo Bunko. Johnston, Edward H., ed. 1950. The Ratnagotravibhāga Mahāyānottaratantraśāstra. Patna: Bihar Research Society. (Includes the Ratnagotravibhāgavyākhyā.) Jones, Christopher V. 2015. “The Use of, and Controversy Surrounding, the Term Ātman in the Indian Buddhist tathāgatagarbha Literature.” PhD thesis, St. Peter’s College, Oxford University. Kano, Kazuo. 2016. Buddha-Nature and Emptiness. rNgog Blo-ldan-shes-rab and A Transmission of the Ratnagotravibhāga from India to Tibet. Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 91. Wien: Arbeitskreis für Tibetische und Buddhistische Studien der Universität Wien. Kapstein, Matthew. 2002. The Tibetan Assimilation of Buddhism: Conversion, Contestation, and Memory. Oxford: Oxford University Press. Lai Whalen. 1977. “The Meaning of ‘Mind-Only’ (Wei-Hsin): An Analysis of a Sinitic Mahāyāna Phenomenon.” Philosophy East and West 27.1: 61–77. Lamotte, Étienne. 1962. L’Enseignement de Vimalakīrti. Bibliothèque du Muséon, vol. 51. Louvain: Publications universitaires. Lévi, Sylvain, ed. 1925. Vijñaptimātratāsiddhi. Bibliothèque de l’École des Hautes Études, Sciences Historiques et Philologiques 245. Paris: Champion. (Includes the Triṃśikābhāṣya.) Liu Zhen 2015. See DDhS. Mathes, Klaus-Dieter. 2008. A Direct Path to the Buddha Within, Gö Lotsawa’s Mahāmudrā Interpretation of the Ratnagotravibhāga. Boston: Wisdom Publications. Mimaki, Katsumi. 1982. Blo gsal grub mtha’: Chapitres IX (Vaibhāṣika) et XI (Yogācāra) édités et Chapitre XII (Mādhyamika) édité et traduit. Kyoto: Université de Kyoto, Zinbun Kagaku Kenkyusyo. Monier-Williams, Monier. 1899. A Sanskrit-English dictionary etymologically and philologically arranged with special reference to cognate Indo-European languages. New edition, greatly enlarged and improved with the collaboration of Ernst Leumann and Carl Cappeller. Oxford: The Clarendon Press. Nagao, Gajin. 1978. Chūkan to Yuishiki. Tōkyō: Iwanami Shoten. Nanjio, Bunyiu, ed. 1923. The Laṅkāvatāra sūtra. Kyoto : Otani University Press. Paul, Diana Y. 1984. Philosophy of Mind in Sixth Century China: Paramārtha’s “Evolution of Consciousness.” Stanford: Stanford University Press. Radich, Michael. 2008. “The Doctrine of *Amalavijñāna in Paramārtha (499– 569), and Later Authors to Approximately 800 C.E.” Zinbun 41: 45–174. Radich, Michael. 2015a. “Tathāgatagarbha Sūtras.” In Jonathan Silk, Oskar von Hinüber, and Vincent Eltschinger, eds., Brill Encyclopedia of Buddhism. Handbuch der Orientalistik. 2. Abt., Indien, 29, vol. 1. Leiden: Brill: 261–273. Radich, Michael. 2015b. The Mahāparinirvāṇa-mahāsūtra and the Emergence of tathāgatagarbha Doctrine. Hamburg Buddhist Studies 5. Hamburg: Hamburg University Press. Radich, Michael. 2016. “Pure Mind in India: Indian Background to Paramārtha’s *Amalavijñāna.” Journal of the International Association of Buddhist Studies 39: 249–308. Sakuma, Hidenori. 1990. Die Āśrayaparivṛtti-Theorie in der Yogācārabhumi. Alt- und Neu-Indische Studien 40, vols. 1‒2. Stuttgart: Franz Steiner Verlag. Schmithausen, Lambert. 1971. “Philologische Bemerkungen zum Ratnagotravibhāga.” Wiener Zeitschrift für die Kunde Südasiens 15: 123–177. Schmithausen, Lambert. 1977. “Textgeschichtliche Beobachtungen zum 1. Kapitel der Aṣṭasāhasrikā Prajñāpāramitā.” In Lewis Lancaster and Luis O. Gómez, eds., Prajñāpāramitā and Related Systems: Studies in Honor of Edward Conze. Berkeley Buddhist Studies Series 1. Berkeley: University of California Press: 35–80. Schmithausen, Lambert. 1987. Ālayavijñāna: on the Origin and the Early Development of a Central Concept of Yogācāra Philosophy. 2 vols. Studia Philologica Buddhica, Monograph Series 4. Tokyo: The International Institute for Buddhist Studies. Schmithausen, Lambert. 2014. The Genesis of Yogācāra-Vijñānavāda: Responses and Reflections. Tokyo: International Institute for Buddhist Studies of the International College for Postgraduate Buddhist Studies. Seyfort Ruegg, David. 1969. La théorie du tathāgatagarbha et du gotra; études sur la sotériologie et la gnoséologie du bouddhisme. Publications de l’École Française d’Extrême-Orient 70. Paris: Maisonneuve. Seyfort Ruegg, David. 1973. Le Traité du tathāgatagarbha de Bu ston rin chen grub. Publications de l’École française d’Extrême-Orient 88. Paris: École française d’Extrême-Orient. Seyfort Ruegg, David. 1981. “Autour du lTa ba’i khyad par de Ye šes sde (Version de Touen-houang, Pelliot Tibétain 814).” Journal Asiatique 269: 207–229. Seyfort Ruegg, David. 1989. Buddha-nature, Mind and the Problem of Gradualism in a Comparative Perspective. On the Transmission and Reception of Buddhism in India and Tibet. Jordan Lectures 1987. London: School of Oriental and African Studies, University of London. Seyfort Ruegg, David. 2000. Three Studies in the History of Indian and Tibetan Madhyamaka Philosophy. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien. Shahidullah, Muhammad. 