Vasubandhu’s Argument Against Atomism in the Twenty Verses

Vasubandhu's Argument Against Atomism in the Twenty Verses

by Lorand Bruhacs

July 20, 2008To my daughter Miriam, for her kindness and understanding.

Contents 1 Introduction 1 1.1 About Vasubandhu 2 1.2 The Vimsatika in a Nutshell 5 1.3 Rationale and Method of the Investigation 12 1.4 Buddhism as Empiricist Mysticism 17 2 The Atomist Metaphysics of the Abhidharma 30 2.1 Realism and Atomism 30 2.2 Early Buddhist Phenomenalism 32 2.3 The Origins of Buddhist Atomism 39 2.4 The Vaibhasika School 42 2.4.1 Time, Space and Atom in Vaibhasika 44 2.4.2 Critical Review of Vaibhasika Presentationalism ... 52 2.5 The Sautrantika School 57 2.5.1 Time, Space and Atom in Early Sautrantika 59 2.5.2 Atomism in the Late Sautrantika Phase 65 2.5.3 Critique of Sautrantika Representationalism 67 3 The Anti-Atomist Metaphysics of Vijnaptimatra 71 3.1 Readings of Mind-Only 72 3.1.1 Revisionist Interpretations of Mind-Only 73 3.1.2 The Traditional Interpretation of Cittamatra 79 3.1.3 Distinguishing True and False Aspectarianism 81 3.2 Argument against Atomism 85 3.2.1 The Scheme ofthe Argument 86 3.2.2 Against Spatially Differentiated Simples 89 3.2.3 Against Point-like Atoms 96 3.2.4 Against Integral Wholes 98 Contents 3.2.5 Against the Causal Theory of Perception 101 3.2.6 Resumee of the Argument 102 3.3 Visions of Vijnaptimatra in Empiriocriticist Eyes 104 4 The End of the End of Epistemology 111 4.1 Anti-Cartesian Meditations 112 4.2 The Empire Strikes Back and the Return of the Empiricists . 117 4.3 Concluding Reflections 119

Bibliography 121

Index of Subjects and Technical Terms 133

Index of Personal Names 139

Appendix I

1 Introduction

The world is given but once. Nothing is reflected. The original and mirror-image are identical. The world extended in space and time is but our representation. (Erwin Schrodinger)

HIS PAPER WILL conduct a critical investigation of the famous ar­gument against atomism first made by the 4th century CE Indian Buddhist philosopher Vasubandhu in his idealist treatise Vimsatika

Vijnaptimatratasiddhi (The Twenty Verses of Mind-Only). Although the present exposition will be more conceptual than historical in focus, it will first unfold the Abhidharmic Buddhist precursors of the Mind-Only epistemology. With the necessary background in place, I shall then attempt a rational reconstruction of the substance of Vasubandhu’s argument against atomism, rendering it intelligible to the modern reader by transposing it into contem­porary philosophical idiom. Finally, I will employ the analysis of atomism and the external world in the Mind-Only school as a point of departure from which to further probe closely related concerns of Buddhist transcendental philosophy having to do with the nature of empirical knowledge, the power of skeptical argument, and the status of apperception.

Before we begin, a brief clarification—the sense of the term atom as used throughout this paper should be carefully distinguished from the so-called “atom” of the modern physical sciences. Unless otherwise noted, I will be using the term atom to mean a mereological atom, or indivisible, concrete particular functioning as an elementary building block of phenomenal expe­rience.

The theory that the world is constituted by such mereological atoms is commonly believed to be the exclusive province of Western philosophical tradition, beginning with the Greek atomists, but is now known to have evolved, more or less independently from the West, within classical Indian philosophy, and last but not least, among the Buddhist scholastics of the Abhidharma sects.

In the Buddhist tradition, Vasubandhu occupies the unique position of being counted as one of Abhidharma’s greatest proponents as well as one of its foremost critics. The Vimsatika is a Mahayana text and belongs to the later, critical phase of his thought.

1.1 About Vasubandhu

Vasubandhu tends to be a somewhat enigmatic figure even to those familiar with East Asian philosophy or the particulars of classical Indian thought. Before going into more detail on the Vimsatika’s philosophical ideas, I would like to give some brief background on Vasubandhu’s biography and other treatises (sastras). The extant biographies of Vasubandhu contain many contradictions and legendary elements. Lamotte has noted the difficulty of coming to any solid conclusions about Vasubandhu as a person: “Vasubandhu a trop vecu, trop pense, trop ecrit; et avant de se prononcer sur sa personalite, il faudrait avoir lu, critique et compare toutes ses oeuvres.” Vasubandhu was born around the year 316 CE in Purusapura, at the time the seventh most populous city in the world, located in what are now the North-West Territories of Pakistan. The city of Purusapura had become a regional capital of the empire of Gandhara under the Kushan king Kaniska I, who reigned from approximately 127 CE, and who is famous for having erected a giant stupa, perhaps the tallest building in the world of that time, to house the Buddha’s relics. Although the Empire of Gandhara was already past its heydey by the time Vasubandhu was born, Purusapura was still a great center of Buddhist learning. Anacker relates that it was the birthplace of the scholars Dharmasri and the Bhadhanta Dharmatrata, both of the Sarvastivada school of Buddhism, and that it was known as the seat of the Pascatiyas (“Western masters”) of Abhidharma.

Vasubandhu received his initial education in the Abhidharma of the Sar­vastivada sect. The sources are in some disagreement as to who his teachers were—according to Paramartha, Vasubandhu’s 6th century Chinese trans­lator, Vasubandhu’s teacher was a scholar by the name of Buddhamitra, whereas other sources say that he studied under the Sautrantika teacher Manoratha and never mention Buddhamitra. The medieval Tibetan doxog- rapher Bu-ston relates that he was a student of the master Sarighabhadra, a staunch defender of neo-orthodox Kashmiri Vaibhasika scholasticism. Whether or not Sanghabhadra actually was his teacher, Vasubandhu was firmly grounded in the Abhidharmic doctrines of his era, though tending already in the early phase of his thought towards a Sautrantika perspective.

For this reason, the two scholars came to develop an intense rivalry and were often pitted against each other in debate. Later in his career, however, Vasubandhu had largely lost his zest for competitive debating—it is reported that he brushed off a challenge to debate Sanghabhadra with a caustic witticism showing that there was no friendship between the two: “Though the lion retires far off before the pig, nonetheless the wise will know which of the two is best in strength.” Vasubandhu’s ironic eulogy for Sanghabhadra shows that he felt him to have been a sophistic debater lacking in the profound insight that he valued: “Sanghabhadra was a clever and ingenious scholar. His intellective powers were not deep, but his dialectics were always to the point.”

This remark, however, is more than merely a matter of personal animosity. It brings to the fore the ideological divide between the non-Mahayana and Mahayana inflections of philosophical-religious thought. For Sanghabhadra, dialectic is apologetic, whereas Vasubandhu employs it in a mainly anti- apologetic role. Against Sarighabhadra’s orthodox dogmatism, Vasubandhu is a figure of transition and a creatively deconstructive thinker.

The early works of Vasubandhu are commonly classified as belonging to the Sautrantika school, e.g. the Abhidharmakosa (Treasury of Higher Doctrine') and its auto-commentary. The later Vasubandhu is however considered to be a Mahayanist and an adherent of the idealist Mind-Only school. The Mind-Only school is variously referred to in the literature by the Sanskrit terms Yogacara, Cittamatra and Vijnanavada. I am assuming for the sake of simplicity that these names can be used interchangeably, but some commentators assert that the terms designate philosophically distinct filiations of Mahayana thought, e.g. NAGAO, who uses Yogacara as a term for the beginnings and Vijnanavada for the later evolution of the school. Some of the important Mahayana sastras of Vasubandhu’s later phase are the Karmasiddhiprakarana (Discussion for the Demonstration of Action), the Vimsatika-karika (Twenty Verses of Mind-Only) and its auto-commentary, the Trimsika (Thirty Verses), the Madhyantavibhangabhasya (Commentary on the Separation of the Middle from Extremes) and the Trisvabhavanirdesa (Treatise on the Three Natures). However, even his later Mahayana works evince many Abhidharmic themes.

