Contemplative Practice , Doxographies , and the Construction of Tibetan Buddhism : Nupchen Sangy é Yesh é and The Lamp for the Eye in Meditation

In this article, I would like to reframe our understanding of the role played by doxographies or classification of views (Skt. siddhānta, Ch. panjiao Religions 2018, 9, x; doi: FOR PEER REVIEW www.mdpi.com/journal/religions Article Contemplative Practice, Doxographies, and the Construction of Tibetan Buddhism: Nupchen Sangyé Yeshé and The Lamp for the Eye in Meditation

either the Sanskrit term dhyāna 2 or 'meditative absorption', or the other Sanskrit term bhāvanā 3 or 'cultivation'.Whatever the precise Sanskrit equivalent, though, Guenther's remark stands, and reflects a powerful paradox at the heart of the Buddhist tradition.On the one hand, Buddhism established, at least rhetorically, its very own foundations on the transformative meditative experience that Siddharta Gautama, the Buddha, had over 2500 years ago sitting under the bodhi tree while searching for the ultimate answers to his disillusionment with the world and the human condition, eventually culminating with his experience of enlightenment. 4 At the same time, as Guenther also points out, and as anyone who has even a superficial knowledge of various forms of Buddhist practice, what Buddhists meant (and mean) by 'meditation', how to define it, how it should be practiced, and even its final goal, became a constant issue of contention over the centuries.
'Meditation' is one of those Buddhist concepts that seems to have a foundation of solid rock, particularly in the eyes of believers, but when examined closely it resembles shifting sands. 5It is by exploring those shifting sands that I will argue that 'meditation', as a concept as well as a practice, functions as a fertile and contested ground which has allowed the Buddhist tradition to assert an important degree of continuity and unity (at least rhetorically) with the teachings of the Buddha and with that primordial contemplative experience of its founder.At the same time, that same contested nature opened the door to important discontinuities and diversity, as reflected in the emergence of new Buddhist contemplative traditions and schools of thought over the centuries. 6ne tool used within the Buddhist tradition to articulate this diversity was the genre of doxographies or classification of views (Skt.siddhānta, Ch. panjiao 判教, Tib.grub mtha').Doxographies played a central role in the expansion of the Buddhist tradition from its native India to other parts of Asia, particularly Central and East Asia, by becoming a tool for organizing the historical and geographical diversity of Buddhist scriptures, as well as of ritual and doctrinal practices that developed over time.In China, for example, Peter Gregory has argued that "doctrinal classification has often been said to be the hallmark of Chinese Buddhism". 7Gregory, though, focused mostly on the sectarian purposes of doxographies, and claimed that they served very specific "hermeneutical, sectarian, 2 See the entry for "Dhyāna" in Buswell et al. (2014): "Dhyāna.(P.jhāna; T. bsam gtan; C. chan/chanding; J. zen/zenjō; K. sȏn/sȏnjȏng).In Sanskrit, 'meditative absorption', refers to specific meditative practices during which the mind temporarily withdraws from external sensory awareness and remains completely absorbed in an ideational object of meditation."Gómez, in his definition of meditation in the Encyclopedia of Buddhism, equates "meditation" with "dhyāna": see "meditation" entry by Gómez in Buswell (2004).
3 See the entry for "Bhāvanā" in (Buswell et al. 2014): "(Tib.sgom pa; Ch. xiuxi; Jap.shuj ū; Kor.sus ȗp).In Sanskrit and Pāli, "cultivation" (lit."bringing into being"), a Sanskrit term commonly translated into English as "meditation."It [ . . .]  has a wide range of meanings including cultivating, producing, manifesting, imagining, suffusing, and reflecting.It is in the first sense, that of cultivation, that the term is used to mean the sustained development of particular states of mind."Germano and Waldron equate "meditation" with bhāvanā in their entry to "Buddhist Meditation" in the Encyclopedia of Religion.See (Jones et al. 2005).4   The classic biographies of the Buddha, such as Aśvaghos .a's Buddhacarita, and the Lalitavistara S ūtra, maintain a dramatic tension between his birth already as a Buddha, and his life as a human and ultimate experience of enlightenment under the Bodhi Tree.Despite a tradition of the Buddha as an already-enlightened being, as found in those narratives, there is no doubt that the experience of enlightenment marks a before and after in the life narrative of the Buddha.On this particular tension, see (Silk 2003).
5 Luis Gómez uses this notion of "shifting sands" to define a different key Buddhist concept, "Nirvān .a", but his ideas equally apply to our exploration of the concept of meditation.For Gómez, "Nirvān .a" "has acquired a patina that makes many assume its meaning is obvious.Yet, it is a word about which Buddhists themselves have never reached agreement."Following Gómez, "it may be that when we ask: 'What is Nirvān .a?' [or in our case, "What is Meditation"?]we seek to answer the wrong question.Instead we need to ask: How have Buddhists used the term?With what polemical or apologetic purposes?What human aspirations might these uses reveal?"See his "Nirvān .a" entry in the Encyclopedia of Buddhism in (Buswell 2004).
6 Before I continue with my argument, let me add a caveat: I am not suggesting here that "meditation", or "meditative experience", is the essence of Buddhism.Robert Sharf has pointed out in his work the dangers of reconstructing the Buddhist tradition through what he considers a Western obsession with the notion of experience that he traces back to William James and his The Varieties of Religious Experience.For Sharf, "[W]hile meditation may have been important in theory, it did not occupy the dominant role in monastic and ascetic life that is sometimes supposed."See (Sharf 1995).What I do argue, though, is that the concept of "meditation" has functioned as an important marker that has helped define boundaries between various Buddhist traditions.7 He also adds: "Although this judgment is surely one-sided-ignoring as it does many rich areas of more 'popular' developments-it is certainly no exaggeration when applied to Chinese Buddhist scholastic writing.Doctrinal classification and soteriological purposes.That is, [doxography] organized into a coherent and internally consistent doctrinal framework the diverse corpus of [Buddhist] sacred scriptures [...]; it legitimized the claims of different [Buddhist] traditions to represent the supreme, orthodox, or most relevant teaching of the Buddha; and it provided a map of the Buddhist path". 8Chanju Mun expanded this important but narrow understanding of doxographies by adding what he calls an "ecumenical" type of doxographies, or classification of views that were not so worried about establishing sectarian differences, and more focused on articulating and integrating a universal type of Buddhist discourse. 9Jacob Dalton, in his overview of late Indian and early Tibetan doxographies, highlighted the role that these texts had in articulating Buddhist ritual and doctrine during the emergence of Tantra in India in the eighth century, as well as in articulating and organizing the ritual and doctrinal diversity of Buddhist practice as it was being introduced in Tibet beginning in the ninth century.
In this article, I would like to expand our understanding of the role of doxographies in the Buddhist tradition by examining Tibetan attempts at defining and organizing the diversity of Buddhist contemplative practices that made their way into Tibet since the introduction of Buddhism on the Tibetan plateau in the seventh century, all the way up to the collapse of the Tibetan Empire in the ninth century.In order to do that, I want to focus on one such doxography, the Lamp for the Eye in Meditation (bsam gtan mig sgron), 10 composed in the 10th century by the Tibetan scholar Nupchen Sangyé Yeshé.As part of my main argument, I will use Jonathan Z. Smith's understanding of the concepts of definition, comparison, and classification to explore how Nupchen used the doxographical genre not only as a vehicle to organize and articulate doctrinal and contemplative diversity, but also used it as a tool that allowed him to claim the superiority of a new and indigenous system of Tibetan Buddhist practice known as the Great Perfection (rdzogs chen), while legitimizing it within the larger context of Buddhist contemplative systems.Following Jonathan Z. Smith, I will also argue that Nupchen was not only describing in the Lamp Buddhism as he saw it, but he was interpreting Buddhism as he imagined it, playing a central role during this period into the transformation of Buddhism from a foreign, imported tradition, into something new: Tibetan Buddhism.

Doxographies: The Role of Comparison and Classification in the Construction of Buddhism
The term doxography was coined by the German classical scholar, Hermann Alexander Diels, to describe the work of classical historians that included discussions of Greek and Roman philosophers of the past. 11Modern Buddhist scholars have adopted this term to describe a similar genre within the Buddhist tradition called siddhānta in Sanskrit, panjiao in Chinese (判教), and grub-mtha' in Tibetan. 12is one of the most striking features of Chinese Buddhist scholasticism, and it is impossible to understand how medieval Chinese Buddhist scholars thought without understanding panjiao."(Gregory 1987, p. 93).8 (Gregory 2002, p. 115).9 (Mun 2006). 10Henceforth referred to as 'the Lamp'. 11(Zhmud 2001). 12I would like to add, though, an important terminological clarification.Following most of the scholarship on this subject, this article will use the term doxography in order to describe these Buddhist systems of doctrinal classification.Doxography means etymologically to "write the views", and its origins are attributed to the Greek tradition.The term describes a literary genre in which views of past philosophical systems are organized, usually, in some sort of historical fashion.The use of this term in the context of Indian, Tibetan, and Chinese classification of views, quickly shows that the term does not always describe accurately the nature and goals of these texts.In India the term used for this genre of literature is siddhānta, in Chinese panjiao (Ch.判教), and in Tibetan they are called grub mtha'.In all cases the meaning generally refers to the description of systems of practice (of realization, literally).There are some general similarities among these Buddhist terms, but there are also important differences.In some of the early classifications (especially in the Indian ones) the main goal was not to classify hierarchically, but to argue specific doctrinal or philosophical views.That is why some scholars are proposing that the term summa may be more accurate to describe the siddhānta or grub mtha' literature.On these issues see (Mestanza 2005, p. 85).To clarify these issues is beyond the scope of this article, but it is important to point out that we need to be careful when using Western terminology to describe Buddhist categories, since they bring with them unintended meanings to the object of our study.Another issue related to the use of the term doxography is the problem of determining what counts as one.In most of the scholarship reviewed, we find that the study of classificatory Buddhist systems usually begins with the earliest one that has survived, Bhāvaviveka's Madhyamakahr .dayakārikā.But what about the classification of

Doxographies: Solving a Buddhist Hermeneutical Problem
As Gregory pointed out in his study of Chinese doxographies, the importance of the genre within the Buddhist tradition should not be underestimated, since it responded to a very important hermeneutical problem at the core of the constant emergence of new texts, rituals, and practices. 13opez traces the root of this problem to the Buddha himself and, in particular, to the conundrum presented to his followers in some of the Buddha's last words before his death. 14These words are presented in the poignant scene in the Mahāparanibbanasutta, in which the Buddhist community is witness to the imminent death of the Buddha: Then [the Buddha] addresses the venerable Ānanda: "It may be, Ānanda, that some of you will think.'The word of the Teacher is a thing of the past; we have now no Teacher.'But that, Ānanda, is not the correct view.The Doctrine and the Discipline, Ānanda, which I have taught and enjoined upon you, is to be your teacher when I am gone".