1928. Les chants mystiques de Kāṇha et de Sahara: Les Dohā-koṣa (en apabhraṃśa, avec les versions tibétaines) et les Caryā (en vieux bengali). Paris: Adrien-Maisonneuve. Sparham, Gareth, tr. 1993. Ocean of Eloquence. Tsong kha pa’s Commentary on the Yogācāra Doctrine of Mind. In collaboration with Shotaro Iida. SUNY Series in Buddhist Studies. Albany: SUNY Press. Stearns, Cyrus.  2010. The Buddha from Dolpo: A Study of the Life and Thought of the Tibetan Master Dolpopa Sherab Gyaltsen. Ithaca, New York: Snow Lion Publications. Steinbock, Anthony J. 1995. Home and Beyond: Generative Phenomenology after Husserl. Evanston: Northwestern University Press. Takasaki, Jikido. 1966. A Study on the Ratnagotravibhāga (Uttaratantra) Being a Treatise on the Tathāgatagarbha Theory of Mahāyāna Buddhism. Serie Orientale Roma XXXIII. Roma: Istituto Italiano per il Medio ed Estremo Oriente. Thompson, Evan. 2007. Mind in Life: Biology, Phenomenology, and the Sciences of Mind. Cambridge, MA: Harvard University Press. Ueda, Yoshifumi. 1967. “Two Main Streams of Thought in Yogācāra Philosophy.” Philosophy East and West 17.1/4: 155–165. Waldron, William S. 2003. The Buddhist Unconscious: The Ālaya-vijñāna in the context of Indian Buddhist Thought. New York: Routledge Curzon. Wallace, Vesna A. 2001. The Inner Kālacakratantra a Buddhist Tantric View of the Individual. New York: Oxford University Press. Wangchuk, Dorji. 2007. The Resolve to Become a Buddha: A Study of the Bodhicitta Concept in Indo-Tibetan Buddhism. Studia Philologica Buddhica Monograph Series 23. Tokyo: The International Institute for Buddhist Studies. Wittgenstein, Ludwig.  2001. Philosophical Investigations. Oxford: Blackwell Publishing. Zahavi, Dan. 1999. Self-Awareness and Alterity: A Phenomenological Investigation. Evanston: Northwestern University Press. Zimmermann, Michael. 2002. A Buddha Within: The Tathāgatagarbhasūtra, The Earliest Exposition of the Buddha-Nature Teaching in India. Bibliotheca Philologica et Philosophica Buddhica 6. Tokyo: The International Research Institute for Advanced Buddhology. Zimmermann, Michael. 2014. “The Process of Awakening in Early Texts on Buddha-Nature in India.” In Chen-kuo Lin and Michael Radich, eds., A Distant Mirror. Hamburg: Hamburg University Press: 513–528.
This paper addresses the age-old question of how buddha nature (tathāgatagarbha) relates to Yogācāra psychology, focusing on the Eighth Karma pa Mi bskyod rdo rje’s (1507–1554) responses to this question. In the centuries following the first appearance of tathāgatagarbha doctrines in India (circa 2nd c. CE), opinions became divided over whether buddha nature should be identified with or distinguished from the Yogācāra idea of a substratum consciousness (ālayavijñāna). The topic attracted a great deal of discussion and debate among Buddhist scholars, both within and beyond the borders of India. At stake were a set of specific doctrinal issues as to whether and how the Yogācāra ālayavijñāna-vāsanā model could be reconciled with  buddha nature theory  tantric buddha nature
proxies such as the unconditioned ground (gzhi) and causal continuum (rgyu rgyud)  Indian and Chinese Buddhist conceptions of an immaculate consciousness (amalavijñāna) and  certain anti-foundationalist strains of Middle Way (Madhyamaka) philosophy that rejected any transcendental basis of consciousness. The Karma pa’s repeated forays into these contested subject areas reveal time and again his commitment to reconcile two contrasting lines of Buddhist thought and praxis:  the affirmative appraisal of the nature of mind and reality emphasized in Yogācāra and tathāgatagarbha classics, the tantras, and the songs and writings of the Buddhist mahāsiddhas and  the metaphysically disinclined stance of the Middle Way (Madhyamaka) philosophy that avoided extremes of
affirmation and denial, existence and nonexistence. To adequately appreciate his contributions to such issues, I will first sketch in rough strokes the historical evolution of the ālayavijñāna doctrine and its complex confrontations with tathāgatagarbha doctrine in India. Against this backdrop, attention will turn to the Karma pa’s contextualist framing of the ālayavijñāna-tathāgatagarbha relationship in terms of a progressive understanding that begins with
differentiation and culminates in unity. His is a view that stresses the need to initially distinguish between conditions of spiritual awakening (such as tathāgatagarbha) and delusion (such as ālayavijñāna) in order to eventually realize their underlying unity (zung ’jug) by recognizing buddha nature as an ever-present continuum (rgyud) of awareness that is a precondition of the substratum consciousness that derives and deviates from it. In his attempts to strike a balance between traditional differentiation and unity models, we encounter a thinker who was as confident about the mind’s ability to discover its own unborn and nonconceptual nature as he was skeptical about its ability to discover any underlying metaphysical foundation.