It has been conjectured (e.g. by FRAUWALLNER) that Vasubandhu is a composite figure. The issue of the identity of Vasubandhu remains to this day a major topic of controversy, but for the theses being argued here, it will be sufficient if the Vasubandhus of the Kosa and the Vimsatika are the same author. This circumstance is fortunately not under dispute— according to Frauwallner’s distribution, both works belong to “Vasubandhu the Younger.”

1.2 The Vimsatika in a Nutshell

As it will not be possible to comprehensively examine all of the Vimsatika’s arguments in their minute details, I will only give a lightning tour of its key points and excerpt the epistemological themes of Vasubandhu’s thought in this treatise. The Vimsatika is a locus classicus of the Mind-Only school, and is written in the style of a debate, presenting the idealist thesis of vijnapti- matra and defending it against various interlocutors who argue realist views: “The goal of Vimsatika is to repudiate the view that there is an external world corresponding to the images of objects.”

Vasubandhu indicates in I that he uses the nomenclature citta, manas, vijhana and vijhapti interchangeably. The translations of these terms are respectively ‘mind,’ ‘cognition,’ ‘awareness,’ and ‘percept.’ Given that other In­dian schools of philosophy do distinguish some of these terms quite sharply, Vasubandhu is making a far from trivial move in simply equating their meanings. Doesn’t it make a considerable difference as to whether one in­terprets the doctrine of Mind-Only as mentalism (as the sole existence of consciousness), or as positivism (all knowledge is based on sense experi­ence)? Perhaps not. The treatment of these terms as exact synonyms implies far-reaching consequences for the Mind-Only understanding of the nature of experience and function of consciousness. The fact that Vasubandhu amal­gamates so many different terms indicates that mentalistic vocabulary is probably an implicit locus of philosophical synthesis.

The Vimsatika begins in 1 with a succinct statement of the fundamental principle that everything is perception-only, and that the appearance of the non-existent external referent (bahyartha) of perception can be compared to the optical illusions seen by the taimirikas, i.e. those persons who perceive illusory hairs or moons in their field of vision owing to a medical disorder of the eyes—a famous statement that epitomizes the Mind-Only idealist view. Wood characterizes this view essentially as a brand of idealist phe­nomenalism—the epistemologically naive, common-sense apprehension of an external world is (on that particular count) wildly inaccurate, because people actually only perceive their own percepts. They are familiar only with the phenomenal workings within their own minds.

Several caveats before we proceed with the discussion—in contemporary scholarship, the primacy of the idealist interpretation has been at times hotly contested, e.g. by LUSTHAUS, WAYMAN, Anacker and KOCHUMUTTOM. These scholars advance revisionist readings that in recent years have gained an increasing amount of currency. The revisionist interpretations generally do not take Vasubandhu to deny external objects, thereby placing themselves in opposition to the “traditional interpretation” advocated by many Western scholars such as WOOD, as well as by medieval Tibetan doxographical accounts of the Cittamatra system of tenets. I will have more to say about the topic of these variant interpretations later on in this paper, but for the time being, I will adopt the frame of traditional interpretation as so not to prematurely introduce unnecessary complexities into the discussion. Interpretive frame-shifts will be less confusing once the fundamentals are in place.

In verse 2, Vasubandhu immediately encounters the first objection of the realist—if they are mere mental creations, how can objects be determinate and definite as to time and place? After all, some of the most salient charac­teristics of physical objects are their apparent spatiotemporal determinacy, stability and continuity (niyama), their actuality or causally efficacious func­tioning (krtakriya), and their public perceptibility, i.e. the indetermination (aniyama) of mental continua with regard to public percepts. External objects, argues the realist, are not perceived in an arbitrary fashion because their behavior conforms to certain objective regularities and constraints— they do not shift about unpredictably, vanish and reappear at random, or gratuitously alter into completely different objects.

Vasubandhu does not deny that the observed world evinces such regulari­ties. He merely questions whether they necessarily imply the metaphysical hypothesis of an objectively existing external world, especially since this hypothesis could be held to imply the anathematic existence of acausal ob­jects. He agrees with the requirement that experience be subject to causal constraint, but with an internalist twist—that the conditioning factors are to be found in the mind alone, and not in an external world. In accordance with Buddhism’s generally Humean notion of causality, the phenomenal laws governing the behavior of the supposedly external objects are no more and no less strictly necessary and universal than the inner laws of own subjectivity. Their lawlike behaviour can be accounted for without any appeal to a transcendentally objective realm. Vasubandhu claims that the dream hypothesis, despite seeming counterintuitive at first blush, ultimately offers a better explanation for the regularities in our experience than meta­physical realism.

Verse 3 introduces Vasubandhu’s famous analogy of perception “as in dreams” (svapnavat). Consciousness projects percepts as external to the mind. The mere experience of objectivity therefore does not suffice as proof that the objects of perception exist independently. In the special case of dreams, even a hard-boiled physical realist will have no difficulty in saying that phenomena are perceived in absence of contact with an external world. But Vasubandhu’s thesis is more than merely an example of argument from illusion—he proposes that this inside-out mode of perception, far from being an exceptional case, is in fact the general rule. Dreaming is the paradigmatic example for perception.

As for the question of spatial and temporal location, Vasubandhu off­handedly dismisses the realist objection, pointing out that the spatiotemporal determination experienced in dreams is perfectly sufficient to account for the perception of an apparently external world. In attributing the structure of experience to purely internal substrates, and absorbing the transcendental object into the mind, Vasubandhu makes objectivity wholly parasitic upon subjectivity. In this respect, he can be rightly said to veer sharply towards idealism.

The indetermination of mind-streams (non-exclusiveness of percepts), Vasubandhu argues, is just as in the case of hungry ghosts (pretas), who all see rivers of water as pus due to their identical dispositional bias of perception. That sentient beings perceive a common world proves only that their karmic propensities have the same common denominator. As an example, Vasubandhu mentions beings whose negative karmic propensities will eventually come to fruition as a rebirth in a mentally projected infernal realm. Thus, Vasubandhu argues that his vijnapti-matra thesis is perfectly in accordance with the accepted Buddhist moral theory of karma (which we might characterize in the modern terminology of philosophical ethics as having a consequentialist orientation).

To Vasubandhu, causality and lawlike regularities, being merely empirical regulatives and therefore wholly internal to experience, do not in and of themselves suffice to offer conclusive proof of objectivity. In 4, he compares the efficacity of action (arthakriya) to nocturnal emission of semen. We might unpack this rather risque example as trying to illustrate how even wholly imaginary actions (e.g. having sexual intercourse in a dream) can nonetheless produce pragmatic effects (the emission of semen). Vasubandhu expands upon this example in his ingenious hell-guardian argument in 5, where he urges that the demonic guardians (naraka-palas) who torture their victims are nonexistent because they themselves are not being tortured. It is an axiomatic assumption of early medieval Buddhist cosmology that hell-dwellers were reborn in a hellish realm in order to experience infernal sufferings as retributions for their past actions. But since the naraka-palas inflict suffering upon others rather than experience it themselves, it follows that it is impossible for their true manner of existence to accord with their appearance of being real, sentient beings. Vasubandhu thus concludes that hell and its guardians are mental projections, suggesting that the doctrine of karma logically implies his idealist thesis.