Mahāparanibbanasutta 60 15
For over 40 years, the Buddha had created a community that relied on him not only as their teacher, as the transmitter of the truth he had discovered under the Bodhi tree, but also as its interpreter.Whenever there was a doubt about the possible meaning of his teachings, the community could always confer with the Buddha, and the Buddha could offer a definitive answer that settled any existing doubts.After his death, though, the community, as the Mahāparanibbanasutta makes clear, is not only sad about the departure of their teacher, but they are also deeply worried about how to preserve the meaning of his words and, most importantly, how to retrieve it.While the language of his teachings may have appeared, at times, equivocal, the Buddha always could guarantee the stability of its meaning.
The moment described in the above quote also signals, as Donald Lopez has argued, the very beginning of Buddhist hermeneutics, since, as he pointed out, "[T]hose who are not yet enlightened must interpret". 16On the one hand, the Buddha ensures in the Mahāparanibbanasutta the continuity of his teachings, since "the teaching [ . . .] is to be your teacher".On the other hand, he also opens the door to interpretation, as he does not explain what this statement exactly means.This tension is very important in understanding the development of Buddhism over the next 2500 years, particularly the way in which it has allowed the tradition to preserve a certain degree of unity within an astonishing degree of doctrinal, soteriological, and ritual diversity.
The notion that the teaching is the teacher allows the later Buddhist tradition a great degree of freedom to determine what the teachings of the Buddha are.At the same time, Buddhists over the centuries had to develop a series of interpretative strategies in order to allow the tradition to rein in that plurality of understandings under the umbrella (even if this was a rhetorical one) of the unity of all the teachings of the Buddha.This is not the place to discuss some of the mechanisms that the Buddhist systems found in the earlier Sam .dhinirmocana S ūtra?This s ūtra presents a hierarchical classification, in the form of the Three Turnings of the Wheel, and it is also polemical, arguing the superiority of Yogācāra over other Buddhist teachings.The point here is not to deconstruct the notion of doxography (or siddhānta, or grub mtha') to the point that it is not useful anymore.The goal is just to make us aware of the constructed nature of the term, and the fact that we need to be aware of the fluidity of the category as it was used differently in different historical and cultural contexts.Dalton also makes an interesting point when he says that another important difference between Western doxographies and Indian ones is that "whereas the former are generally philosophical works and restrict themselves to the views held by each school, the tantric classification systems of India [ . . .] are largely concerned with differences in ritual practice", see (Dalton 2005, p. 119).Additionally, for important studies of the doxographical genre in Buddhism, see (Hopkins 1996; Mun 2006; Eimer 1992;  Gregory 1987). 13(Gregory 2002, p. 20). 14 (Lopez 1988, pp. 1-2). 15(Warren 1973, p. 107). 16 (Lopez 1988, p. 9).
Buddhist tradition developed in order to solve this hermeneutical tension, 17 but I want to highlight in this section the important role that doxographies played in articulating key differences between various Buddhist traditions, integrating different doctrines, rituals, and practices within a cohesive system, while organizing those differences hierarchically in order to establish the superiority of certain doctrinal developments in contrast to competing or old ones.
The various functions of the doxographical genre are a reflection of its simultaneous descriptive and prescriptive nature.On the one hand, we can examine these texts and trace the evolution of Buddhism in India, China, as well as Tibet.Doxographies then can be seen as describing a historical process.At the same time, we can see doxographies as a narrative written with clear agendas by various competing Buddhist groups that were attempting not only to organize (sometimes contradictory) ritual and doctrinal systems, but also to prescribe a particular doctrinal or sectarian point of view. 18Doxographies were, from this perspective, tools to create an ideological reality.Being able to see where these two aspects intersect, as well as when they differ, is one of the important roles in our scholarly research of these texts.
A quick overview of the history of Buddhist doxographies in India, China, and Tibet, will help us to better contextualize the contents of Nupchen's Lamp for the Eye in Meditation, as well as his goals when writing it. 19Not all the doxographies discussed in this brief summary had a direct influence on Nupchen's Lamp, but I have included them because I want to highlight the role that doxographies had all over Asia in making sense of the constantly changing Buddhist tradition, as well as its constantly evolving and constructed nature.The goal of this overview is also to show how Nupchen used the doxographical genre not only to passively organize the various forms of Buddhism that were imported into Tibet during his time, but also to actively and creatively engage in the construction of a unique Tibetan Buddhist view.Nupchen was, ultimately, a deeply innovative thinker working within a very traditional framework.

Indian Precedents
In India, doxographies were originally used as an intellectual tool to define Buddhist positions against non-Buddhist traditions. 20The emergence over time of discrepancies in the interpretation of certain Buddhist ideas forced doxographies to eventually incorporate various intra-sectarian doctrinal issues.In the earlier form, doxographies were articulated around the discussion of particular doctrinal or philosophical views in which the Buddhist view always prevailed against non-Buddhist positions.When Buddhist doxographies shifted mainly to discussions of different Buddhist philosophical views (with non-Buddhist traditions being peripheral to the main argument of the text or not even included), we see a shift in the genre to a hierarchical organization of those diverse philosophical 17 Regarding some of the most important hermeneutic strategies developed by Buddhism to solve this tension, see (Lopez 1988). 18For an understanding of history as narrative, see (Jenkins 2003). 19Here, I will not discuss Japanese doxographies since they did not have a direct influence in the Tibetan world, which is the focus of my article.The doxographical genre, nonetheless, also played a major role in the process of assimilation of Buddhism in Japan.In K ūkai's works (K ūkai 1972, 774-835 CE), for example, we see the use of doxography as a way to articulate the superiority of the Shingon tradition over the Tendai school.For K ūkai's use of doxography, see (Hakeda 1972).
In his discussion of K ūkai and Shingon, McMullen also argues that "esoteric Buddhist scholasticism emphasized the tools of doxography, taxonomy, and lineage in an effort to explicate the secret teachings of the buddha.However, which texts, teachings, and lineages were 'esoteric' has always been a matter of debate."See (Mcmullen 2016).K ūkai's contemporary and founder of the Tendai school, Saicho (767-822 CE), will also use doxographical classifications in order to argue the superiority of Tendai over that of rival schools.I want to thank Prof. Paul Groner for his insights on the use of doxographies in Japan. 20Although there were also non-Buddhist examples such as Śa ṅkarācārya's Brahmas ūtrabhās .ya, which "sets forth the views of nāstika (heterodox) and āstika (orthodox) schools and shows the weaknesses and strengths in each as a strategy to demonstrate the superiority of Śa ṅkara's own Advaita Vedān .ta philosophy.None of these Indian works were written simply as informative textbooks about the tenets of different Indian schools of thought.They instead have clear polemical agendas: namely, demonstrating the superiority of their own position, and showing how the lesser philosophies are either a hindrance or a stepping stone to their own philosophy, as revealed by the Buddha in the case of Buddhist siddhānta, and by the Vedas in the case of non-Buddhists."See Śa ṅkarācārya entry in (Buswell et al. 2014).
positions.Doxographies, as a literary genre, offer Buddhism a structure in which diversity could be interpreted and organized, arguing for the unity of the Buddhist path, while allowing space for new and diverse interpretations.For reasons of space, I will only refer to the most important examples of the doxographical genre in India, mainly those of Bhāvaviveka, Śāntaraks .ita, Buddhaguhya, and Vilāsavajra, which will give us a sense of the evolution of late Indian tantra and the attempts to organize new ritual systems of practice into cohesive, clearly organized, hierarchical schemes.212.2.1.Bhāvaviveka's Madhyamakahr .dayakārikā (500-570 CE) Bhāvaviveka (500-570 CE)'s Madhyamakahr .dayakārikā22 with its auto-commentary, the Tarkajvālā, is traditionally considered by scholars to be the earliest extant Indian Buddhist doxography (although, as we will see, there were earlier Buddhist doxographies in China).He is believed to have been influenced by Indian scholar and grammarian, Bhartr .hari. 23The Madhyamakahr .dayakārikā offers a detailed discussion of the most important schools of thought, Buddhist as well as non-Buddhist, that were influential during Bhāvaviveka's time. 24he structure of the work does not follow the hierarchical organization that will be common in later doxographical systems.Bhāvaviveka seems more concerned with arguing specific doctrinal points against non-Buddhist traditions than in ranking those traditions hierarchically.Nonetheless, we already find in the Madhyamakahr .dayakārikā vigorous internal Buddhist debate between the Svātantrikas, of which Bhāvaviveka is considered one of the earliest proponents, and the Prāsa ṅgikas, whose position was defended by Buddhapālita. 25lthough later Buddhist doxographies will differ in content as well as in structure, there is no doubt that we find in Bhāvaviveka the foundation for a new way of thinking within the Indian Buddhist tradition.Doxographies will be used, from then on, to argue within the Buddhist tradition in the same way that had been used in the past to argue against the doctrinal issues of other religious traditions.Buddhism will gradually become more concerned about the internal structure and importance of its own teachings than in arguing against other traditions.262.2.2.Śāntaraks .ita's Tattvasa ṅgraha (725-788 CE)   Śāntaraks .ita's Tattvasa ṅgraha 27 can be considered one of the earlier (if not the earliest) Indian doxographical systems to make it onto the Tibetan plateau.Śāntaraks .ita is considered, after all, the first abbot of Samyé monastery and had a very important role in the introduction of the monastic and, by extension, scholastic Buddhist tradition into Tibet. 28Śāntaraks .ita's work, especially as transmitted into Tibet by his main disciple, Kamalaśīla, 29 will be enormously influential to Nupchen, especially in his understanding of what he labeled as "the Gradual Vehicle" in the Lamp for the Eye in Meditation. 30āntaraks .ita's Tattvasa ṅgraha, following Bhāvaviveka's Madhyamakahr .dayakārikā, does not offer a hierarchical classification of philosophical views, but presents an overview of important philosophical concepts and the way they were understood by various Buddhist, as well as non-Buddhist schools, like the Nyāya, Mīmām .sā, Sām .khya, Jainas, and the Aupanis .adikas, as well as of some Buddhist traditions like the Vātsīputrīya. 31n retrospect, we can see the Tattvasa ṅgraha as part of a transitional period in the history of Buddhism.Śāntaraks .ita's Tattvasa ṅgraha reflects a Buddhism in which external debates still overshadow internal Buddhist tensions.It is also important to point out the lack of any meaningful discussion of tantric systems, especially since these systems are thriving during this period and will become the center of any doxographical discussion in the upcoming decades.
2.2.3.Buddhaguhya's Commentaries to the Mahāvairocana-Abhisam .bodhi Tantra (Eighth Century) In the eighth century, the Indian scholar Buddhaguhya wrote two different doxographical classifications that can be found in his two commentaries on the Mahāvairocana-abhisam .bodhi Tantra, 32 and are structured around a twofold system that differentiated Mahāyāna and tantric teachings.He also included a division within the tantric teachings of Kriyā and Yoga Tantras. 33ith Buddhaguhya, we clearly see that Buddhist doxographical classifications are taking a new direction from the one we have seen in the works of Bhāvaviveka and Śāntaraks .ita.We see this in five main areas: First, Buddhaguhya's writings are not as concerned in arguing against non-Buddhist views as in trying to make sense of the enormous variety of new Buddhist teachings being produced during his time.Second, he also proposes a clear separation between Mahāyāna teachings and the new cycles of tantric teachings being produced in India.Third, he is one of the first Buddhist scholars to focus mainly on tantra, making all other previous Buddhist teachings merely introductory or peripheral to the new tantric cycles.Fourth, we can also see a change of attention from doctrine to ritual concerns, which Dalton sees as the central feature of Tantric doxography. 34Even if we argue for the profound doctrinal relevance of these new ritual vehicles, the new classification will pay attention to ritual aspects in order to organize them.Fifth, Buddhaguhya classifies the new tantras around a dichotomy that will become very important in later tantric thought: external vs. internal tantras.This 29 Kamalaśīla (740-95), Śāntaraks .ita's most important student, wrote an important commentary to Śāntaraks .ita's Tattvasa ṅgraha, the Tattvasam .grahapañjikā, although he did not add any major doctrinal changes to Śāntaraks .ita's text.He became a relevant figure in Tibetan history due to his role in the so-called Samyé Debate in which he defended the views of the gradualist tradition, represented by his teacher Śāntaraks .ita, against the 'Sudden Approach', represented by Chan and defended by the Chinese monk known as Hashang Mahāyāna.The historicity of the debate has been called into question by some scholars, see (Van der Kuijp 1984, ; Ruegg 1989, who called it a "dehistoricized topos"), but there is no doubt that the narrative about the debate framed what was considered an important tension in early Tibetan Buddhism, mainly the methods by which to achieve enlightenment.Some important sources for the debate are (Demiéville 1952; Tucci 1956; Houston 1980; Faber 1986).On the issue of the Samyé Debate and its connection to doxographical classifications (see Ruegg 1989). 30Ruegg provides the following description of Śāntaraks .ita's classification: Chapters 1-3 offer discussion of the Sām .khya prakr .ti and theory of causation and discuss the concept of God; Chapter 4 examines the doctrine of a world endowed with own being; Chapter 5 is an examination of the theory of śabdabrahman; Chapter 6 explores the theory of purus .a; in Chapter 7, Śāntaraks .ita examines the views of Nyāya, Mīmām .sā, Sām .khya, Jainas, and the Aupanis .adikas, as well as the Buddhist Vātsīputrīya's pudgala ideas; Chapter 8 is an analysis of the doctrines of permanent and stable entities; Chapter 9 deals with Karma and the result of actions; Chapter 10 outlines the six categories (padārtha) of Substance (dravya); Chapter 11 is devoted to the notion of Quality (gun , a); Chapter 12 focuses on Action (karma); and the Chapter 13 discusses the notion of the Universal (sāmānya) (see Ruegg 1989). 31(Jha 1937). 32(Hodge 2003, p. 4). 33To be more specific, Buddhaguhya divides the Buddhist teaching in two: the teachings of the Mahāyāna, focused on the practice of the perfections (pāramitā), and the teachings of Tantra, focused on mantra recitation, which Buddhaguhya subdivides into Kriyā (focused on outward practice or "objective supports") and Yoga Tantras (focused on inward techniques or on the "practice of the profound and vast").See (Hodge 2003; Dalton 2005, pp.121-24).
dichotomy, in which "exterior" practices are inferior to "interior" ones, will help organize new tantras in a hierarchical way. 352.2.4.Vilāsavajra's Spar Khabs (Eighth Century) In the eighth century, the Indian Master Vilāsavajra wrote commentaries to the Chanting the Names of Mañjusrī (Skt.Mañjuśrīnāmasam .gīti, Tib.'jam dpal mtshan brjod) and the Guhyagarbha Tantra, in which he offered two different doxographical classifications.The one in the Mañjuśrīnāmasam .gīti adds an intermediate category, Caryā, between the Kriyā and Yoga divisions found in Buddhaguhya's commentary to the Mahāvairocana-abhisam .bodhi. 36However, it is the classification in his commentary to the Guhyagarbha Tantra (spar khabs) that deserves special attention, since it has some characteristics that will be very influential for later Tibetan thought.In fact, Vilāsavajra will be one of the main references, together with the ninth century Tibetan scholar, Pelyang, informing Nupchen's understanding of the Mahāyoga tradition.
There are several important aspects to his classification.First, although Vilāsavajra includes some non-Buddhist vehicles, their inclusion is clearly not as relevant as in Bhāvaviveka's or Śāntaraks .ita's systems.They seem to simply be included in order to add legitimacy to the later Buddhist systems.Second, we can see the new relevance of the Mahāyoga system with which the Guhyagarbha was first associated.The Guhyagarbha and, especially, its 13th chapter will be highly influential in the development of later tantric systems.Third, and this will be important for our later discussion of Nupchen's Lamp for the Eye in Meditation, we find in Vilāsavajra some of the earliest references to the Great Perfection.Although Vilāsavajra discusses Atiyoga as separate from Mahāyoga, Dalton thinks that "in practical terms [Atiyoga] is still dependent on it". 37In the spar khabs, Atiyoga seems to be a separate ritual moment that emerges from Mahāyoga, but not a different vehicle with its separate textual tradition and contemplative practices, as we will see later with Nupchen.Fourth, we also see something that will be very relevant for the later New Schools (gsar ma) traditions, namely, the organization of these tantric vehicles around gender categories: "the male tantras focused primarily on method, the female tantras focusing on wisdom, [plus a category of] neuter tantras". 38Fifth, the fact that we can find two different classification systems in Vilāsavajra's writings (something we will also see in Nupchen's corpus) can point to a certain fluidity in the use of these categories during this period, but we can also see how tantra is taking shape in a way that will be clearly recognizable in the later Tibetan tradition.

