Below is a review that was published in Confluence: Online Journal of World Philosophies, Vol. 2, 2015. The journal "is a bi-annual, peer-reviewed, international journal dedicated to comparative thought. It seeks to explore common spaces and differences between philosophical traditions in a global context." to subscribe to the journal, visit its website at http://www.verlag-alber.de/e-journals/confluence/.
During the lifetime of the Buddha and in subsequent centuries, the philosophical traditions of India commonly accepted the existence of an eternal, substantive self (ātman). Among Buddhism’s most novel and noteworthy tenets was the rejection of this view and the acceptance of the doctrine of selflessness or the non-existence of the self (anātman). The non-existence of the self was, however, controversial even among Buddhists, due in part to the Buddha’s conflicting comments on the question and to the Buddha’s use of personal pronouns. This led some to believe that he endorsed the existence of the self. As a consequence, various schools interpreted the doctrine in various ways. Several schools, particularly the Vātsīputrīyas and the Sammitīyas, maintained that some sort of “inexpressible person” (pudgala) must exist in order to make sense of personal continuity and rebirth and that this inexpressible person did not contradict the non-existence of the self. These schools became collectively known as Pudgalavādins. Other schools, particularly the Sarvāstivādins and the Sautrāntikas, maintained that the self was a conceptual fiction, constructed out of more fundamental elements called “dharmas.” Still another school, the Madhyamakas, considered the self, along with all objects, to be without independent existence. The Madhyamaka view was developed first by the second century philosopher Nāgārjuna and subsequently by other philosophers, including the seventh century philosopher Candrakīrti.
Works written by Buddhist philosophers on the self are well-worth reading for any philosopher outside of the Buddhist tradition as they offer theses that are at times analogous to ones found in the European tradition as well as theses that have no clear analogy. Among the most important works is the Abhidharmakośabhāṣya by Vasubandhu, particularly its ninth chapter, Refutation of the Theory of the Self (Ātmavādapratiṣedha or Pudgalapratiṣedhaprakaraņa). This work presents several important theories of the self. It outlines the view held by the Sarvāstivādins, the Pudgalavādins, and the Sautrāntikas. For the Madhyamaka tradition, one would do well to read Verses on the Fundamentals of the Middle Way (Mūlamadhyamakakārikā) by Nāgāruna and several works by Candrakīrti: Clear Words (Prasannapadā), Introduction to the Middle Way (Madhyamakāvatāra), and his Autocommentary on the Introduction to the Middle Way (Madhyamakāvatārabhāṣya). None of these are easy reading for anyone not steeped in the concepts and terminology of the Buddhist tradition. Happily, James Duerlinger has provided us with two mostly clear and insightful guides to much of this literature.
The first work is his 2003 book, Indian Buddhist Theories of Persons: Vasubandhu’s “Refutation of the Theory of the Self” which provides us with a translation of Vasubandhu’s Refutation of the Theory of the Self.1 The second is his 2013 book, The Refutation of the Self in Indian Buddhism: Candrakīrti on the Selflessness of Persons which provides us with a translation of verses 120-165 of Candrakīrti’s Autocommentary on the Introduction to the Middle Way. A full review of Duerlinger’s 2003 work is beyond the scope of this review, but readers would be well served to read at least the introduction to the 2003 work. This will give the reader a background that will make reading Duerlinger’s 2013 work more meaningful.
The Refutation of the Self in Indian Buddhism is composed of three parts. The first part is a general introduction to the root text and an overview of the issues that it addresses (pp. 1-54). The second is the translation of the root text (pp. 55-89). The third is Duerlinger’s own verse-by-verse commentary on the root text (pp. 90-194). In the introduction, Duerlinger describes and explains the views that Candrakīrti attributes to a several Buddhist schools: the Sāṃmitīyas, the Āryasāṃmitīyas, the Sarvāstivādins, and the Sautrāntikas, as well as the non-Buddhist Tīrthikas. Duerlinger also provides a relatively clear expression of Candrakīrti’s criticisms of these views as found in Candrakīrti’s Clear Words, Introduction to the Middle Way, and his Autocommentary on the Introduction to the Middle Way. The introduction is composed of four sections. The first section distinguishes Duerlinger’s translation and commentary from the existing English translations and commentaries. The second places Candrakīrti’s Autocommentary in the context of the Mahāyāna and Madhyamaka traditions and explains the ten stages of the Bodhisattva path of meditation and its fruit as Candrakīrti understands it from the Sūtra on the Ten Stages (Daśabhūmika Sūtra). The third presents valuable explanations of several critical terms used by Candrakīrti, and the fourth section relates Candrakīrti’s theory of persons to other Indian Buddhist theories.