Moreover, if hell, the worst logically conceivable case of suffering, is in­admissible as a Moorean proof for the metaphysically objective existence of hell, then the same applies a fortiriori to all other possible states of expe­rience. Vasubandhu even coopts such realist proofs as support for his own thesis. Hell, Vasubandhu would diagnose, is nothing other than the painful symptom of a very severe case of Johnsonian stone-kicking.

In response to Mach’s sardonic aperqu?1 about the Pyrrhonist in Moliere’s Mariage force who is beaten and does not go on saying “Il me semble que vous me battez,” but takes his beating as really received, Vasubandhu might answer that his case against the metaphysical existence of an external world is predicated on a specifically transcendental sense of “seems talk.” In modern terms, we might construe Vasubandhu’s position of transcendental illusionism as advocating an internalist account of knowledge similar to the analysis of Cohen. The hell-dweller argument tries to show that if the principle of ethical consequence is to be preserved, then suffering (such as receiving a beating)—and everything else that masquerades as the external world—must be a quasi-oneiric manifestation of mental causative potentials (vasanas). In a dream, the dreamed tortures and terrors are experienced as very real indeed, but upon awakening, they evaporate. The dreamer recognizes that all of his experiences were, in truth, no more than projective appearances of his mind’s inherent muddle.

Next, Vasubandhu gives an argument in 8 against the existence of two types of sense-base (ay at ana)—sensory faculties and their external sensory fields—asserting that external sensory fields (visibles, tangibles, audibles, etc.) do not actually exist. Vasubandhu justifies this claim by means of a scriptural hermeneutic. He maintains that the statements regarding the existence of the sensory fields have a merely provisional meaning as metaphorical figures of speech set forth with the hidden (neyartha) intention of accomodating realists to the Buddhadharma.

Taught in one way, Vasubandhu argues in 10, the doctrine of the Buddha leads to the lower understanding of the insubstantiality of the personality (pudgala-nairatmya) and taught in its deeper meaning as Mind-Only, it leads to the higher understanding of the insubstantiality of the elements of experience (dharma-nairatmya). But aside from this hermeneutic justifica­tion, he also gives a rational line of reasoning—that the existence of external sensory fields implies metaphysical realism, which seems an acceptable enough theory at first glance, but is revealed upon closer examination to be unreasonable.

Vasubandhu then proceeds in 11-15 to give an inventive reductio demon­strating why the various forms of metaphysical realism fail to be a coherent alternative to his dream hypothesis of perception. The transcendental ob­jects postulated by realism do not exist because it is impossible for them to be composed of atoms. Since Vasubandhu crystallizes the dispute over transcendental objectivity (i.e. the existence of a mind-independent external world) around the issue of atomism, he devotes a sizeable portion of the Vimsatika to a demonstration that the metaphysical existence of atoms is unproven—a treatment which is akin to the Kantian analysis of the issue, as we will explore in detail in Chapter 3. I shall argue that, unlike Kant, Vasubandhu also rejects the necessity of noumenal simples to the constitu­tion of phenomenal reality. Kant’s deep suspicion about the legitimacy of such a philosophical move is due to his correct intuition that it would signify a final departure from realism. But this is of course precisely the paradigm shift that Vasubandhu wishes to urge.

In 16, Vasubandhu refutes the causal theory of perception and then an­swers to a number of further realist objections against the dream theory of experience, having to do with how recollection of objects or differences in the dream and waking state can be accounted for. Just as those who are dreaming do not realize the inexistence of the objects in their dreams, he argues in XXXIII, those who lack world-transcending gnosis fail to realize insubstantiality.

In a similar vein, Essler remarks that just as people rarely realize they are dreaming while they are having a dream, they are likewise unaware of the extent to which they epistemologically constitute their world when awake. Such is the nature of dreaming, Mach ironizes—while in a dream, one may reflect about the dream and recognize it as such by its oddities, but one immediately puts one’s mind to rest about them. Vasubandhu’s com­parison of this constitutional creativity of the mind to dream consciousness is suggestive of the language of the Mandukya Upanisad? It strongly hints at the transformative aim of the text—if it is really true that our world is like a dream, then we can hope to awaken from it.

Since his realist interlocutor in 18 is also Buddhist, Vasubandhu is in the position of using scriptural authority to illustrate his idea that minds do not require the intermediation of a mind-independent external world to influence each other. In the Mind-Only paradigm, the existence of a tele­pathic influence directly connecting the different mind-streams is not only invoked in order to absolve Mind-Only of solipsism or to explain anomalous cases, but is employed as a central explanatory mechanism for ordinary perception.In realism, common perceptions are accounted for by the external object, which causally determines the sensory core of perception. In Mind-Only idealism, the external object is merely a common locus item.

In the intriguing argument of the forest-dwelling seers in 20, Vasubandhu attempts to rebut the charge of solipsism by proposing that mental states and perceptions exercise a direct causal influence on other consciousnesses, as demonstrated by a scriptural example involving the extraordinary mental powers of certain forest hermits. This causal influence places the various minds in immediate mind-to-mind contact (paras-patah) with each other. Despite that we are unaware of this contact, Vasubandhu suggests, it is the means by which we intersubjectively constitute afolie simultanee that then appears to us as the objective world.

In the penultimate verse 21, Vasubandhu directly addresses the other minds problem. Not only do we not truly know other minds, we do not even know our own. Essentially, Vasubandhu frames the other minds problem as a special case of the more general problem of the epistemic status of the transcendental subject, for which he offers a deeply skeptical response.

1.3 Rationale and Method of the Investigation

Why single out Vasubandhu’s argument against atomism for further exami­nation? The Vimsatika’s argument against atomism demarcates a crucial juncture both in Vasubandhu’s own thought as well as in Northern Buddhist thought in general—the two non-Mahayana schools of Northern Buddhism (the Sarvastivada/Vaibhasika and the Sautrantika) commonly accept the existence of atoms, whereas the two Mahayana schools, the Yogacara/Citta- matra (Yogic Practice/Mind-Only) and the Madhyamika (Middle Way) do not. Also, in using Vasubandhu’s argument as a mirror in which to study the epistemological interrelation of mind and object, I will offer some contem­porary reflections on two perennial themes of philosophy—empiricism and Cartesian external world skepticism.

However, this effort runs into certain initial difficulties, concerning not so much the analysis of Vasubandhu’s anti-atomist argument in and of itself— which will prove fairly straightforward—but rather, having to do with the hermeneutic challenges of properly assessing its significance against the broader backdrop of classical Buddhist thought and analyzing it in the light of contemporary epistemology.

On the one hand, any attempt to draw out the meaning of Vasubandhu’s argument must take account of the unique historical context and the dis­tinctive philosophical moves generated by the Buddhist program. On the other hand, an exegesis from a contemporary standpoint will necessarily be grounded, at least in part, within the matrix of the Occidental philosophical traditions, which prima facie seems very far removed from the worldview of Indian Buddhism around the 4th century CE. Herbert V. Guenther has often remarked on the difficulty of applying West­ern philosophical categories to Eastern patterns of thought, warning that an overly facile characterization in Western terms runs the risk of blurring or even entirely misconstruing the distinguishing features of Buddhism.

Extrapolating grand trends and tendencies is tricky business in any philo­sophical context, and especially so when it comes to the history of Buddhist thought. One must be careful not to cavalierly colonize the texts of the Other. But a comparative analysis unavoidably requires a certain degree of creative extrapolation and generalization. Clearly, in the worst case, such an approach may reveal little more than the expositor’s hermeneutic bias. Yet it would be equally naive, in a philosophical sense, to desist from bringing in current vocabulary and concerns in an attempt to eliminate Western perspective from the equation.