Chinese Precedents
Tibetan doxographies, due to their interest in Tantric systems, followed Indian models for the most part, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, 35 In a very different context this inward vs. outward categorization found an echo in the earlier denigration by the Mahāyāna of all the early Buddhist traditions, using the label Hinayāna, or inferior vehicle, to refer to it.The use of these categories in order to articulate ideas can be very powerful since they affect the way we look at those earlier teachings.There is no doubt that earlier Buddhist practitioners did not consider their early practices and rituals as merely outwards and were, probably, a complete soteriological path of their own. 36(Dalton 2005, pp.124-25). 37(Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170). 38(Dalton 2005, p. 128). 39 (Dalton and van Schaik 2003; Meinert 2003). 40 (Dalton 2005, p. 116).
in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan ( xographies, due to their interest in Tantric systems, followed Indian models for the he unusual fourfold classification found in Nupchen's Lamp has made some scholars ill see, a possible Chinese influence. 39While in India and Tibet, classification systems around doctrinal or ritual issues (especially with the development of tantra), a acteristic of classification systems in China was that they "were usually arranged es of the Buddha's life, often with the aim of promoting a particular text over all ously, in India and Tibet, texts or cycles of texts were important in doxographical ut at least rhetorically, they were not at the center of these doctrinal battles.I will iscussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan ed and composed some of his main works in China), whose doxographies could be some similarities to the one composed by Nupchen.
(慧觀)'s Five Periods Teachings (五時教) (d.453)   here are some earlier doxographical classification systems in China prior to the one mārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely f organizing the Buddhist teachings that was particularly influential for the later distinctive characteristics of Huiguan's system is that he proposes a division of the ngs into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra ra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he e different periods, each one associated with a particular textual tradition.The final oks as follows: 43 aching (dun jiao 頓教): corresponds to the Huayan Sūtra aching (jian jiao 漸教) ree vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha The Four Noble Truths to the Śrāvakas p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term d does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170) ) and Chegwan ( Tibetan doxographies, due to their interest in most part, but the unusual fourfold classification suggest, as we will see, a possible Chinese influenc were organized around doctrinal or ritual issue distinctive characteristic of classification systems around narratives of the Buddha's life, often wit others". 40Obviously, in India and Tibet, texts or classifications, but at least rhetorically, they were include a brief discussion of two of the scholars fro (諦觀) (who lived and composed some of his mai seen as having some similarities to the one compo 2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五 Although there are some earlier doxographic presented by Kumārajīva's disciple Huiguan (慧觀 under the name of the Five Periods Teachings (五 Chinese way of organizing the Buddhist teachin tradition. 42ne of the distinctive characteristics of Huig Buddhist teachings into sudden approaches (dun (Avataṃsaka Sūtra) and are also identified with Ch divides into five different periods, each one assoc classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): correspond 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lo The Four Noble Truths to the Ś 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsav rdzogs chen, and does not employ it in any specific te 38 (Dalton 2005, p. 128).
) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.

Huiguan (
an also see how tantra is taking shape in a way that tradition. t in Tantric systems, followed Indian models for the n found in Nupchen's Lamp has made some scholars nce.39  While in India and Tibet, classification systems ues (especially with the development of tantra), a s in China was that they "were usually arranged ith the aim of promoting a particular text over all or cycles of texts were important in doxographical re not at the center of these doctrinal battles.I will rom China and Korea, Huiguan (慧觀) and Chegwan ain works in China), whose doxographies could be posed by Nupchen.
五時教) (d.453) hical classification systems in China prior to the one 觀) (d.453), 41 scholars seem to agree that his system, 五時教), was one of the first to articulate a uniquely hings that was particularly influential for the later iguan's system is that he proposes a division of the n jiao 頓教), which correspond to the Huayan Sūtra han, and gradual teachings (jian jiao 漸教), which he ociated with a particular textual tradition.The final ds to the Huayan Sūtra Lotus Sūtra) were taught separately by the Buddha e Śrāvakas avajra "does not give any special precedence to the term c technical sense."See (Van Schaik 2004, p. 170).
)'s Five Periods Teachings ( Tibetan doxographies, due to their interest in Tantric systems, followed Indian models for the most part, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (諦觀) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.
2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453)   Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan ( rimarily on method, the female tantras focusing on wisdom, [plus a category of] Fifth, the fact that we can find two different classification systems in Vilāsavajra's ng we will also see in Nupchen's corpus) can point to a certain fluidity in the use s during this period, but we can also see how tantra is taking shape in a way that ognizable in the later Tibetan tradition.ents graphies, due to their interest in Tantric systems, followed Indian models for the unusual fourfold classification found in Nupchen's Lamp has made some scholars l see, a possible Chinese influence.39  While in India and Tibet, classification systems round doctrinal or ritual issues (especially with the development of tantra), a teristic of classification systems in China was that they "were usually arranged of the Buddha's life, often with the aim of promoting a particular text over all sly, in India and Tibet, texts or cycles of texts were important in doxographical t at least rhetorically, they were not at the center of these doctrinal battles.I will cussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan and composed some of his main works in China), whose doxographies could be me similarities to the one composed by Nupchen.觀)'s Five Periods Teachings (五時教) (d.453)   re are some earlier doxographical classification systems in China prior to the one ārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, f the Five Periods Teachings (五時教), was one of the first to articulate a uniquely rganizing the Buddhist teachings that was particularly influential for the later istinctive characteristics of Huiguan's system is that he proposes a division of the s into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra ) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he ifferent periods, each one associated with a particular textual tradition.The final s as follows: 43 hing (dun jiao 頓教): corresponds to the Huayan Sūtra hing (jian jiao 漸教) e vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha The Four Noble Truths to the Śrāvakas 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
) (d. 453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings ( Tibetan doxographies, due to their interest in Tantric systems, followed Indian models for the most part, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (諦觀) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.
2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453)   Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings ( 五時教 ), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao Tibetan doxographies, due to their interest in Tantric systems, followed Indian models for the most part, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (諦觀) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.
2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453)   Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
), which correspond to the Huayan S ūtra (Avatam .saka S ūtra) and are also identified with Chan, and gradual teachings (jian jiao Tibetan doxographies, due to their interest in Tantric systems, followed Indian models for the most part, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (諦觀) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.
2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453)   Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.
Sudden Teaching (dun jiao gs (something we will also see in Nupchen's corpus) can point to a certain fluidity in the use se categories during this period, but we can also see how tantra is taking shape in a way that e clearly recognizable in the later Tibetan tradition.
inese Precedents ibetan doxographies, due to their interest in Tantric systems, followed Indian models for the art, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars st, as we will see, a possible Chinese influence.39  While in India and Tibet, classification systems organized around doctrinal or ritual issues (especially with the development of tantra), a ctive characteristic of classification systems in China was that they "were usually arranged d narratives of the Buddha's life, often with the aim of promoting a particular text over all ".40  Obviously, in India and Tibet, texts or cycles of texts were important in doxographical ications, but at least rhetorically, they were not at the center of these doctrinal battles.I will e a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (who lived and composed some of his main works in China), whose doxographies could be s having some similarities to the one composed by Nupchen.
Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453) lthough there are some earlier doxographical classification systems in China prior to the one ted by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, the name of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely se way of organizing the Buddhist teachings that was particularly influential for the later on.42 ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the ist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra ṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he s into five different periods, each one associated with a particular textual tradition.The final ication looks as follows: 43 udden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra radual Teaching (jian jiao 漸教) .1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas ): corresponds to the Huayan S ūtra 2.
Gradual Teaching (jian jiao e in Nupchen's corpus) can point to a certain fluidity in the use d, but we can also see how tantra is taking shape in a way that ter Tibetan tradition.
heir interest in Tantric systems, followed Indian models for the classification found in Nupchen's Lamp has made some scholars inese influence.39 While in India and Tibet, classification systems or ritual issues (especially with the development of tantra), a ation systems in China was that they "were usually arranged life, often with the aim of promoting a particular text over all Tibet, texts or cycles of texts were important in doxographical lly, they were not at the center of these doctrinal battles.I will he scholars from China and Korea, Huiguan (慧觀) and Chegwan me of his main works in China), whose doxographies could be he one composed by Nupchen.scholars seem to agree that his system, Teachings (五時教), was one of the first to articulate a uniquely ddhist teachings that was particularly influential for the later ristics of Huiguan's system is that he proposes a division of the proaches (dun jiao 頓教), which correspond to the Huayan Sūtra tified with Chan, and gradual teachings (jian jiao 漸教), which he each one associated with a particular textual tradition.The final ): corresponds to the Huayan Sūtra ) ribed in the Lotus Sūtra) were taught separately by the Buddha Truths to the Śrāvakas ds that Vilāsavajra "does not give any special precedence to the term n any specific technical sense."See (Van Schaik 2004, p. 170).rt 2003). ).
The three vehicles (as described in the Lotus S ūtra) were taught separately by the Buddha 2.1.1.The Four Noble Truths to the Śrāvakas 2.1.2.Interdependent Origination to the Pratyekabuddhas 2.1.3.The Six Perfections to the bodhisattvas.

2.2.
The Buddha taught the common teaching of the three vehicles; mainly, the Perfection of Wisdom.2.3.
In the third period, the Buddha made clear the inferiority of the teachings of the Śrāvakas and elevated the teachings of the bodhisattvas.This is mainly found in the Vimalakīrti Sūtra.

2.4.
During this period, the three vehicles were subsumed under one vehicle.This teaching was presented in the Lotus S ūtra.

2.5.
The final period, as presented in the Nirvān .a S ūtra, taught the eternality of the Buddha.
Huiguan is considered to be the first to establish the polemical notions of gradual and sudden as classification categories to organize the various Buddhist teachings, 44 a reflection of a debate that will be part of the Buddhist intellectual life in China in the following centuries.Even though Huiguan's and Nupchen's understandings of the notion of 'sudden' are quite different, one of the aspects that makes a discussion of Huiguan's classification relevant to our discussion of Nupchen's Lamp is that we will see how this tension between sudden and gradual also plays a central role in Nupchen's doxographical classification.In the Lamp we do not only see a clear distinction between the Indian gradual and the Chinese sudden systems, but even within Mahāyoga there is a distinction between gradual and sudden approaches. 41(Mun 2006; Wu and Wilkinson 2017). 42 (Gregory 2002, p. 111). 43 (Gregory 2002, pp.111-12). 44(Gregory 2002, p. 113).

Chegwan (
suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (諦觀) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.