The second part of the work, the translation of the root text, is informed by what is perhaps the most important contribution that Duerlinger makes toward understanding Candrakīrti’s arguments: the distinction between a self “with person-properties” and a self “without person-properties.” By selves “with person-properties,” Duerlinger means beings that possess minds and bodies, perceive, think, feel, act, etc. When English speakers use the term “self” (and personal pronouns), we commonly refer to beings with such properties. This is most evident in our use of the reflexive pronouns “myself,” “yourself,” “himself,” “herself,” “ourselves,” “yourselves,” and “themselves.” In each case, we refer to beings that have person-properties. Even in the case of “itself,” we commonly use the term to refer to beings with person-properties, e.g., “the mouse trapped itself in the box.” The neuter pronoun merely elides our ignorance of the mouse’s sex. There are, however, some instances when we use “itself” (and even “themselves”) to refer to objects without person-properties, e.g., “the building collapsed on itself” or “the bean stalks entwined themselves around the poles.” In these instances, we appear to suggest a degree of agency (a feature of personhood) that on more careful analysis we would reject. So while it is not always true, on the whole our use of “self” refers to beings with person-properties.
The use of the term “atman” to refer to persons is less consistent in Buddhist texts. The world “ātman” is normally translated as “self,” but it ambiguously refers to beings with person-properties and objects without person-properties. According to Duerlinger, by carefully attending to the ambiguities in the Buddhist texts and marking them with his person-property terminology, we can better understand the arguments made by Candrakīrti. Duerlinger writes, “The distinction [between selves with and without person-properties] is not to my knowledge explicitly drawn by Candrakīrti and his Madhyamaka (Middle Way) followers,” but he goes on to write, “The distinction is needed to explain why he [Candrakīrti] represents his fellow Buddhists as asserting the thesis that a self exists by itself when they deny that a self exists by itself” (p. 4). Perhaps it is because Duerlinger does not find explicit evidence for his person-property terminology that he does not us the terminology in his translation of the root text, but it helpfully appears in both his introduction to the root text and in his commentary on the root text.
The third part of the work is Duerlinger’s verse-by-verse commentary on Candrakīrti’s Autocommentary. It is based on seven Tibetan commentaries written in the Madhyamaka tradition, six of which are from English translations. Among the value-added features of Duerlinger’s commentary are quotations from Candrakīrti’s Clear Words, his commentary on Nāgārjuna’s Verses on the Fundamentals of the Middle Way. These quotations provide additional helpful perspective on Candrakīrti’s views.
To delineate the various Buddhist views of the self as Duerlinger believes Candrakīrti understands them, we should start by describing a view of the self held by the non-Buddhist Tīrthika school. This is the most robust view of the self considered by Candrakīrti. We can compare it to something like (but only something like) a Cartesian substantive self. It is an eternally existing mind that is temporarily associated with a particular body. This is in contrast to the Buddhist view that sees the self as identical to or at least dependent upon the body. Perhaps the simplest version of this contrasting view is held by the Sarvāstivādins and the Sautrāntikas. They maintained that the self is identical to the aggregates (skandhas), i.e., collections of elementary “dharmas” which we might recognize as (i) physical atoms, (ii) sensations, (iii) perceptions, (iv) volitional actions and external forces that condition our circumstances, and (v) consciousness. The classic explanation of this view appears in the Questions of Milinda (Milindapañha), written in the first century. In this text, Nāgasena explains to King Malinda that the self is like a chariot, composed of parts, and while one might say that each part exists, the chariot only exists dependently upon the parts; hence, its ontological status is different than the ontological status of the parts. The chariot does not exist in the strictest sense. The word “chariot” is only a convenient way to refer to the collection of parts that alone exist. Similarly, the word “I” is merely a convenient way to refer to the collection of parts or “aggregates” which make up the self.