Although it would be possible to restrict oneself to a purely text-immanent interpretation of Vasubandhu’s Vimsatika, such a treatment of this type would be of little more than historiographical or antiquarian relevance. Indeed, it would be rather difficult to conceive of a philosophically relevant study of Vasubandhu’s position on atomism that does not attempt to line him up with Western thinkers and commensurate his ideas to familiar Western categories.

One challenge is therefore to employ the inescapable perspectival slant of a contemporary interpretation to good effect in bringing Vasubandhu’s arguments to bear upon topics of interest to present-day philosophers. As B. K. Matilal notes, contemporary writing on classical Indian philosophy must somehow strike a balance between the reconstruction of historical views and the critical examination of similar modern views. Some knowledge of his­torical development will prove indispensable as a foundation for abstracting the underlying conceptual structure of Vasubandhu’s line of argument and reconstructing its main epistemological themes and issues. The project of philosophical re-synthesis in the guise of comparative analysis requires that the discussion fluidly shift back and forth between ideo-historical exploration and rational reconstruction.

In particular, I hope to identify some of the central philosophical features of Buddhist thought by casting the Vimsatika, in an empiricist mold, juxtapos­ing Buddhist thinkers such as Gautama Sakyamuni and Vasubandhu with Kant as well with Avenarius and Mach, Kant’s 19th century empiriocriticist successors. The main reason why I do not elaborate the obvious parallels to Berkeley or Schopenhauer is not because it would be uninstructive or inappropriate, but because the comparisons have already been extensively pursued elsewhere. Instead, I will emphasize the comparison of Mind-Only epistemological idealism to to the post-Kantian empiriocriticist phenom­enalism of Mach. Since Vasubandhu’s Abhidharmakosa phase of thought arguably already makes a Kantian turn and the Vimsatika represents a further critical step in his development, it seems that Mind-Only might profitably be re-evaluated as a continuation of the Kantian trajectory of crit­ical empiricism (in other words, as an epistemology of “Experience-Only”) rather than a reversion to the pre-Kantian idealism of Berkeley.

Another reason for reading Vasubandhu in an empiriocriticist vein is that both Mach and Yogacara Buddhism may be said to have advocated a pure phenomenalism. The parallel to Mach is crucial to my interpretation due not only to his sophisticated (albeit unsystematic) phenomenalist point of view, but also to his notoriously die-hard skepticism about the theory of physical atomism, which will help to illustrate a plausible philosophical rationale for Vasubandhu’s anti-atomism. Just as the empiriocriticism of Mach and Avenarius may be regarded as an empiricism that has purified itself of transcendental objectivity and ontological absolutes, the Vijnapti- matra of Vasubandhu amounts to a Sautrantika view that has eliminated the external object (bdhydrtha). I am aware that I am suggesting a deliberately prochronistic interpretation of Vasubandhu’s views. It is by no means an authoritative reading, but hopefully, an interesting and illuminating one.

The various philosophical ideas of the Indian ancients will be articulated largely from the perspective of analytic philosophy, while also attending to the unique concerns that they implicitly convey. The basic motivations of the classical Indian traditions seem largely alien to a contemporary philosophical audience because they were influenced by a very different historical setting and cultural outlook. Like most classical Indian philosophy, Buddhist thought of the early medieval era intimately fuses (and inevitably, often intimately confuses) rigorous philosophical analysis with dogmatic and sote- riological concerns, to a degree rarely encountered in post-Enlightenment Western philosophy.

Moreover, Mahayana Buddhist thinkers are almost universally mysti­cally inclined, and Vasubandhu, a mystic of the first water, is certainly no exception. An in-depth exploration of Mahayana mysticism would unfortu­nately be beyond the scope of this paper—a brief survey will have to suffice. But despite that for the purposes of this discussion, I will be singling out the philosophical aspect of Vasubandhu’s thought (i.e. the level of rational argument), I will not be able to avoid making some basic interpretative allowances for its religio-mystical dimensions. The subtle interplay of philo­sophical argument and mystical anagoge might seem difficult to reconcile to the modern philosophical mind. However, the lines of rational argument are fortunately easier to discern in the Vimsatika than in other writings of his later phase, e.g. the Trimsika or the Trisvabhavanirdesa.

The exceedingly condensed, epigrammatic nature of Vasubandhu’s writing poses another interpretive challenge. In stark contrast to contemporary Western philosophy, which tends to valorize lengthy prose expositions that discursively expand upon their themes in exhaustive technical detail, the Indian philosophy of that period placed a premium on brevity and poetic elegance of expression. It had evolved a philosophical poetry, written in a strictly metered form typical of classical Sanskrit composition. Key philo­sophical themes were epitomized in gnomic verse and expanded upon in the author’s auto-commentary. The most brilliant examples of this style evoke an air of tension, paradoxicality and mystery with their austere reductions, unexpected double-turns and double-entendre, and constant shifting of the level of discourse.

Part of the reason for the use of this poetic form was to facilitate memo­rization, but another motivation is that verse can more easily convey elusive aesthetic subtleties and multiple levels of meaning. A poem can be read as a philosophical epitome, operating on the level of rational explanation, and on a deeper reading, function as a mystical pith instruction, which follows the entirely different modus operandi of deliberately perplexing the intellect so as to induce transcendence of conceptual limitations.

The paradoxical mode of expression is an integral part of the mystical atti­tude, and past a certain point, it defies interpretative reduction to a purely rational meaning. Vasubandhu concludes the Vimsatika in 22 with a Trac- tarian cliff-hanger—even he, the author himself, cannot fully understand the meaning of what he has written: It is impossible for people like me to consider it [the doctrine of perception-only] in all its aspects, because it is not in the range of dialectics. And in order to show by whom it is known entirely as a scope of insight, it is said to be the scope of Buddhas.

1.4 Buddhism as Empiricist Mysticism

Unless it is read in the context of a meta-philosophical “grand narrative” or background hermeneutic, the technical analysis of Vasubandhu’s arguments against atomism will seem rather dry. The deeply apophatic finale of the Vimsatika offers a hint as to where the meta-narrative ought to begin. We will bring the hermeneutic lens of Vasubandhu as mystical proto-empirio- criticist into sharper focus by offering some intertextual reflections on the meaning of the silence of the Buddha. This is the silence that Gautama Sakyamuni is said to have maintained with regard to the famous fourteen unanswered questions (catvari-avyakrta- vastuni) posed to the Buddha by the wandering ascetic Vatsagotra:

Whether the cosmos has a beginning, or not, or both, or neither. Whether the cosmos has an end, or not, or both, or neither. Whether the Tathagata exists after death, or not, or both, or neither.

Whether the mind is identical with the body or different from it.

Although he held silence for many other reasons as well, Sakyamuni’s silence about these questions of a specifically metaphysical nature is par­ticularly significant, in that it prefigures the later Madhyamika program of insubstantialism and null ontology. Madhyamika develops the Buddha’s silence about absolutes into a profoundly meontological mysticism, deploying skeptical argument against all points of view in an attempt to demonstrate that the most important thing the Buddha said was nothing. The fourteen unanswered questions moreover present a fascinating paral­lel to the Kantian antinomies of pure reason, where Kant raises aporiai in demonstrating that the four ideas of rational cosmology each lead to two opposing positions that appear to be equally valid.