Huiguan (慧觀)'s Five Periods Teachings (五時教) (d. 453)
Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
)'s Four Types of Content of Buddhist Teachings (化法四教) (d.971) The 10th century classification, the Four Types of Content of Buddhist Teachings (化法四教), developed by the Korean monk Chegwan ( Tibetan doxographies, due to their interest in Tantric systems, followed Indian m most part, but the unusual fourfold classification found in Nupchen's Lamp has made s suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classifica were organized around doctrinal or ritual issues (especially with the development distinctive characteristic of classification systems in China was that they "were usua around narratives of the Buddha's life, often with the aim of promoting a particular others". 40Obviously, in India and Tibet, texts or cycles of texts were important in d classifications, but at least rhetorically, they were not at the center of these doctrinal b include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) a (諦觀) (who lived and composed some of his main works in China), whose doxograph seen as having some similarities to the one composed by Nupchen.2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453)  Although there are some earlier doxographical classification systems in China pri presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree tha under the name of the Five Periods Teachings (五時教), was one of the first to articulat Chinese way of organizing the Buddhist teachings that was particularly influential tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a di Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the H (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教 divides into five different periods, each one associated with a particular textual traditi classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by 2.1.1.The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special preceden rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 1 38 (Dalton 2005, p. 128).
) (d. 971), was not particularly influential in the Chinese tradition, but it is important that we briefly discuss it here, since it has been mentioned as a possible direct influence on Nupchen's Lamp. 45Chegwan, just like Nupchen, offers a fourfold scheme that includes, in the two lower levels, the Gradual and the Sudden, while the two upper levels include what he calls Secret and Indeterminate (see Table 1).will be clearly recognizable in the later Tibetan tradition.

Chinese Precedents
Tibetan doxographies, due to their interest in Tantric systems, followed Indian models for the most part, but the unusual fourfold classification found in Nupchen's Lamp has made some scholars suggest, as we will see, a possible Chinese influence. 39While in India and Tibet, classification systems were organized around doctrinal or ritual issues (especially with the development of tantra), a distinctive characteristic of classification systems in China was that they "were usually arranged around narratives of the Buddha's life, often with the aim of promoting a particular text over all others". 40Obviously, in India and Tibet, texts or cycles of texts were important in doxographical classifications, but at least rhetorically, they were not at the center of these doctrinal battles.I will include a brief discussion of two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan (諦觀) (who lived and composed some of his main works in China), whose doxographies could be seen as having some similarities to the one composed by Nupchen.
2.3.1.Huiguan (慧觀)'s Five Periods Teachings (五時教) (d.453)   Although there are some earlier doxographical classification systems in China prior to the one presented by Kumārajīva's disciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, under the name of the Five Periods Teachings (五時教), was one of the first to articulate a uniquely Chinese way of organizing the Buddhist teachings that was particularly influential for the later tradition. 42ne of the distinctive characteristics of Huiguan's system is that he proposes a division of the Buddhist teachings into sudden approaches (dun jiao 頓教), which correspond to the Huayan Sūtra (Avataṃsaka Sūtra) and are also identified with Chan, and gradual teachings (jian jiao 漸教), which he divides into five different periods, each one associated with a particular textual tradition.The final classification looks as follows: 43 1.Sudden Teaching (dun jiao 頓教): corresponds to the Huayan Sūtra 2. Gradual Teaching (jian jiao 漸教) 2.1.The three vehicles (as described in the Lotus Sūtra) were taught separately by the Buddha 2.1.1.
The Four Noble Truths to the Śrāvakas 37 (Dalton 2005, p. 130).Van Schaik adds that Vilāsavajra "does not give any special precedence to the term rdzogs chen, and does not employ it in any specific technical sense."See (Van Schaik 2004, p. 170).
) Sudden (ston mun cig car 'jug pa) is period, but we can also see how tantra is taking shape in a way that in the later Tibetan tradition.
due to their interest in Tantric systems, followed Indian models for the ourfold classification found in Nupchen's Lamp has made some scholars sible Chinese influence. 39While in India and Tibet, classification systems ctrinal or ritual issues (especially with the development of tantra), a classification systems in China was that they "were usually arranged ddha's life, often with the aim of promoting a particular text over all ia and Tibet, texts or cycles of texts were important in doxographical hetorically, they were not at the center of these doctrinal battles.I will two of the scholars from China and Korea, Huiguan (慧觀) and Chegwan osed some of his main works in China), whose doxographies could be ities to the one composed by Nupchen.
Periods Teachings (五時教) (d.453) e earlier doxographical classification systems in China prior to the one isciple Huiguan (慧觀) (d.453), 41 scholars seem to agree that his system, Periods Teachings (五時教), was one of the first to articulate a uniquely the Buddhist teachings that was particularly influential for the later haracteristics of Huiguan's system is that he proposes a division of the den approaches (dun jiao 頓教), which correspond to the Huayan Sūtra lso identified with Chan, and gradual teachings (jian jiao 漸教), which he eriods, each one associated with a particular textual tradition.The final s: 43 jiao 頓教): corresponds to the Huayan Sūtra jiao 漸教) (as described in the Lotus Sūtra) were taught separately by the Buddha Noble Truths to the Śrāvakas chaik adds that Vilāsavajra "does not give any special precedence to the term ploy it in any specific technical sense."See (Van Schaik 2004, p. 170).

3; Meinert 2003). son 2017).
) Gradual (tsen men rim gyis 'jug pa) Although there are important differences in their understandings of each of those categories, what is relevant to us is the possible structural borrowing of them by Nupchen.This would explain the obvious difference of the Lamp's classification when compared with late Indian systems, which, as we saw, were more concerned with tantric developments than with doctrinal and philosophical issues.The fourfold doctrinal system and the centrality of the tension between sudden and gradual approaches may indicate a certain amount of Chinese influence in the work, and the inclusion of Chinese Chan as one of the Buddhist traditions discussed in the Lamp definitely indicates that Nupchen was quite familiar with the Chinese Buddhist tradition.
There are some issues when trying to suggest some direct borrowing by Nupchen of Chegwan's system.First, Chegwan wrote his classification system at the end of the 10th century in Southern China, and his text did not circulate widely at that time, so it would be difficult to prove that it influenced Nupchen, even though both composed their texts around the same time.Second, the four categories within Chegwan's system are related to methods of teaching and not to actual teachings, as they are in Nupchen's classification.Despite the problematic attribution of direct influence of Chegwan's system over Nupchen's classification, there are enough interesting correlations in both systems that further research on this issue may be needed to settle this point.46

Early Tibetan Classifications
It is now time to look at some of the early Tibetan classification systems that preceded the Lamp and that show the way Tibetans were assimilating and reinterpreting late India tantric developments before Nupchen composed his doxography.For the most part, Tibetans will follow Indian doxographical models concerned with late Indian tantric developments, like the ones by Buddhaguhya and Vilāsavajra.Nupchen was also very much part of this process of assimilation of late Indian tantric Buddhism, 47 but he was also an exception, and proposed in the Lamp an alternative classification system that did not follow its Indian precedents and had more in common with Chinese doxographical (panjiao) systems.
One of the main differences between the Tibetan and the Indian approaches to the doxographical genre is that while Indian systems were more concerned with organizing and classifying new tantric ritual developments, Tibetans were more interested in doctrinal issues. 48This may have to do with the foreign nature of Buddhism in Tibet.Another important aspect of the early Tibetan doxographies is that they show Buddhism in Tibet in a process of creative transformation, before the later Ancient (rnying ma) and New (gsar ma) schools settled for standard nine and four vehicle systems, respectively, during the later centuries. 49wo of the most important precedents to Nupchen's Lamp were Padmasambhava's Garland of the Views of the Secret Instructions (man ngag lta ba'i phreng ba) and Pelyang's Lamp for the Mind (thugs kyi sgron ma). 50The former is important for its connection to the early developments of the Great Perfection tradition as a separate vehicle, an important aspect of Nupchen's Lamp.The latter one will have an important impact on Nupchen's understanding of the Mahāyoga tradition.Nupchen quotes from both of them repeatedly in the Lamp.

Padmasambhava's Garland of the Views of the Esoteric Instructions (man ngag lta ba'i phreng ba)
The importance of the Garland of the Views of the Esoteric Instructions ('the Garland' from now on) for our understanding of the early development of Buddhism in Tibet has been extensively discussed by several scholars. 51There are a few aspects that make this text relevant: it is considered to be the earliest Tibetan doxography; 52 its seven vehicle classification is a clear Tibetan precedent to the nine vehicle system later adopted by the Nyingma tradition; and finally, the Garland is also seen as very influential in the formation of the early Great Perfection tradition.The fact that the text is quoted in Nupchen's Lamp testifies to its old pedigree, and although Nupchen does not follow the seven-vehicle classification of the text, he uses it as a source for his understanding of tantric contemplative practices.
The Garland shares with Vilāsavajra's spar khab the fact that both are a commentary of Chapter 13 of the Guhyagarbha Tantra.This complex and obscure chapter proved to be very fertile ground for the development of tantric thought and lead to the early speculations that would be the seeds for the Great Perfection.The text offers a sevenfold classification that, as we can see, follows earlier Indian developments, as discussed in our previous section.The Garland divides all Buddhist views into three main vehicles: the S ūtra, the Outward Tantra, and the Inner Tantra: S ūtra 1. Śrāvaka

Pratyekabuddha
48 (Dalton 2005, p. 116). 49"In Tibetan, 'Ancient', is the name of one of the four major sects of Tibetan Buddhism.The name derives from the sect's origins during the 'early dissemination' (snga dar) of Buddhism in Tibet and its reliance on translations of Tantras made during that period; this is in distinction to the new (gsar ma) sects of Bka' brgyud, Sa skya, and Dge lugs, all of which arose during the later dissemination (phyi dar) and make use of newer translations.The Rnying ma is thus 'ancient' in relation to the new sects and only began to be designated as such after their appearance."(Buswell et al. 2014). 50There are other doxographies, both from the eighth century: Yéshé Dé (ye shes sde)'s Differentiating the Views (lta ba'i khyad par) and Kawa Peltsek (ska ba dpal brtsegs)' Esoteric Instructions of the Stages of the View (lta ba'i khyad pa'i man ngag) that I will not be discussing here for lack of space and because they are less relevant for my argument.For Yéshé Dé's text, see (Ruegg 1981).Van Schaik also has argued that Kawa Peltsek's text may not be by him and of later composition, making it irrelevant for our discussion.See (Van Schaik 2004, p. 188). 51See (Karmay 1988; Dalton 2005), and Germano in two of his unpublished manuscripts, The Secret Tibetan History of Buddhist Tantra in Ancient Tibet and Mysticism and Rhetoric in the Great Perfection. 52Dalton claims that the text is "our earliest extant text entirely devoted to setting forth a tantric classification system" (Dalton 2005, p. 132).According to Dalton, there is substantial evidence for the attribution of the text to Padmasambhava, see (Dalton 2005, p. 132, n. 41).Ruegg thinks that the first Tibetan doxography is the lta ba khyad par, see (Ruegg 1981) Here, as we saw in the late Indian tradition developments, Yoga Tantra is pushed into the Outward Tantra, showing that during this period the system is still very much in play, with newer developments rapidly pushing aside (or outside, or lowering, depending on the image the text wants to present) older ones.What we find here, though, at least in regard to Inward Tantra, is not yet a system of separate vehicles, but a blueprint for a gradual, integrated system of meditation that outlines the different stages of contemplative practice.The Great Perfection tradition during this period is still part of a broader system of ritual practice and contemplation. 542.4.2.Pelyang's Lamp for the Mind (thugs kyi sgron ma, Late Eighth-Early Ninth Centuries) 55 Pelyang's doxographical system also shows clear Indian precedents (particularly Buddhaguya), although, unlike the Garland, his main concern is the Mahāyoga tradition.As we have said, it is very difficult to know if this classification is earlier or later than the Garland, but, nonetheless, an examination of these two texts, which were probably composed very close to each other in time, shows that Tantra is undergoing a very creative period, with new developments continuously emerging and with various classification systems trying to articulate this evolution in some sort of cohesive way.
Pelyang's Lamp for the Mind includes, again, non-Buddhist as well as Buddhist philosophical views. 56We can also see some Chinese influence in Pelyang's system in his inclusion of a vehicle of Gods and humans, which is common in many early Chinese doxographies.Dalton also points out the fact that while the Garland had mainly a ritual focus, following Indian models, Pelyang's focus is doctrinal, showing a new Tibetan sensibility that seems to pave the way for the philosophical approach of Nupchen's Lamp. 57e can see Pelyang's influence on Nupchen's Lamp on two different levels.On a more general one, Nupchen's Lamp shows the doctrinal concern of Pelyang in his presentation of the four different 53 (Dalton 2005, pp.124-31). 54(Dalton 2005, p. 88). 55For an excellent study of Pelyang's life and works, see (Takahashi 2009). 56The final classification looks like this: (1) Gods and humans, (2) Śrāvakas, (3) Pratyekabuddhas, (4) Bodhisattvas, (5) Tantras (these divided in (a) Kriyā, (b) Upāya, (c) Yoga, (d) Mahāyoga).See (Dalton 2005, p. 137). 57Takahashi agrees with this portrayal of Pelyang's classification: "The first half of the Lamp of the Mind is a Buddhist doxographical presentation of non-Buddhist and Buddhist systems, refuted or criticized in the standard hierarchical progression of lower to higher systems, ending with a summation of the highest system.In this light Dpal dbyangs's work is nothing out of the ordinary.However, while most Indian Buddhist tantric doxographies center on distinctions in the rituals of the various Tantras, Dpal dbyangs's Lamp of the Mind hardly mentions ritual.The distinctions that concern Dpal dbyangs are exclusively related to view; he does not comment on the role of ritual in distinguishing whether a system is effective.Furthermore, although later Tibetan doxographies resemble Dpal dbyangs's Lamp of the Mind in relying upon view and perspective, rather than practice or ritual for their rankings, it appears that Dpal dbyangs's Lamp of the Mind is among the first, and perhaps the oldest, extant Tibetan doxography to do so.Once again, Dpal dbyangs appears to have been an innovator among Tibetans in this field."(Takahashi 2009, p. 104).
vehicles and, in fact, he has no interest in the ritual aspects of the various traditions he discusses.
A more direct influence can be seen in Nupchen's particular presentation of the Mahāyoga tradition, which is strongly influenced by Pelyang's works, as these are cited frequently in the Lamp.Nupchen, just like Pelyang, will avoid, for the most part, any discussion of the sexual or violent rhetoric of the tradition and instead will focus on its more philosophical aspects.