David Hume comes closest to holding this particular view of the self. For Hume, personal identity is a bundle of overlapping impressions and ideas. Hume would not agree with the Buddhist enumeration of the aggregates (the strands that make up the bundle that is the self), but the important point is that the self is a composition of elementary parts and does not have an independent existence. The Sarvāstivādins differed from the Sautrāntikas on a number of points, but most importantly the former maintained the existence of the past, present, and future, while the latter only accepted that the present exists. In other words, the Sarvāstivādins accepted a kind of duration of the dharmas that the Sautrāntikas rejected. At the same time, the Sautrāntikas accepted the spatial extension of the bodily dharmas while the Sarvāstivādins held that they were infinitely divisible. Importantly, they agreed that the self was identical to the aggregates and that a self with person-properties did not exist independently of those aggregates.
In contrast, the Pudgalavādins held a view that lay precariously between the Tīrthika view and the Sarvāstivādin-Sautrāntikan view. For the Pudgalavādins the self was dependent upon the aggregates. In this respect it was like the Sarvāstivādin-Sautrāntikan self; however, the Pudgalavādin self did possess person-properties. This latter feature made the Pudgalavādin self similar to the Tīrthika self, but as distinct from the Tīrthikas, Pudgalavādins did not maintain that the self was eternal. It could, though, transmigrate from body to body in rebirth. That the Pudgalavādin self was dependent upon the aggregates, but at the same time possessed person-properties while the aggregates did not, meant that the self and the aggregates were neither the same nor different from each other. The Pudgalavādin self was, in this way, “inexpressible.” Perhaps the closest Western notion to the Pudgalavādin view is that of a form of supervenience. The self is dependent upon the aggregates, but does have the same ontological status as the aggregates. It is not substantive as is a Cartesian or Tīrthikan self; yet, it does possess a mind and body and has the capacity to perceive, think, feel, act, etc. It is no wonder that orthodox Buddhists greeted this view with extreme skepticism.
Candrakīrti rejected all of these views and carried to completion the refutation of the self begun by the Buddhist tradition. His refutation relied on a distinction that all of the previous schools of Buddhism accepted but did not make the most of. Each school recognized two forms of truth: conventional (saṃvṛtisatya) and ultimate (paramārthasatya). By asserting that the self is a collection of aggregates and that reference to the self was a short hand for referring to the aggregates, Buddhists were able to maintain that the existence of the self was of a different order than the existence of the aggregates. That is, the self existed conventionally, while what ultimately existed were the aggregates or the elementary dharmas that composed the aggregates. This allowed Buddhists to maintain that it was conventionally true that the self “existed,” while at the same time maintaining that it was not ultimately true. When the Buddha spoke of the self or made use of personal pronouns, he was asserting facts that were merely conventionally true. Both the Pudgalavādins and the Sarvāstivādin-Sautrāntikan made use of this distinction and both accepted that the aggregates – or more precisely, the dharmas – had an ultimate existence.
It is this last claim that Candrakīrti and the Madhyamikas rejected. Their critical premise was that all things with which we are normally acquainted arise dependently. That is, their existence relies on the existence of other things. This includes even the dharmas, the elemental building blocks of the aggregates. In light of this, the self had no ultimate basis at all. All things, including the self, neither existed (independently) nor did not exist. Instead, they maintained what provisionally might be thought of as a third ontological status between existence and non-existence known as “emptiness” (śūnyatā). One might see this as similar to the Pudgalavādin claim that the self was neither the same as nor different from the aggregates, but the similarity is only superficial. The Pudgalavādins located the “inexpressible self” within the conventional realm, while accepting the ultimate reality of the aggregates. Consequently, the self had a basis in the ultimate realm. Against this, the Madhyamikas drew the conventional-ultimate distinction not between the self and its component parts, but between all experience and a transcendent realm accessible only to the enlightened. The illusion of the self as ultimate or as being composed of ultimate elements was what anchored us in samsāra – this delusional world of suffering. Candrakīrti and the Madhyamikas were thus able to acknowledge the purely conventional existence of the self while completely purging it of any ultimate reality. This, more than any other Buddhist theory of the self, was able to interpret the doctrine of anātman in its most rigorous form, while making sense of our (and the Buddha’s) use of personal pronouns.
Relying solely on Buddhism’s root texts upon which these distinctions are based makes for difficult study. Consequently, commentaries and other secondary literature are of great value. James Duerlinger’s The Refutation of the Self in Indian Buddhism along with his early work Indian Buddhist Theories of Persons stand among the most helpful aids to understanding the critical and intriguing Buddhist doctrines of the self.
1. Duerlinger, J. Indian Buddhist Theories of Persons: Vasubandhu’s “Refutation of the Theory of the Self,” (London: Routledge, 2003).