The first eight unanswered questions of Vatsagotra closely approximate the first Kantian antinomy of pure reason, as both reduce to aporia the question as to whether or not the world is spatiotemporally limited. The second of Kant’s antinomies, which happens not to be among the fourteen unanswered questions of sutric tradition, has to do with whether or not the world of experience is metaphysically constituted of simple or primitive elements, i.e. atoms. The atomists of the Abhidharma schools answer this question in the affirmative, whereas in the Vimsatika, Vasubandhu categorically responds in the negative.

As for what position Gautama Sakyamuni himself may have taken on atomism—although the earliest Buddhist corpus of texts, the Pali Canon, does not mention atomist doctrines or offer any clues as to whether or not the Sakyamuni adhered to a theory of simples, in light of his silence on Vatsagotra’s questions, one can easily picture what his response might have been. I conjecture that he would have recognized the aporetic character of the issue and remained silent.

Concerning his own rationale for keeping silence about the fourteen ques­tions, Sakyamuni does not have much to offer in the way of explanation, aside from the famous fire simile in the discourse (sutra) of Vaccagotta on Fire. It seems that Sakyamuni intended the simile to specifically address Vatsagotra’s questions about the post-mortem status of a monk who has reached nirvana, but it could also be construed, without too much of a stretch, to apply analogously to the other questions as well:

“And suppose someone were to ask you, ‘This fire that has gone out in front of you, in which direction from here has it gone? East? West? North? Or south?’ Thus asked, how would you reply?” “That doesn’t apply, Master Gotama. Any fire burning depen­dent on a sustenance of grass and timber, being unnourished— from having consumed that sustenance and not being offered any other—is classified simply as ‘out’ (unbound).”

Aside from the Kantian analogy, there is also a Machian parallel to the fire discourse. Mach uses examples that are strikingly similar to the sutric simile when he accuses his fellow physicists of naively misconstruing concepts which properly apply strictly to sensations to refer to objective physical quantities. In the opinion of Mach, many putative scientific problems can be traced back to naive category errors and unmasked as philosophical pseudo-questions: ‘Where does the light go when it has been extinguished and no longer fills the room?’ Such are the questions of a child. Con­founded by the sudden shrinking of a hydrogen balloon, the child continues to search for the large body that had still been present just a moment ago. ‘Where does the warmth come from? Where does it go?’ Such childish questions coming from the mouths of mature men determine the intellectual climate of the current century.

In view of Mach’s quintessentially positivist theory of the pseudo-problem (Scheinproblem) and its purported root cause of metaphysical thinking, one might deduce that he would be very reluctant to acknowledge the existence of genuine aporiai, i.e. philosophical conundrums that are fundamentally and intrinsically intractable. Mach believes that philosophical problems are either solved or discovered to be nonsensical and futile.

It is certainly not too much to assume that Sakyamuni was aware of the existence of pseudo-problems. Therefore, we may guess that if Gautama Sakyamuni had believed the fourteen unanswered questions to be no more than pseudo-problems, he would have plainly said as much. Panikkar conjectures Sakyamuni’s silence about the questions to be not an epistemic, but an ontic apophaticism of a metaphysically avyakta nature. In Indian philosophy, the term avyakta also carries the connotations ‘unmanifest,’ ‘all­pervading,’ ‘formless,’ ‘non-differentiated,’ ‘indeterminate’ and ‘unexpressed.’ If understood in an ontological sense, we can see how avyakta might perhaps be comparable to Anaximander’s concept of the apeiron. Moreover, it is worth noting that the notion of avyakta can be considered in some ways a precursor of the concept of sunyata, which later attained its full development in Mahayana thought. The ontologically negative orientation of sunyata is a Mahayanistic reprise of the early Buddhist anatman principle.

Therefore, it may be argued that the Buddha’s silence was more than merely a positivistic epoche about the metaphysicalities beyond the limits of experience—it was the metaphysical silence, or apophaticism of the cen­terlessness and ineffability of experience itself, in its immanent immediacy. This is, generally speaking, the Madhyamika position on the issue. We will not be able to conclusively establish whether Sakyamuni’s own understanding of avyakta questions was similar to Mach’s doctrine of the pseudo-problem arising from superstitions, misunderstandings, category mistakes and misapplications of language, or whether he in fact considered that there are some truly unanswerable philosophical questions pointing towards a mystical reality of “metaphysical silence.”

Even if the Buddha did not articulate an explicit positivism, the empiricist tendency of eliminating metaphysics and adverting to experience is evident. Inasmuch as his silence about the fourteen questions is interpreted as a dismissal of pseudo-problems (a wholly justified interpretation) we may consider it to be the germ of a strong proto-positivist subtext in Buddhism. There are quite a number of writings in early Buddhism that clearly state empiricist viewpoints, e.g. that speculative theories which go beyond the lim­its of experience generate confusion and vexation rather than knowledge. In early Buddhism, one finds—other than instances of the tetralemmatic negation (catuskoti)—little direct evidence of the deeply paradoxical formu­lations and the penchant for puzzling dialectic that were later to become the signature style of the Mahayana schools. On the contrary—Sakyamuni either speaks clearly or keeps silence.

Though he is known in many instances to have expressed a dislike of philosophical debate, I do not conceive Gautama Sakyamuni to have been a rigorous misologist. Rather, I construe his position to have been that con­structive philosophical inquiry and theorizing have a primarily diagnostic or therapeutic purpose—in the language of Buddhist simile, they are rafts that must be abandoned once one has crossed over to the further shore, or ladders that should be cast away after they have been climbed. This soteriological instrumentalism is a characteristic feature of the Buddhist doc­trine, and, especially in the Mahayana, is extended into an epistemological instrumentalism. This is also Vasubandhu’s interest. In the Vimsatika, he introduces concepts not as absolute truths, but solely as an expedient means for dissolving the conceptual rigidities of the metaphysical realist.

The Buddhist skeptical mysticism was later to attain its fullest expression in the Prajnaparamita corpus of scriptures and their various exegeses. The central theme of this corpus is the wisdom (prajha) of voidness (sunyata). The notion of sunyata, though already present in some writings of early Bud­dhism, is taken up and analytically developed within the critical-skeptical framework of the Madhyamika school of philosophy, into the doctrine that the dependently arising elements {dharmas) of empirical experience are ineffable {anabhilapya) and void {sunya)—or to restate the same matter in terms more appropriate to mysticism—apophatically negated. Phenomena in other words cannot be said to have essences, core bearers, intrinsic identities, existence as unitary objects, and so on. The Prajnaparamita radicalizes the metaphysical silence about what lies beyond the limits of experience into a meontology of metaphysical silence.

Based on this distinction of a mystical and a positivistic layer to the silence of the Buddha, we can attempt to hermeneutically project the Buddhist epistemological discourse along two parallel and mutually complementary avenues. The “cataphatic” discourse along the via positivistica (i.e. the posi­tivistic avenue exemplified by the critically purified Abhidharmika approach) attempts to constructively epistemologize experience, whereas the via nega- tiva (paradigmatically, the Madhyamika) tends to frame its epistemology in terms of a skeptical and deconstructive practice.

The hypothesis that there came to be these two closely related but nonethe­less distinct understandings of the Buddha’s silence can provide us with a model for plotting the structural dynamic of the Vimsatika. We will designate the the via positivisticd's metaphysical silence about the metaphysical as the “Cartesian epoche” indicating that a bracketing of the transcendental object is taking place. This catchprase is meant to suggest characteristic modes of philosophical expression such as Cartesian external world skepticism, and to hint at the notion of phenomenological epoche developed by Husserl in his Cartesian Meditations.