Nupchen Sangye Yeshe: Writing the Lamp for the Eye in Meditation to Bring Light into the Darkness
Dalton has pointed out that "Chinese panjiao systems [ . . .] reflect cultural interests and anxieties that were uniquely Chinese.The panjiao often tells us less about the Indian Buddhist teachings that they organize than about the Chinese concerns that were at stake in the Sino-Indian encounters." 58This will also be the case in Tibet, which is why, if we want to understand Nupchen's Lamp for the Eye in Meditation and its importance in the establishment of a unique understanding of the Buddhist path in Tibet, it is important to look at the biographical and historical context in which it was written.
We are not sure about Nupchen's dates, although we can roughly situate him between the mid-9th and mid-10th centuries. 59Various accounts present us with a figure deeply engaged in the intellectual Buddhist world of the 9th-10th centuries, traveling to India, Nepal, and Central Asia from a very young age, learning from a wide variety of teachers, collecting and translating texts, and composing commentaries and treatises on some of the most cutting-edge Buddhist literature of the period, such as that of the Anuyoga and the Great Perfection traditions.He was also central in articulating a coherent vision for the Great Perfection drawing from the earliest textual corpus of the tradition, the Eighteen Texts of the Mind (Tib.Sems sde bco rgyad). 60He was also a unique witness to the political events that unraveled after the collapse of the Tibetan Empire in the mid-ninth century.One of the most famous stories about him involves a (probably) fictitious encounter with Langdarma (glang dar ma), who is traditionally portrayed in Buddhist sources as responsible for the persecution of Buddhism in the ninth century that almost erased the Buddhist tradition from the Tibetan plateau.In a short biographical account by the 18th century Tibetan scholar Guru Tashi (gu ru bkra shis) the encounter is described as follows (the translation is mine): At the time when King Langdarma was destroying the teachings of the Buddha [Nupchen]  scared this evil king.The king asked him: "What powers do you have?" and Sangyé Yeshé replied: "Look at the power of my mantra!" [ . . .] He pointed his index finger towards a rock, and a lightning bolt destroyed it into pieces.Then, the terrified king said: "I will not harm your followers!"It is clear that due to Sangyé Yeshé's kindness the mantrins with white robes and long hair were not harmed and, in general, [this was of] great benefit to the teachings of the Buddha. 61 another episode described in an alleged autobiography, we find a further instance in which he uses again wrathful mantras (what some people may call "black magic") 62 in order to kill a group of soldiers who were surrounding a fortress in which Tibetan monks were looking for shelter. 63The figure 58 (Dalton 2005, pp.115-16). 59One of the first problems we face when studying the life of Nupchen is the uncertainty surrounding his dates.For a detailed discussion of this issue as discussed in Tibetan as well as Western sources, see (Lopez 2014). 60For a history of the Mind Series literature, see (Lopez 2018). 61The Tibetan is "khyad par du rgyal po glang dar mas sangs rgyas kyi bstan pa bshig pa'i dus slob dpon 'dis rgyal po sdig can skrag par mdzad de/rgyal pos khyod la nus pa ci yod zer bas ngas sngags tsam bzlas pa'i nus pa 'di la gzigs shig ces [ . . .] da dung nus pa 'di la gzigs shing zer nas sdigs mdzub kyis thog phab ste pha ri'i brag la bsnun pas tshal bar song/der rgyal po 'jigs shing skrag nas khyed 'khor bcas la gnod pa mi byed do zer nas btang ste/sngags 'chang gos dkar lcang lo can rnams la gnod pa ma byung na khong gi drin du mngon te sangs rgyas kyi bstan pa spyi la'ang phan pa cher byung ngo."In Guru Tashi: 167. 62For more on the connection of Nupchen to black magic see (Cantwell 1997, pp.107-8, n. 5). 63See bka' shog rgya bo che 20a-21b.that emerges from these accounts is that of a scholar, but also of a powerful protector of the Buddhist teachings, who will do anything in his power to protect them. 64is works, or, we must say, the few that have survived, such as the Lamp for the Eye in Meditation or The Armor Against Darkness (mun pa'i go cha), 65 a commentary on the main text of the Anuyoga tradition, the S ūtra of the Gathered Intentions (dgongs pa 'dus pa'i mdo), reveal a remarkable intellectual, one who was well-versed in a wide variety of Buddhist contemplative traditions.His texts are filled with constant citations from hundreds of Buddhist scriptures, as well as the works of other Buddhist intellectuals from India and China, offering the most comprehensive window to the textual and intellectual world of Tibet during that period. 66ccording to various accounts, Nupchen wrote the Lamp in the early 10th century at the age of 61 67 to atone for all the negative actions he engaged in to protect the Dharma during the political and military upheaval that followed the collapse of the Empire.He also claims to have written the text in order to defend the emergent Great Perfection from claims of being a mere Tibetan copy of Chinese Chan, a line of attack that will be used by later Tibetan scholars such as Sakya Pan .d .ita. 68 In the larger historical context, it is important to remember that Buddhism was imported into Tibet as part of the great military, economic, and cultural transformation that the Land of Snows underwent during the Imperial era, between the seventh and the ninth century, (618-842 CE), which culminated with the establishment of this foreign tradition as the official state religion in the eighth century.During this period, Buddhism was imported and strictly regulated by the state through a few sponsored monastic institutions that controlled what texts were to be translated, and what kinds of rituals and practices were allowed.
The collapse of the Empire in the ninth century (842 CE), to which Nupchen had a front row seat, opened a brief but crucial period in the history of Buddhism in Tibet that has been traditionally labeled as the Tibetan Dark Age (a clear reference to the European Middle Ages) 69 and that has been described in traditional Tibetan historiography, and by most contemporary scholars in the field, as a period of intellectual decay, in which Buddhism almost disappeared from the Tibetan plateau.The problem with this description is that while the notion of a Dark Age may accurately evoke the collapse of the state-sponsored, monastic-centered form of Buddhism introduced by the empire, it also obscures the creative and dynamic changes that Buddhism underwent in Tibet during the same period.It is possible to argue that liberated from the restraints of Buddhist orthodoxy imposed by monastic institutions and the no-longer-existent state, Tibetans (and Nupchen was a key participant in this process) were able to transform what had been a foreign religion imposed by the state into a vehicle that was able to express genuine Tibetan religious ideas and concerns: Buddhism during the Dark Age became 64 For the complex intersection of Buddhism and violence see (Dalton 2011). 65Kapstein has some doubts about the attribution of this work to Nupchen.Although I agree with Dalton in considering this a work by Nupchen, more research on this topic may bring some light to the issue of its authorship.See (Dalton 2016;  Kapstein 2017). 66The Lamp also offers an invaluable widow to the textual world of Buddhism in Tibet before the collapse of the Empire.
The Lamp includes close to 750 quotes from almost 150 different sources, many of them no longer extant, or that have only survived in a fragmentary state.Nupchen quotes from s ūtras like the La ṅkāvatāra, Avatam .saka, and the Perfection of Wisdom S ūtras (Prajñāpāramitā); Nupchen also quotes more than 50 different Chan masters whose names only survive in the Lamp and in Dunhuang sources.There has been some scholarly work on some of these sources, particularly by Japanese scholars on the Chan chapter, and by others on the Great Perfection one, but there has not been a systematic attempt to trace all of the quotes of the text back, at least when possible, to their original sources.Donati (2006) and Esler (2018) have both translated this massive work. 67See bka' shog rgya bo che 20a-21b.See a discussion and a translation of this critical passage from the Bka' shog rgya bo che in (Lopez 2014, pp.79-80). 68On Sakya Pan .d .ita's attacks to the Great Perfection see his Clear Differentiation of the Three Vows (Tib.sDom gsum rab dbye), fol.25b. 69For a discussion on periodization in Tibetan history see (Cuevas 2006; Dalton 2011; Lopez 2014).On the use of the label 'Dark Age' to describe this period of Tibetan history see (Snellgrove 1987, p. 464; Kapstein 2000, pp.10-17; Denwood 2010,  p. 1; Manchester 1992, pp.3-5; Lopez 2014, pp.35-58).
Tibetan Buddhism.As we will see, doxographies like the Lamp played a very important role in this creative process.

The Lamp for the Eye in Meditation
In my brief overview of early Indian, Chinese, and Tibetan doxographies we have seen the tension at the heart of the genre: On the one hand, doxographies were conservative tools that tried to preserve the unity of the Buddhist tradition in the face of astonishing and sometimes contradictory views, while, on the other hand, they also offered a vast amount of room for creativity and innovation, as new ritual and doctrinal systems would be added to the older ones.
Before we focus on the Lamp, it is important to point out that Nupchen, just like some of the scholars we have explored earlier, developed a second doxographical system during his lifetime, which can be found in his other major surviving work, the Armor Against Darkness (mun pa'i go cha).This commentary on the S ūtra of the Gathered Intentions, a key Anuyoga text, clearly engages with the previous Indian classifications of these ritual and contemplative systems as elaborated by Buddhaguhya and Vilāsavajra in India, and the Garland in Tibet.Nupchen, though, argues in his Armor Against Darkness for a nine-vehicle classification that includes the Great Perfection (as Atiyoga) as a separate vehicle, something that none of the previous systems had proposed.Atiyoga 70 In fact, van Schaik has argued that Nupchen's doctrinal classification in the Armor Against Darkness may be the first instance of the nine-vehicle system found in Tibet. 71As Dalton points out, the Garland and Pelyang's text "divide their highest class of Mahāyoga into three 'techniques' (tshul) or 'stages' (rim) of development, perfection, and great perfection.In the writings of Nupchen, these three stages are enshrined as three separate classes". 72What is remarkable in Nupchen's classificatory system in the Armor is that it shows that he is part of a contemplative and textual movement in Tibet that is arguing for the unique nature of the Great Perfection, a movement that wants to separate the Great Perfection from the earlier tantric ritual and contemplative stages, transforming it into its own vehicle.In the case of the Armor, though, Nupchen seems to tackle the Great Perfection within the specific context of Indian Buddhism, while in the Lamp, as we will see, he is more concerned with establishing the Great Perfection in the much larger context of Buddhist contemplative practices across Asia.
The Lamp is divided into eight chapters (le'u), but has its core in chapters four through seven, where, in a hierarchical form, four different Buddhist vehicles are explained.The four vehicles and their degree of non-conceptuality are, as follows: the Gradual Approach (tsen men rim gyis 'jug pa), which describes the traditional Mahāyāna textual and scholastic tradition coming from India and represented mainly by the Yogācāra-Madhyamaka school; the Instantaneous Approach (ston mun cig car 70 (Dalton 2005, p. 140). 71See (Van Schaik 2004, p. 188).Germano has suggested that the nine vehicle doctrinal classification may be a Tibetan indigenous interpretation of the Buddhist path that is based on the Tibetan myth of the nine-runged ladder that connects heaven and earth (in David Germano's unpublished manuscript Mysticism and Rhetoric in the Great Perfection (rDzogs Chen}).
'jug pa), represented by Chinese Chan; Mahāyoga (rnal 'byor pa chen po), which reflects the new tantric developments that had become popular in India starting in the seventh-eighth century (particularly in the Guhyagarbha Tantra); and Atiyoga (rdzogs pa chen po), a new contemplative system that reflects an innovative Tibetan understanding of Buddhist doctrine and practice.
For Nupchen, the four Buddhist systems share something in common, an acceptance of the notion of "non-conceptuality" (Skt.avikalpa; Tib.mi rtog pa) as the defining characteristic of meditation.What allows him to differentiate between the various systems and hierarchically classify them is their methods (thabs) to reach that state of non-conceptuality, and their philosophical views (lta ba) or descriptions of that state.For Nupchen, each tradition has an increasingly subtler understanding of the notion of non-conceptuality, and therefore a more profound understanding of the nature of the meditative experience.

Defining and Comparing: Meditation and Non-Conceptualization
In a section dedicated to taxonomy and comparison in the chapter "When the Chips are Down" of his book, Relating Religion, Jonathan Z. Smith explains how his early interests in botany and fascination with taxonomy led him directly to a concern with comparison.For Smith, "[p]ut simply, taxonomy seemed a comparative enterprise which sought similarity across obvious individual variations and which asserted significant difference even in the face of apparent resemblances."73For Smith, comparison and classification are inseparable, since "comparison is never a matter of identity.Comparison requires the acceptance of difference as the grounds of its being interesting, and a methodical manipulation of that difference to achieve some cognitive end." 74mith's remarks will help us frame Nupchen's goal in Chapter 3 of the Lamp, in which he compares and classifies the four Buddhist contemplative systems discussed in the text (i.e., Gradual, Sudden, Mahāyoga, and Atiyoga).Nupchen's classification implicitly acknowledges the "acceptance of difference", since, in his view (i.e., in the Buddhist view), there is place for diversity of interpretations of the Buddhist path.At the same time, Nupchen also engages in a "methodological manipulation of that difference in order to achieve a cognitive end"; in this case, to establish a hierarchy and to impose the superiority of the Great Perfection over other Buddhist traditions.
In terms of the structure of the text, Chapter 3 is, without a doubt, the most important in Nupchen's Lamp. 75If Chapters 1 and 2 offer a general introduction to the practice of meditation and Chapters 4 through 7 are a detailed and individualized discussion of each of the four contemplative traditions, Chapter 3 reveals the methodological rationale behind the structure of the whole text.As Dalton and van Schaik have argued, Chapter 3 is "a microcosm of the text as a whole".764.1.The Avikalpapraveśa-Dhāran .ī and "Non-Conceptuality" in Tibet In Chapter 3 of the Lamp, Nupchen, in a very systematic way, describes the methodology behind the classification that will unfold in the following chapters.For Nupchen, all Buddhist systems share something in common, which will allow him to compare and, at the same time, rank them, as their different understandings (or should we say Nupchen's interpretation of their understanding) of that which they have in common is what gives him the ability to position them.What they have in common is their acceptance of the notion of "non-conceptuality" (Skt.avikalpa; Tib.mi rtog pa) as the defining characteristic of the Buddha's enlightened experience.What sets them apart is how they approach that state (their contemplative methods; Tib.thabs), and how each tradition understands it (their philosophical views; Tib.lta ba).
Although from its very early stages Buddhism had considered the Buddha's experience of enlightenment as being a non-conceptual, non-discursive, direct experience of the nature of reality, the emphasis on identifying non-conceptuality as the defining characteristic of Buddhahood seems to have taken hold within the Indian Siddha movement. 77The notion of 'non-conceptuality' played a very important role in many of the philosophical debates about the nature of enlightenment during the period of the early introduction of Buddhism in Tibet. 78It was at the heart of the so-called Samyé Debate. 79A text that played a key role in the introduction of this concept into Tibet was The Supreme Dhāran .ī of Entering into Non-Conceptuality (Skt.Āryāvikalpapraveśa-dhāran .ī; Tib.'phags pa rnam par mi rtog par 'jug pa zhes bya ba'i gzungs). 80Ueyama and Meinert argue the very important role that this scripture played in the dissemination of Buddhism from Central Asia into Tibet. 81Its importance can also be seen in the influence that the text had on the thought of some of the most important figures of Tibetan Buddhism during its early stages of introduction onto the Tibetan plateau, such as the Indian scholar, Kamalaśīla, who wrote a commentary on this scripture (phags pa rnam par mi rtog par 'jug pa zhes bya ba'i gzungs kyi rgya cher 'grel pa), 82 and the Chinese monk, Hashang Mahāyāna. 83The Dunhuang manuscript, Pelliot tibétain 116 (PT 116), a text that reflects Nupchen's understanding of the Chan tradition, also quotes the text.
In the Āryāvikalpapraveśanāmadhāran , ī scripture, 84 the historical Buddha is teaching the importance of entering into the non-conceptual (rnam par mi rtog pa) state to an assembly of bodhisattvas, and teaches a fourfold, gradual approach to achieve it in which the practitioner cuts off the marks (spong ba) of conceptual thinking (rtog pa) by not engaging in whatever thoughts appear in the mind (mngon du gyur pa dag yid la mi byes pas). 85The four different stages that will lead the practitioner to the state of non-conceptuality are: (1) perception (dmigs pa); (2) non-perception (mi dmigs pa); (3) non-perception of perception (dmigs pa mi dmigs pa); and (4) perception of non-perception (mi dmigs pa dmigs pa).Meinert describes these four stages as follows: In the first step of "perception" one cognizes all dharmas as the manifestation of "mere cognition" (rig pa tsam), that is, all dharmas are an expression of one's own mind [...] In the second step one cognizes "non-perception" of objects, to which the ordinary apprehension generally adheres [...] In the following step of "non-perception of perception" one trains oneself in the non-perception of the perception that "mere cognition" is non-existent.Since cognition is not possible without an object, cognition itself is also impossible [...] In the final step of "perception of non-perception" one perceives neither an apprehending subject not an apprehensible object.As subject and object are not of separate natures, non-duality may be realized. 86he relevance of the Āryāvikalpapraveśa-dhāran .ī and, in particular, the importance of the notion of non-conceptuality as the essence of the enlightenment experience can be seen in the arguments of the two factions of the Samyé Debate.Hashang Mahāyāna, the defender of the Chinese Chan approach, argued for a complete suppression of all mental activity in order to reach this non-conceptual state. 87amalaśīla, on the other hand, in his commentary to the Āryāvikalpapraveśa-dhāran .ī, as well as in various references in his Bhāvanākramas, argued against Mahāyāna's position since non-conceptuality in the Āryāvikalpapraveśa-dhāran .ī did not simply imply a complete suppression of all thoughts, or even a negative view of mental processes, as the Chinese master argued, but a gradual process in which conceptual thoughts play a role in gradually entering into a final non-conceptual state. 88his is not the place to discuss the nuances of this debate.Here, suffice it to say that the Āryāvikalpapraveśa-dhāran .ī and its notion of non-conceptuality had a very important role during the introduction of Buddhism in Tibet: It influenced some of the most important philosophical debates of the period and, as we will see, its fourfold path was probably the main influence for Nupchen's fourfold doxographical classification in the Lamp.

Non-Conceptuality and the Fourfold Structure of the Lamp for the Eye in Meditation
As we saw during our historical overview of the doxographical genre in Tibet in the previous section, Nupchen's classification in the Lamp is unusual in offering a fourfold scheme when the trend, particularly for the doxographical systems that originated in India, was towards a nine-vehicle system that was able to accommodate all the new ritual and contemplative tantric innovations of the eighth and ninth centuries. 89Although the origin of Nupchen's fourfold classification scheme is still not completely clear, Meinert makes a strong case when suggesting that Nupchen used the fourfold scheme of the Āryāvikalpapraveśa-dhāran .ī as a model for the one used in the Lamp. 90There is no direct acknowledgement by Nupchen of this influence, but all the evidence seems to point to the influence of the Āryāvikalpapraveśa-dhāran .ī on the structure of the Lamp: its importance in the introduction of the concept of non-conceptuality in Tibet; its influence in shaping some of the main arguments in the Samyé Debate; the fact that Nupchen is familiar with it, and quotes it in his text (including Section 3); and, finally, the similarities between the fourfold schemes in both texts. 9187 Ueyama argues that Hashang Mahāyāna's interpretation of non-conceptuality "took as the basic cause of birth-and-death the fact that thoughts (想xiang) grasp external objects, that is, that mental impulses arise.He held that awakening consists of cutting off the thoughts which are the cause of delusion and entering into no-examining/no-thought (Ch.不思不观)", (Ueyama et al. 1983, p. 36). 88Ueyama argues that Kamalaśīla "held that to deny all thoughts (manasikāra) is to deny even the correct examining or investigation (pratyaveks .ā=so sor rtog pa) which is necessary for arriving at non-discriminative wisdom, that is, awakening.His criticism was that subsuming the conditions for becoming a Buddha under insight alone neglects the good practices of upāya and contradicts the Mahāyāna" (Ueyama et al. 1983, p. 36).Higgins writes in his discussion of the related issue of mentation (manasikāra) vs. non-mentation (amanasikāra) in the Samyé Debate that Kamalaśīla's critique of Hashang Mahāyāna (although the Chinese master is not mentioned in the text, scholars agree that he is the recipient of Kamalaśīla's criticism) "is that amanasikāra does not imply the suppression or cessation of mental activity but rather its progressive refinement through the gradual elimination of subjective distortions [...] Thus, Kamalaśīla singles out mistaking amanasikāra for the absolute non-existence of manasikāra as the major misinterpretation of this concept, a point he further clarifies in his Nirvikalpapraveśadhāran .ī-t .īkā [...] Conceptual meditation, in other words, is a necessary condition for non-conceptual realization", in (Higgins 2009, pp.258-59).Gomez (2004) has defended a more nuanced understanding of Mahāyāna's positions that do include what some may call an acceptance of the Gradual Approach. 89This is, in fact, the doxographical approach used by Nupchen in his Armor Against Darkness (mun pa'i go cha). 90See (Meinert 2003).Dalton and van Schaik suggest a possible Chinese influence since scholars like the Korean scholar, Chegwan (d.971), taught "'four methods of conversion' (hua-i): gradual (chien), sudden (tun), secret (pi mi), and indeterminate (pu-ting)" (Dalton and van Schaik 2003, p. 156). 91Non-conceptuality also plays a very important role in a text quoted by Nupchen in the Lamp, Vimalamitra's Stages of the Path: "One who wishes to dispel obscuration and quickly achieve omniscience should concentrate on achieving quiescence and insight, thereby generating the altruistic mind.I, like a blind person, cannot demonstrate this, but relying on the words of Buddha and other scriptures, I will explain it.By right cultivation of quiescence and insight, non-conceptual intuition arises.
Through the arising of this [non-conceptuality], one can abandon all obscurations and attain omniscience that is the result of the longtime practice of quiescence and insight.Therefore, one should strive for quiescence and insight.If one dwells in the state of non-conceptualization, one can see the reality of all things.It is certain that the purification of all obscuration and the achievement of omniscience is dependent upon causes; therefore, one should contemplate non-conceptualization.
Nupchen structures the Lamp around the notion of "non-conceptuality" and identifies each of the different contemplative systems he discusses in the text with a different and increasingly subtler and refined understanding of that concept, with the Gradual tradition at the bottom, and Atiyoga, or the Great Perfection, at the top 92 : The Gradual tradition's goal is the non-conceptualization of appearances (snang ba mi rtog pa), the one of the Sudden Approach is the non-conceptualization of non-appearances (mi snang ba mi rtog pa), the Mahāyoga tradition attempts the non-dual non-conceptuality (gnyis su med pa'i mi rtog pa), and the Great Perfection, or Atiyoga, tradition's goal is the spontaneously present non-conceptuality (lhun gyis grub pa'i mi rtog pa).Let us explore how Nupchen explains each of these approaches to non-conceptuality, since they are key to the process of construction of doctrinal difference, as well as being central to Nupchen's project of building a new structural hierarchy that will allow him to claim the Great Perfection as the pinnacle of Buddhist contemplative practice.4.2.1.The Gradual Approach and "Non-Conceptualizing Appearances" (snang ba mi rtog pa) In order to describe the understanding of non-conceptuality of the Gradual Approach, Nupchen uses the Āryāvikalpapraveśa-dhāran .ī: As for the "tsen men 93 Gradual Approach, when [the practitioner] gradually abandons the four characteristics [which are] conceptualizing the nature, (conceptualizing the antidote, conceptualizing suchness, and conceptualizing the attainment), [the practitioner] enters into the non-conceptual [state].The Dhāran .ī of Non-Conceptuality (i.e., Avikalpapraveśa-dhāran .ī) says: "The people who want the wish-fulfilling jewels must use their minds to excavate under the extremely solid rock in order to see them.[So when people excavate] under these rocks they [find] the three precious stones: silver, gold, and emeralds.Under these four very firm rocks, all of them excavate [some more] and obtain the wish-fulfilling jewel, which emerges as the perfection of everything for the sake of self and others.[Similarly, those who want to reach] the non-conceptual [state] should free themselves from the four conceptualizations: conceptualizing the nature, conceptualizing the antidote, conceptualizing suchness, and conceptualizing the attainment.If [the practitioner] meditates about suchness without creating existents (yod pa) [the practitioner] will accomplish the three characteristics".They gradually meditate on the three gates of emptiness itself (stong pa nyid), without signs (mtshan ma med pa), and without aspirations (smon pa med pa), and they also meditate [using the techniques] of Śamatha and Vipaśyanā. 94malaśīla, one of the main proponents of the Gradual Approach, argued the importance of mental processes (Skt.manasikāra; Tib.yid la byed pa) and conceptual thinking (Skt.vikalpa; Tib.rtog pa) As one who reaches the top of a lofty mountain can see almost all the surrounding places, similarly, one who dwells in the state of non-conceptualization can see all things without obstacle or impediment.Therefore, one should contemplate on quiescence and insight."In (Jamspal 2000, p. 1). 92In this same chapter, Nupchen sketches an alternative threefold non-conceptual classification that does not develop in the text.For a brief discussion of this alternative system, see (Dalton and van Schaik 2003, p. 154). 93It is interesting to point out that Nupchen uses the Tibetan (rim gyis), as well as the Chinese term (tsen men), to designate the Gradual Approach.For the Sudden school he also uses both languages (Tib.cig car; Ch. ton mun). 94Lamp 55.6-57.1:de la tsan med rim gyis 'jug pa ni/rang bzhin (56.1) la rtogs pa'i mtshan ma la sogs ba <dang gnyin po rtog pa dang/de kho na nyid la shrtogs pa dang/thob pa la rtog pa dang/> bzhi rim gyis spangs nas/mi rtog pa la 'jug pa ste/de yang mi rtog pa'i gzungs las/yid bzhin gyi nor bu rin po (56.2) che 'dod pa'i mis brag shin tu rsa zhing 'greg pa'i 'og na yid kyis rkos shig//rin po che 'di lta ste//(56.3)brag de'i 'og na/dngul rin po che dang/gser dang/rdo'i snying po'o//de dag la rin po che'i brag gsum//brag shin tu sra ba dang bzhi'i 'og nas/de yo brkos (56.4) pa dang/yid bzhin nor bu bdag dang gzhan don thams cad rdzogs par 'byung ba 'thob pa de bzhin tu/rnam par mi rtog pa 'dod pas/rang bzhin la rtog pa'i mthan (56.5) ma dang/gnyen po la rtog pa'i mtshan ma dang/de kho na nyid la rtog pa'i mtshon ma dang/thob pa la rtog pa'i mtshan ma de bzhi dang bral bar byas nas/mtshan (56.6) ma gsum po la de kho na yis yod la ma byas par bsgom na 'grub bo//zhes gsungs pas/rim par bsgom pa dang/gzhan stong pa nyid dang/mtshan (57.1) ma med pa dang/smon pa med pa'i sgo gsum rim par bsgom pa dang/zhi gnas lhag mthong la sogs pa bsgom pa'o//.
as the starting point of the meditation path, as well the central difference with the Sudden Approach.In this tradition, then, we see a gradual abandonment of the process of conceptualization. 954.2.2.The Sudden Approach-Non-Conceptualizing Non-Appearances (mi snang ba mi rtog pa) According to Nupchen, the practitioners of the Sudden Approach do not need to begin meditative practice by engaging in conceptual thinking, since they are able to instantaneously enter into a non-conceptual state: From the beginning, without alternating [methods], they train in the unborn ultimate [...]  The entrance gate to the Sudden Approach has been explained by many teachers [such as] the Great Master Bodhidharma, who said "Turn in the correct direction and abandon conceptuality".When you abide in wall-gazing, 96 there is no self and other.Common people and noble ones are one and the same.When abiding steadily, without moving, from then on, [you] will not go after words and teachings.[You] will abide in the state of the true meaning, where there is non-conceptualization, and no activity. 97e superiority of the Sudden over the Gradual is, according to Nupchen, based on the acknowledgment of the superiority of non-conceptuality over conceptuality as the starting point of contemplation.The problem, though, as Guenther pointed out, is that the Sudden Approach still "emphasizes the non-thematic [i.e., non-conceptuality] over and against the thematic [i.e., conceptuality] without realizing that experience is made up of both, the thematic and the non-thematic". 98Also, Nupchen critizices the Chinese Sudden tradition since, in their efforts to distance themselves from any type of conceptuality (and therefore, separate themselves from the Gradual Approach), they seem to have reified the idea of non-conceptuality almost without noticing it. 99As Guenther argues, "To speak 95 Nupchen describes it as a process in which the three types of conceptualization are gradually abandoned.These three types of conceptual discrimination (trivikalpa) are described in the Princeton Dictionary of Buddhism as: "(1) Intrinsic discrimination (svabhāvavikalpa), which refers to the initial advertence of thought (vitarka) and the subsequent sustained attention (vicāra) to a perceived object of the six sensory consciousnesses (vijñāna), that is, the discrimination of present objects, as when visual consciousness perceives a visual object.
(2) Conceptualizing discrimination (abhinir ūpan .āvikalpa), which refers to discursive thought on ideas that arise in the sixth mental consciousness when it adverts toward a mental object that is associated with any of the three time periods of past, present, or future.(3) Discrimination involving reflection on past events (anusmaran .avikalpa), which refers to discriminative thought involving the memory of past objects." 96The term "wall-gazing" (lham mer gnas pa) refers to the Chinese term biguan (( The superiority of the Sudden over the Gradual is, according to Nupchen, based on the acknowledgment of the superiority of non-conceptuality over conceptuality as the starting point of contemplation.The problem, though, as Guenther pointed out, is that the Sudden Approach still "emphasizes the non-thematic [i.e., non-conceptuality] over and against the thematic [i.e., conceptuality] without realizing that experience is made up of both, the thematic and the nonthematic". 98Also, Nupchen critizices the Chinese Sudden tradition since, in their efforts to distance themselves from any type of conceptuality (and therefore, separate themselves from the Gradual Approach), they seem to have reified the idea of non-conceptuality almost without noticing it. 99As Guenther argues, "To speak of something as never having come into existence is to have a concept of something, to which we may make attributions.This is precisely to be back in the thematic aspect of experience, and this the follower of the 'instantaneous' [i.e., Sudden] Approach will not admit." 1004.2.3.Mahāyoga-Non-Dual Non Conceptuality (gnyis su med pa'i mi rtog pa)