The status of the transcendental subject is left untouched by the Carte­sian epoche, allowing for various modalities of metaphysically negating or asserting it that lead to weaker or stronger types of idealism. However, we hypothesize that the weaker modes of idealism (i.e., those obtained by metaphysical negation of the transcendental subject) generally fall closer to the centerline of the via positivistica because they make an effort to ap­proximate a phenomenalist middle ground between realism and idealism. By metaphysically negating the transcendental subject, they are trying to counteract their inherent tendency to gravitate towards ontological ideal­ism. These are in other words idealisms trying to disguise themselves as forms of neutral monism or pluralism.

The metaphysical silence of the Buddhist via negativa, on the other hand, indicates the apophatic negation of the subject. It will simply be equated with the terms sunyata, dharma-nairatmya and anatman. Along with pratltya-samutpada (dependent origination), the doctrine of anatman must count as one of the most distinctively Buddhist doctrines, the anatman doctrine may be viewed as a challenge to the primacy of transcendental subjectivity in the orthodox Vedic philosophies (astika darsanasY While there is no evidence that early Buddhism rejected any kind of metempirical self, it is important to keep in mind that none of the arguments attempting to warrant the opposite thesis (that Buddhism acknowleged an atman) are substantive, nor is there any evidence that Sakyamuni taught an atman theory.

That there is no evidence either way of course does not prove the hy­pothesis that his doctrine of anatman involved an epoche of transcendental subjectivity, but at least allows for the possibility. As for the empirical reality of human existence, Buddhism almost uniformly proposes a strictly bundle-theoretic account (with the exception of the Pudgalavada schools of early Buddhism which maintained the ultimate reality of a personal self). The bundle theory of personhood is also a recurrent theme of Western empiricism—we find it e.g. in the writings of Hume.

In its most radical form, the the Buddhist via negativa may be construed as amounting to a thorough epoche of any and all types of subjectivity. The substantiality of the empirical self, the Cartesian and the Kantian cogitos, as well as the transcendental ego of Husserl and the Tractarian notion of the transcendental subject as world-limit, are all to be apophatically negated.

Buddhism seeks to achieve egolessness by “dissolving” the transcendental ego into the stream of empirical consciousness. After subjectivity has been bracketed, there remains only the pure “difference” of experience without any unitary forms of subjectivity or objectivity which could serve as a foundations for constructing a world or a self. As it is in fact the atman which is the source of all forms of unity and ontological categories falsely projected upon experience, bracketing the atman also eliminates these false projections.

The atman and its cognates therefore become the Buddhist soteriology’s main targets of attack. Early Buddhist scholasticism pursues the positivistic trajectory and attempts to eliminate the subject by a reductive analysis of the elements of experience into objectified least parts. This program is played out as an atomistic sense-data phenomenalism. Vasubandhu, however, puts this Abhidharmic reductionist program to an end by reasoning against transcendental objectivity—a maneuver that brings him closer to the Upanisadic monism of the astika pattern, but also to the original via positivistica of Buddhism that had been derailed by the metaphysical realism of the Abhidharmikas.

In the Western history of philosophy, we can identify the epistemology of Mach as an example of this hypothetical Buddhist via positivistica. Mach, though tracing the sources of his inspiration to Hume, Lichtenberg and William James (giving special credit to Lichtenberg’s famous correction to the Cartesian cogitcr. “Es denkt”) was aware of the general tendencies of Buddhist thought and held them in high esteem, as shown in one of Mach’s letters to Fritz Mauthner.

Nowhere does Mach propound an overtly soteriological aim in the vein of the Buddhist nirvana (the complete cessation of suffering and freedom from afflictive states of mind) or the Pyrrhonian ataraxia (the happiness, freedom from worry and imperturbability arising from the skeptical suspension of belief), nor does Mach have a fully evolved apophatic mysticism as does the Madhyamika. His main focus as a physical scientist is a positivistic one— employing quantitative method to investigate the functional relationships between the elements of experience.

If Mach is able to mentally envision the refraction angles of the color spectrums that are generated by shafts of white light falling onto a prism, if he can predict where the Fraunhofer lines will appear and what changes will occur if the prism is shifted or if a thermometer touching the prism gives a different reading, he is satisfied. Because he views the quantitative laws of physics as mere regulatives of his empirical sensory intuition, he feels that he knows everything he can expect to know. Early Buddhism, on the other hand, evolved in a historical setting in which the quantitative scientific method had not yet been developed. As a result, the aim of the Buddhist empiricism of functional relation naturally gravitated towards acquiring moral and mystical insights into the dependent arising of experience, which were elaborated into the characteristic doctrines of karma and selflessness.

Nevertheless, as a key point of comparison to Buddhism, we have Mach’s famous dictum “The ego cannot be saved.” Mach reflected upon this insight as an ethical saving grace rather than a disaster—he believed that it would help to value things according to their real importance and inspire a more emancipated and altruistic way of life. The Machian and the Buddhist worldviews are versions of what Sommer characterizes as the “inversion of gnosis.” If by gnosis we are to understand a mystical knowledge of the metempirically transcendent reality of atman by turning away from the world of empirical experience—as is the declared aim of Upanisadic mysticism—then the “inversion of gnosis,” its polar opposite, would be a negative immanence mysticism that tries to attain knowledge of dharma-nairatmya by turning towards the world of experience. In inverse gnosis, the world is no longer believed to be that which separates us from true knowledge. Rather, it is considered that which we are separated from by the false dualisms of an internal and an external world, of qualia and substances, sensations and atoms.

The hypothesis that a via positivistica was operational in the earliest forms of Buddhism can help to account for the strong Buddhist tendencies towards phenomenalism and epistemological idealism, which are remarkably suggestive of the empiriocriticist views of Avenarius and Mach. The Carnap of the Aufbau is also firmly rooted in the Machian tradition. His Vienna Circle empiricism contends that the concept of objective reality cannot be epistemologically constituted, and that nothing meaningful can be said either way about the things-in-themselves (Dinge an sich).

To recapitulate the “grand thesis” that is being staked out here: Vasubandhu and the empiriocriticists (including most of their positivist successors) have merely taken different on-ramps onto the same philosophi­cal highway. Mach’s epistemology has been argued by some to be influenced by the Madhyamika. In this exposition we will be following up on a slightly

different lead—the conjecture of a philosophical continuity between Mach and Mind-Only. Admittedly, every interpretive scheme is eventually un­raveled by a sufficiently close attention to the facts—this one is certainly no exception to the rule. However, Indian philosophy is such a dense and confusing topic that we will be unable to make much headway into it unless we cut some broad aisles through the undergrowth.

Following the lead of Dharmakirti and Santaraksita, Tibetan doxographers have traditionally rendered Buddhist doctrine as a ladder of philosophical views, using the method of an ascending scale of analysis to move from common-sense realism towards more rarefied epistemological views. Since this has proven to be a useful approach, we will also be adopting it for the presentation of Vasubandhu’s argument in this thesis.

The itinerary for the following chapters will be to tease out the unifying strands of phenomenalism and nominalism that run throughout the differ­ent schools and historical evolutions of Indian Buddhist epistemology. We will explore how these themes were initially worked out into an atomist metaphysics, and in a later stage of development “empiriocritically” revised by Vasubandhu into a non-atomistic idealism that in many respects may be considered a chip off the block of modern empiricism. We will be focusing our discussion mainly on the epistemological rationale for this critical turn. But the hermeneutic background assumption for this analysis will be that the Vimsatikd’s skeptical recovery of the hypothetical Buddhist via positivistica is ultimately motivated by a profound shift—evident throughout the later Mahayana phase of Vasubandhu’s thought—towards the via negativa. While the metaphysical higher ground of Vijnaptimatra—the restoration of idealis­tic phenonemalism—may the final rung of Vasubandhu’s epistemological ladder, it is not his ultimate position, but rather, a provisional antidote (pratipaksa) to substantialism—a raft to be abandoned after it has been used for its intended purpose.