96
The term "wall-gazing" (lham mer gnas pa) refers to the Chinese term biguan ( 壁觀 ), a key concept for the early Chan tradition as found in Bodhidharma's The Treatise on the Two Entrances and Four Practices (Ch.二入 四行论).As McRae has pointed out, though, "this concept has bedeviled the Chan tradition ever since its introduction [in that text]", since "ultimately, no one really knows what the term means.It only occurs in one other more-or-less contemporaneous source, a list of meditation practices recommended for beginners, where it occurs without explanation.(McRae, p. 30).As McRae points out, "eventually, the term came to be interpreted in the Chan tradition as referring to the act of sitting in meditation facing a wall, but as indicated in the discussion of Bodhidharma's hagiographical evolution above, it took some time for this meaning to take hold."(McRae 2003, p. 30).Meinert has argued that the Tibetan "lham mer gnas pa" should be translated as "abiding in luminosity", and offers a compelling case as to why this should be.In this case, though, I prefer to use "wall-gazing" since the term is used in the context of the Sudden teachings of Bodhidharma, and wall-gazing is the most common term to translate the Chinese biguan, which is what I think the Tibetan term is trying to translate (as Meinert herself acknowledges in her article).See (Meinert 2007, p. 255).I want to thank an anonymous reader for pointing out the need to clarify my use of 'wall-gazing' to translate the Tibetan 'lham mer gnas pai.' 97 Lamp 57.1-58.2:"ston mun ni cig car (57.2) 'jug pa ste/dang po nas re mos pa med par don dam pa ma skyes pa nyid la cig car slob te/shes rab kyi pha rol du phyin pa'i mdo las//dang po sems bskyed pa (57.3) nas rnam pa thams cad mkhyen pa nyid la yid bya'o//[...] mkhan <gzhung bstan/> po chen (57.6) po bo dhe dar mo ta ras bshad pa las/yang dag pa la phyogs shing rtogs pa spangs te/lham mer gnas na/bdag kyang med gzhan yang med/ma (58.1) rabs dang 'phags pa mnyam zhing gcig ste/mi 'gyur bar brtan par gnas na/de phan chad yi ge dang bstan pa'i rjes su mi 'brang ngo//'di ni yang dag (58.2) pa'i don gyi rnal du phab pa rnam par rtog pa med pa/zheng zhing bya ba med pa ste/de ni don la mi 'jug pa'o//".98 (Guenther 1983, p. 356).