Before we can explore Vasubandhu’s strategy for regaining the Cartesian epoche and fully draw out the empiriocriticist parallels to the Vimsatika’s ar­gument contra atomism, it will be advantageous to examine the Abhidharmic context of its thought. We cannot expect to reasonably form an opinion on Vasubandhu’s arguments without a careful reading of his realist opponents.2 The Atomist Metaphysics of the Abhidharma

I am sitting with a philosopher in the garden; he says again and again "I know that that's a tree," pointing to a tree that is near us. Someone else arrives and hears this, and I tell them: "This fellow isn't insane. We are only doing philosophy."

(Ludwig Wittgenstein) IN THE TRADITIONAL and modern commentarial literature, the epistemo­logies of the two Abhidharmika schools are universally described as being realistic or substantialistic. While the Mahayana schools can be viewed as idealist and skeptical variations on the basic Buddhist themes of phenomenalism and nominalism, the non-Mahayana schools of Abhidharma comprise their realist inflections. However, given that the nature of realism is a hotly debated topic in contemporary philosophy, what exactly is meant by this claim? A brief review of realism is in order.

2.1 Realism and Atomism

Although there are many different philosophical senses of realism, in the present context, we are only interested in the sense of realism that concerns the everyday world of macroscopic physical objects and its constituents. As we shall see, the position of the Buddhist atomists may be construed as an attempt to formulate an ontologically realist epistemology that deconstructs common-sense realism about ordinary objects—as opposed, for example, to the atomism of the Indian Nyaya-Vaisesika school, which attempts to defend their objective existence.

Realist claims can be grouped into two broad types—(1) the existence dimension involves claims that physical objects exist extramentally, as do the facts pertaining to them, whereas (2) the independence dimension has to do with claims that physical objects are independent of beliefs, linguistic practices and conceptual schemes. Realisms vary considerably in their degree of ontological commitment, but as a general rule of thumb, the more naive (i.e., pre-critical) the realism, the larger its ontological menagerie. Chatterjee ventures that all forms of realism must have atomistic onto­logies because they are necessarily committed to some form of pluralism or other. Non-atomistic systems of thought such as absolute monism (e.g. Advaita Vedanta) or subjective idealism (e.g. Cittamatra) can evince realistic tendencies, but according to Chatterjee do not qualify as unreserved realism. Clearly, realism implies a commitment to an ontology of some cardinality, but why should it necessarily restrict itself to a denumerable ontology? In fact, it needn’t, and realism often employs more powerful ontologies. But for the purpose of this investigation, we will define the atomistic hypothesis, in general terms, as the view that the ontological constitution of matter, stuff, hyle (or whatever else one wishes to call it) is ultimately discrete.

It is also important to distinguish realisms employing foundationalist theories of epistemic justification, which due to their epistemic atomism are prone to lead to an atomistic ontology, from those forms of realism employing a coherence theory of justification, which are holistic, i.e. epistemically non- atomistic, and therefore far less likely to lead to an ontological atomism.

If and when epistemological atomism slips into ontological atomism, as it almost invariably does (though we will see that there are exceptions to the rule), then atomism entails realism. The reverse case—that realism entails atomism—only holds for foundationalist varieties of realism with denumerable material bases of constitution. Moreover, we can categorize realisms into two types, according to whether they adhere to a two-term or a three-term theory of knowledge.The defining characteristic of the two-term theory is that bare perceptual evidence is inherently veridical in a quite strong sense. For the purposes of the present investigation, we will define two subtypes—(1) direct realism and (2) sense-datum realism. Direct realism (1) holds that physical objects themselves are directly apprehended, i.e. when we say we see a tree, we are not merely seeing sense-data—we are seeing the tree itself as an integral whole. In sense-datum realism (2), we are directly seeing the sense-data, which are accorded intrinsic credibility—one’s awareness of a sense datum is immediate and infallible. Sense-data are moreover held to be ontological ultimates. Thus, the evidentially given sense-data are immune to skeptical assault and constitute objectively true foundational knowledge.

Representational realism (the three-term theory) holds that the “veil of perception”—a representation, aspect or sensation—intervenes between the perceiving consciousness and the object.

A metaphysical realism is any position that has not yet overcome the category of the Ding an sich or the transcendental objectivity. The “tran­scendental realist” (to use Kant’s terminology) believes that he directly experiences absolute reality. But the way we are defining it here, realism is metaphysical even if the transcendental object is never experienced in itself (as is the case in critical realism) because some degree of objective backing is still held to be required as a condition of the possibility of phenomenal expe­rience. Critical realism relies on the transcendental reduction to distinguish which parts of experience are supplied by the mind from the parts that are given in sensory intuition.

2.2 Early Buddhist Phenomenalism

The view of pure or radical phenomenalism can be characterized in several ways: (1) epistemically, that things are just as they are known to be in experience and that there is nothing metaphysical to be said about them— (2) ontologically, that there is no real difference between phenomena and their essences, or even that phenomena are essenceless—or (3) semantically, “that propositions asserting the existence of physical objects are equivalent in meaning to propositions asserting that subjects would have certain sequences of sensations were they to have certain others.”

In other words, radical phenomenalism holds that physical objects are in a real sense constituted strictly by the sense-data of perception (and not vice versa), and that propositions about physical objects reduce to propositions about clusters of sensations or sensibilia. Knowing is grounded in the evidence of the senses, which is foundational and immediate. The bare sensations are self-disclosing (and in the case of Mind-Only, self-cognizing). Familiar objects such as material bodies, objects, wholes have only a nominal, conceptually imputed existence. In reality, they are merely bundles of sensory perceptions that are construed as objects. Phenomenalism therefore disputes the claim that objects exist independently of sensory awareness. If no such awareness were to exist, all existence claims about entities would be false.

In what must be acknowledged as one of the “great texts” of early Buddhist epistemology, the Bahiya Sutta? we find a paradigmatic example—Gautama Sakyamuni advises the bark-cloth-wearing ascetic on how to train himself in a phenomenalistic mode of awareness. Bahya, who is an advanced con­templative, receives only a very brief instruction, to the effect that his sense of ego will disappear as soon as he succeeds in correctly taking account of just his naked sensory impressions and thoughts and no more than that: In reference to the seen, there will be only the seen. In reference to the heard, only the heard. In reference to the sensed, only the sensed. In reference to the cognized, only the cognized. That is how you should train yourself. When for you there will be only the seen in reference to the seen, only the heard in reference to the heard, only the sensed in reference to the sensed, only the cognized in reference to the cognized, then, Bahiya, there is no you in terms of that...

Read as an example of “inversion of gnosis,” this sutra is a classic demon­stration of the anatman perspective in operation. Bahiya is being instructed to bracket his subjectivity. Whereas for Husserl, phenomenological re­duction involves only the suspension of empirical subjectivity, the epoche of anatman is more far-reaching in that it entails a suspension of even the transcendental subjectivity. The changing acts of knowing are purely empirical and do not refer back to an unchanging transcendental knower. Knowing exhausts itself in the known. All traces of subjectivity are dissolved by the projection of consciousness into experience in its momentariness and concreteness. The transcendental subjectivity’s construction of superim­posed noemata is brought to a halt and the thusness of the experiential flux revealed.