99
In the chapter dedicated to the Mahāyoga tradition, Nupchen criticizes the Sudden tradition for thinking that their view is "free from activity", but theirs is only a partial understanding of that concept, since it is not grounded in the superior non-dual view of Mahāyoga: "tshe de nyid la dbang bsgyur chen por 'gyur ba dang 'dra ste/ston mun dang (226.6)nga lta bas bya ba dang bral ba skad zer yang/theg pa chen po gnyis su med pa'i de bzhin nyid/mtha' yang ma mthong ste/don dam pa 'ba' zhig la mi stsol du zad de/(227.1)'jag pa non bgrong pa mthar phyin nas de bzhin gshegs pas dbang bskur nas bzod rdo rje lta bu gnyis med shar nas sngags mthong ba dang/(227.2) de bzhin du da lta blo la brtan par cud zhig 'dug du ma btub kyang/gnyis su ), a key concept for the early Chan tradition as found in Bodhidharma's The Treatise on the Two Entrances and Four Practices (Ch.二入四行论).As McRae has pointed out, though, "this concept has bedeviled the Chan tradition ever since its introduction [in that text]", since "ultimately, no one really knows what the term means.It only occurs in one other more-or-less contemporaneous source, a list of meditation practices recommended for beginners, where it occurs without explanation.(McRae, p. 30).As McRae points out, "eventually, the term came to be interpreted in the Chan tradition as referring to the act of sitting in meditation facing a wall, but as indicated in the discussion of Bodhidharma's hagiographical evolution above, it took some time for this meaning to take hold."(McRae 2003, p. 30).Meinert has argued that the Tibetan "lham mer gnas pa" should be translated as "abiding in luminosity", and offers a compelling case as to why this should be.In this case, though, I prefer to use "wall-gazing" since the term is used in the context of the Sudden teachings of Bodhidharma, and wall-gazing is the most common term to translate the Chinese biguan, which is what I think the Tibetan term is trying to translate (as Meinert herself acknowledges in her article).See (Meinert 2007, p. 255).I want to thank an anonymous reader for pointing out the need to clarify my use of 'wall-gazing' to translate the Tibetan 'lham mer gnas pai.' 97 Lamp 57.1-58.2:"ston mun ni cig car (57.2) 'jug pa ste/dang po nas re mos pa med par don dam pa ma skyes pa nyid la cig car slob te/shes rab kyi pha rol du phyin pa'i mdo las//dang po sems bskyed pa (57.3) nas rnam pa thams cad mkhyen pa nyid la yid bya'o//[...] mkhan <gzhung bstan/> po chen (57.6) po bo dhe dar mo ta ras bshad pa las/yang dag pa la phyogs shing rtogs pa spangs te/lham mer gnas na/bdag kyang med gzhan yang med/ma (58.1) rabs dang 'phags pa mnyam zhing gcig ste/mi 'gyur bar brtan par gnas na/de phan chad yi ge dang bstan pa'i rjes su mi 'brang ngo//'di ni yang dag (58.2) pa'i don gyi rnal du phab pa rnam par rtog pa med pa/zheng zhing bya ba med pa ste/de ni don la mi 'jug pa'o//". 98(Guenther 1983, p. 356). 99In the chapter dedicated to the Mahāyoga tradition, Nupchen criticizes the Sudden tradition for thinking that their view is "free from activity", but theirs is only a partial understanding of that concept, since it is not grounded in the superior non-dual view of Mahāyoga: "tshe de nyid la dbang bsgyur chen por 'gyur ba dang 'dra ste/ston mun dang (226.6)nga lta bas bya ba dang bral ba skad zer yang/theg pa chen po gnyis su med pa'i de bzhin nyid/mtha' yang ma mthong ste/don dam pa 'ba' zhig la mi stsol du zad de/(227.1)'jag pa non bgrong pa mthar phyin nas de bzhin gshegs pas dbang bskur nas bzod rdo rje lta bu gnyis med shar nas sngags mthong ba dang/(227.2) de bzhin du da lta blo la brtan par cud zhig 'dug du ma btub kyang/gnyis su med pa rdo rje lta bu'i ye shes la slob pas/tshe 'di nyid la mthar thug pa'i phyir rol kyang (227.3) of something as never having come into existence is to have a concept of something, to which we may make attributions.This is precisely to be back in the thematic aspect of experience, and this the follower of the 'instantaneous' [i.e., Sudden] Approach will not admit." 1004.2.3.Mahāyoga-Non-Dual Non Conceptuality (gnyis su med pa'i mi rtog pa) According to Nupchen, Mahāyoga tries to break free from this tension between conceptuality and non-conceptuality by embracing the non-dual nature of reality in which those distinctions disappear: As for inner non-conceptual Thusness, the Mahāyoga tradition [proposes] that all phenomena have the radiance of intrinsic awareness [and therefore], the two truths do not exist, the agent does not act [upon phenomena], and it becomes luminous [naturally].[All phenomena] are the non-dual sphere of wisdom.The Guhyasamāja Tantra says, "The object of meditation is not an object, the activity of meditation is not meditation, therefore, because objects do not exist, there is no object of reference in meditation." 101However, as Guenther argues in his interpretation of Nupchen's presentation of Mahāyoga's approach to non-conceptuality: "It is precisely this attempt to preserve the unitary [i.e., non-dual] character of experience as an interiority in which there is neither a 'within' nor a 'without' that prevents the experience from coming to grips with that spontaneity [of] pure experience". 102For Guenther, both the Gradual and Sudden traditions have an ontological approach to contemplation in which the practitioner objectifies reality during the practice of meditation.The Mahāyoga tradition is aware of that pitfall and focuses on non-duality, but in its effort to preserve that non-dual nature of the meditative experience, it does not leave space for the acknowledgement of the spontaneous and dynamic nature of reality.Although I find Guenther's interpretation compelling, he is also in many ways as much of an apologist of Nupchen's work, as he is an interpreter.While he may be right about Nupchen's definition of non-conceptuality within the Mahāyoga tradition, he also seems to ignore that Nupchen is responsible for the interpretative framework in which each of the traditions is defined and finally judged, and which will ultimately allow him to claim the superiority of the Great Perfection tradition.4.2.4.Atiyoya-Spontaneously Present Non-Conceptuality (lhun gyis grub pa'i mi rtog pa) Finally, Nupchen presents what he considers to be the highest vehicle, Atiyoga, or the Great Perfection, would offer the supreme understanding of non-conceptuality, "non-conceptuality which is spontaneously perfect" (lhun gyis grub pa'i mi rtog pa): As for the spontaneously perfected Thusness of Atiyoga, the Supreme Yoga, all appearing and existing phenomena are primordially self-radiant since they remain perfect in the perfectly pure sphere of self-emergent wisdom.As for the cause and the result, which are spontaneously perfected without any need to search [for them], this is the Great Self, where there is no movement, no particles, no names.So what is to meditate in the clear radiance where there is an intrinsic awareness that is not established, does not waver, is not tainted, mi 'byi char yang mi phod de/de ni lta ba'i don gzhi nas khyad par du 'phags pas rgyu'i theg pas thob pa'i byang chub sems dpa' nyid/'dir dang po sems bskyed pas (227.4)zil gyis gnon par gsungs yang/sems dang po skyes pa nas gnyis su med pa'i don la sems bskyed pas/stobs de las gyur pa ste/zhib tu phyis 'chad (227.5)do/". 100(Guenther 1983, p. 357). 101Lamp 59.3-59.6:"nang pa'i mi rtog pa de bzhin nyid ni/mah'a yo ga'i gzhung gis/chos thams cad ni rang rig par gsal ba nyid bden pa gnyis med pa/byed pa pos ma byas pa dang/yong gis 'od gsal ba/dbying ye shes gnyis su med pa ste/gsang ba 'dus pa las/dngos po bsgom pa med pa'i dngos//bsgom par bya ba bsgom pa med//de ltar dngos po dngos med pas//bsgom pa dmigs su med pa'o//zhes gsungs pas". 102(Guenther 1983, p. 360).
does not settle?What is there to recollect?There is nothing [...] Non-conceptuality is itself a [mere] concept. 103re, the non-dual approach of the Mahāyoga tradition is superseded by an understanding of reality in which everything is spontaneously perfect, and therefore there is no need for any effort or practice to reach the stage of non-conceptuality.What allows Nupchen to make this claim is not only the unique understanding of non-conceptuality espoused by the Great Perfection, but also its non-dual, spontaneous nature, concepts that will be at the center of this new Tibetan tradition.4.2.5.Meditation and 'the Rungs of a Ladder' The structural similarity of Nupchen's system with the fourfold presentation of non-conceptuality in the Āryāvikalpapraveśa-dhāran .ī can be seen in Figure 1.
Religions 2018, 9, x FOR PEER REVIEW 23 of 28 the unique understanding of non-conceptuality espoused by the Great Perfection, but also its nondual, spontaneous nature, concepts that will be at the center of this new Tibetan tradition.
4.2.5.Meditation and 'the Rungs of a Ladder' The structural similarity of Nupchen's system with the fourfold presentation of nonconceptuality in the Āryāvikalpapraveśa-dhāraṇī can be seen in Figure 1.Nupchen seems to indicate that the four traditions articulate a gradual path in which the practitioner could follow in stages the contemplative techniques of each of the traditions in order to achieve the supreme form of non-conceptuality, that of the Atiyoga tradition.He insinuates as much in the text when he claims: The differences between these [contemplative systems] are like the rungs of a ladder.Just as a ladder has higher and lower rungs, so are the differences between these four nonconceptualizations [of the Gradual, Sudden, Mahāyoga, and Atiyoga traditions]. 104netheless, there is also an important tension at the core of the system.The overall gradual message implied with the image of the ladder resonates with the approach presented in the Āryāvikalpapraveśa-dhāraṇī.The problem, though, emerges when Nupchen attaches each of the stages of non-conceptuality to a different Buddhist contemplative tradition.While all Buddhist traditions as described by Nupchen may share this notion of non-conceptuality, none of those traditions, besides the Great Perfection, to which Nupchen belongs, would agree with Nupchen's portrayal of their understanding of it, and they would, most certainly, not situate themselves in positions of inferiority vis-à-vis other traditions.They are all, from their respective points of view, perfect and complete contemplative systems, and it is only in the context of Nupchen's project that other systems are demoted.This tension is also accentuated by the fact that throughout the text Nupchen continuously undermines the idea of a gradual approach to his overall system when he keeps dismissing all of the 104 Lamp 60.6-61.1:"de dag gi khyad par skas gyi gdang bu bzhin te/dper na skas gdang la mtho dman yod par dang 'dra ste/mi rtog pa 'di bzhi yang khyad par yod <ston tsen mah'a a ti/>//."The same notion is repeated in the final colophon of the text in Lamp 500.4:"tsen ston mi dmigs nang pa'i mi rtog de bzhin nyid//lhun rdzogs mi rtog khyad par skas bzhin chud par bya'o//" Bold and italics are mine.Nupchen seems to indicate that the four traditions articulate a gradual path in which the practitioner could follow in stages the contemplative techniques of each of the traditions in order to achieve the supreme form of non-conceptuality, that of the Atiyoga tradition.He insinuates as much in the text when he claims:

Meditational
The differences between these [contemplative systems] are like the rungs of a ladder.Just as a ladder has higher and lower rungs, so are the differences between these four non-conceptualizations [of the Gradual, Sudden, Mahāyoga, and Atiyoga traditions]. 104netheless, there is also an important tension at the core of the system.The overall gradual message implied with the image of the ladder resonates with the approach presented in the 103 Lamp 60.3-60.6:"a ti yo ga lhag pa'i rnal 'byor gyi lhun rdzogs de bzhin nyid ni//snang srid gyi chos so cog rang byung gi ye shes rnam par dag pa'i (60.3) klong du sel med par ye nas rang gsal ba la//rgyu dang 'bras bu ril ma btsal bar lhun gyis rdzogs pa ni/bdag nyid chen po pas/de la g-yo rdul ming (60.4) yang med pas/rang rig pa ma bzhag ma g-yos ma bslad ma zhugs par lhan ne lhang nge ye gsal bar ci zhig bsgom/ci zhig dran par byar yod de med/(60.5)med pa'i don de nyid kho na yod/de dang du len pa su zhig ste/ye mi rtog pa chen po la/snang ba bkag pa yang med la/de la rtogs pa med de/mi rtog pa (60.6) nyid kyang bla dwags so//". 104Lamp 60.6-61.1:"de dag gi khyad par skas gyi gdang bu bzhin te/dper na skas gdang la mtho dman yod par dang 'dra ste/mi rtog pa 'di bzhi yang khyad par yod <ston tsen mah'a a ti/>//."The same notion is repeated in the final colophon of the text in Lamp 500.4:"tsen ston mi dmigs nang pa'i mi rtog de bzhin nyid//lhun rdzogs mi rtog khyad par skas bzhin chud par bya'o//" Bold and italics are mine.
Āryāvikalpapraveśa-dhāran .ī.The problem, though, emerges when Nupchen attaches each of the stages of non-conceptuality to a different Buddhist contemplative tradition.While all Buddhist traditions as described by Nupchen may share this notion of non-conceptuality, none of those traditions, besides the Great Perfection, to which Nupchen belongs, would agree with Nupchen's portrayal of their understanding of it, and they would, most certainly, not situate themselves in positions of inferiority vis-à-vis other traditions.They are all, from their respective points of view, perfect and complete contemplative systems, and it is only in the context of Nupchen's project that other systems are demoted.This tension is also accentuated by the fact that throughout the text Nupchen continuously undermines the idea of a gradual approach to his overall system when he keeps dismissing all of the other traditions with the exception of the Great Perfection, criticizing their approaches in each of the chapters.

Conclusions: Comparison, Classification, and the Construction of Tibetan Buddhism
Here, I would like to bring back Jonathan Smith's ideas in order to address the inherent tension in Nupchen's system (and of all doxographies for that matter).In the first place, Smith can help us understand the constructed nature of Nupchen's classification.For Smith, "Comparisons are not a given, they are the result of thought . . .comparison does not necessary tell us how things 'are' . . .like models and metaphors, comparisons tells us how things might be redescribed". 105Under this different interpretation of comparison and classification, we can see Nupchen's project not as a description of Buddhism during his time as existing "out there", but as a creative project in which Nupchen is reenvisioning what Buddhism could have been in Tibet. 106Secondly, for Smith, comparisons: . . .bring differences together within the space of the scholar's mind for the scholar's own intellectual reasons [...] Comparison provides the means by which we 're-vision' phenomena as our data in order to solve our theoretical problems.Comparison, as seen from such view, is an active, at times even playful, enterprise of deconstruction and reconstruction which gives the scholar a shifting set of characteristics with which to negotiate the relations between his or her theoretical interests and data stipulated as exemplary. 107e notion of non-conceptuality allows Nupchen to bring together these four different traditions "within the space of the scholar's mind for the scholar's own intellectual reasons".These intellectual reasons can be seen as twofold.On the one hand, Nupchen wants to make sense of the variety of Buddhist contemplative systems that had made their way into Tibet up until his own time.On the other hand, he wants to establish the superiority of his own contemplative tradition, the Great Perfection, above all other Buddhist traditions.Definition, comparison and classification, using Smith's terms, offered Nupchen the tools to actively and creatively deconstruct the Buddhist tradition only to put it back together in a new, original way that responds not only to the historical developments of the Buddhist tradition over the centuries, but to the present historical context in which the Buddhist tradition was developing in Tibet.
As we have seen above, Nupchen lived during the so-called Dark Age period, and it might well be that it was under the cover of this darkness that Tibetans like him were able to liberate themselves from the restraints of Buddhist orthodoxy imposed by monastic institutions and the no-longer-existent state.Thus, it was in this post-Imperial period that scholars like Nupchen were not simply interested in importing and translating Buddhism wholesale, but have become fully engaged in the wider conversation that is taking place across Asia about Buddhist doctrine and practice.In the Lamp, we do not only see a presentation of established Buddhist views across the continent, but a creative Tibetan attempt to transform what had been a foreign religion into a vehicle that expressed genuine Tibetan religious ideas and concerns.
As Smith also argued, "[T]he world is not a given, it is not simply 'there.'We constitute it by acts of interpretation.We constitute it by activities of speech and memory and judgment.It is by an act of human will, through projects of language and history, through words and memory, that we fabricate the world and ourselves".108I believe that by looking at Nupchen's project through the lens of Smith, we can reenvision Nupchen's narrative, looking at it not as an inevitable, inner-driven, theological Buddhist interpretation of the tradition, but by understanding it in its very specific social and historical context, and especially as an attempt to construct a worldview that translated the needs of a new, evolving culture.Nupchen and doxographies like the Lamp were central to the creative transformation of Buddhism into Tibetan Buddhism.

Figure 1 .
Figure 1.The Lamp for the Eye in Meditation and the Āryāvikalpapraveśa-dhāraṇī.

Figure 1 .
Figure 1.The Lamp for the Eye in Meditation and the Āryāvikalpapraveśa-dhāran .ī.
When you abide in wall-gazing, 96 there is no self and other.Common people and noble ones are one and the same.When abiding steadily, without moving, from then on, [you] will not go after words and teachings.[You]willabide in the state of the true meaning, where there is non-conceptualization, and no activity.97