Mach arrives at a somewhat similar view of things in Analyse der Empfind- ungen, albeit not by a contemplative route but by a speculative thought experiment. After its semantic scope is successively expanded until it en­compasses the totality of phenomena, he imagines that the ego is finally dissolved. It is recognized to be a provisional concept, having at best a merely pragmatic utility—rather than a veridical representation of some metaphysical absolute underlying or overlooking the functional fabric of sen­sations that is revealed in the final analysis to be truly fundamental—colors, sounds, warmths, pressures, spaces, times, etc.:

The contradistinction of ego and world, between sensation or appearance and thing then drops away, and now it is merely a matter of the interrelationship of the elements a j3 7 .... AB C ... K L M ...., of which this contradistinction had been only a partially applicable and incomplete expression. As does the Buddhist theory of the elements of experience {dharmas), Mach advocates a form of sensation pluralism. The Rohitassa Sutta of the Nikayas shows evidence that early Buddhism rejected cosmological world-concepts in favor of a phenomenalistic micro­cosm or body-world constituted by the totality of sense-impression at any given moment:

Yet it is just within this fathom-long body, with its perception and intellect, that I declare that there is the cosmos, the origina­tion of the cosmos, the cessation of the cosmos, and the path of practice leading to the cessation of the cosmos.

The sutric doctrine of the microcosmos of the “fathom-long body” is again strikingly paralleled by the Machian notion of the Korperwelt—a clever double-entendre capturing how the world appears from the perspective of common-sense realism—as a “world of bodies,” i.e. the totality of physical objects—as well as how it appears from the less intuitive but epistemologi­cally more sophisticated point of view of phenomenalism—as a “body-world,” i.e. the logical space of sensibilia comprised by the human body. In keeping with Buddhism’s soteriological pattern of inverse gnosis, the macrocosm is not only reduced to the mind-body microcosm, but the latter comes to be contemplated in its empirical aspect of functional flux and dispersive difference, as aggregates or bundles {skandhas) lacking any kind of real inner unity.

This phenomenalizing reduction of the macrocosm to the microcosm is an important prototype in early Buddhism of the later Mind-Only program of subjective idealism. Vasubandhu’s refutation of Abhidharmic realism in the Vimsatika arguably represents an attempt to urge a return to this earlier ideal type of non-atomic phenomenalism—a revival with a twist. In verse 16 of the Vimsatika, we rediscover the Bahiya Sutta’s grand theme (in the seeing, only the seen) iterated in an idealistic inversion—in the percept, only the perception (pratyaksaY

The awareness of perception is as in dreams. When it exists, the object of perception does not exist. Nominalism Buddhist doctrine has a characteristic tendency to analyze perception and knowledge in terms of particulars, and does not accept the existence or instantiation of abstract entities such real wholes or universals, as do other darsanas (e.g. Nyaya). In the context of the theory of perception, Matilal describes nominalism as the view that we are directly aware of sensible quality-particulars. These qualities do not exist apart from their instantiations, and their instantiations do not exist independently of our sensation of them.

The development of nominalist views about perception is closely connected with the articulation of an ontological nominalism in the context of the Bud­dhist “bundle theory” of the body-world and its associated “no-ownership” stance on personal identity. In the Milindapanha, the monk Nagasena’s famous dialogue with King Menander, Nagasena maintains that under anal­ysis, he does not exist as a person, and that his name “Nagasena” is therefore a strictly nominal designation for the psycho-physical bundles, just as the conceptual label “chariot” is is applied to a phenomenon that does not with­stand a decompositional analysis. Since there is no more to the chariot than its parts, and none of the chariot’s parts are the chariot, Nagasena conveys, in reality there is no such thing as a chariot.

Siderits submits that Nagasena’s argument is proposing a kind of mereo- logical reductionism. Examining a number of different readings of the chariot dialogue, Oetke makes the case that the argument is semantically underdetermined. He believes that it is far from clear precisely what philo­sophical argument Nagasena is trying to make. However, he for some reason misses the parallels drawn by Kapstein to instances of the chariot simile in the Katha Upanisad^ as well as in Plato’s Phaedrus. In this light, I conjecture that purely deconstructive intent offers a plausible reason for why Nagasena’s argument seems underdetermined. Instead of advancing any constructive thesis of his own, he is merely attempting to subvert a classic Upanisadic example of atman theory, i.e. to demonstrate that in actual fact, the example shows the opposite of what its proponents intend it to illustrate. Kapstein, however, argues that Nagasena’s chariot argument is proposing a logical constructivist view of personal identity, and contends that it represents a possible precursor to Vasubandhu’s argument against atomism.

The logico-epistemic tradition of Yogacara worked out a more technical elaboration of this nominalism in the so-called apoha (exclusion) theory presented by Dharmakirti in the Pramanavarttika. Here, universals are interpreted as absences or eliminations—e.g., the class of trees is defined as the mere exclusion of non-trees, which is another way of saying that trees are merely whatever is labeled as a tree, without committing to ontological proliferations such as an underlying universal of “tree-hood,” an extensional set of all trees, and so forth. In articulating the relation of abstract concept to concrete particular along strictly linguistic lines, i.e. one of naming or conceptual imputation, the Buddhist apoha theory avoids trading in metaphysicalites such as instantiation, exemplification or inherence. These examples are only to illustrate that ontological nominalism, or the principle that existence, reality and actuality are in the proper sense attributed only to particulars (if at all), is a deeply rooted aspect of the Buddhist philosophical pattern.

As we will later see in more detail, the “sense-data mosaic” theory of experience is one of severals ways that the Buddhist phenomenalism has been integrated with nominalism. The Rise and Fall of Sense-Datum Atomism A realistic sense-data nomi­nalism allows for two possible accounts of the nature of physical objects—a regressive or a progressive theory. The regressive theory (that the physical object gives rise to sense-data) leads to representationalism, the progressive theory (that the physical object is a construction from sense-data) leads to phenomenalism, or presentationalism. The aim of Vasubandhu’s argu­ment against atomism in 11-16 will be to demonstrate that neither of these accounts work in a realist mode.

To fill in some philosophical substructure for Vasubandhu’s delivery of this coup de grace to metaphysical realism, we will briefly survey Indian atomism, reviewing the Vaisesika doctrine and theorizing about the possible rationale for the development of Buddhist atomism. Following the path of reasoned eclecticism, we shall then construct two Weberian “ideal types” or models of Abhidharma synthesized from four streams of commentary—(1) mod­ern Western scholarship of Indian Abhidharma (mainly STCHERBATSKY, FRAUWALLNER and VON ROSPATT), (2) Gelug doxography and contempo­rary commentary (HOPKINS, BUESCHER), (3) other modern philosophical reconstructions of Abhidharma (notably MATILAL, CHATTERJEE, GRUPP and GUENTHER), and (4) historical review of atomism in Western philosophy (Pyle).

The main focus of the discussion will be to investigate the Abhidharmic evolution of the concept of substance in the light of its critiques of time and perception. We will illuminate how this critical dynamic forced an initial shift from a progressive to a regressive theory of mind-independent physical objects, setting the stage for Vasubandhu’s abandonment of even the regressive theory.

The philosophical ground covered by this evolution is roughly equivalent to the trajectory leading from Hume through Kant to Mach. On an ideo- historical time scale, this is no mean achievement. But the argument we are making here is stronger. We are setting out to retrace this trajectory as it is reflected in the philosophical voyage of a single thinker—in the 4th century CE.

2.3 The Origins of Buddhist Atomism

We will have to leave the questions of where the doctrine of atomism first originated—in ancient India or in Greece—and whether there was any ex­change of ideas between these cradles of philosophy for others to answer. Democritus, the first known Greek atomist, can be dated to approximately 460 BC, whereas Kanada, the probable originator of atomism in India, lived possibly a century or so earlier. However, for the purposes of our investiga­tion, it will suffice to point out that ancient India was home to numerous atomist doctrines. The theory of atomism (paramanuvada) emerged in the astika schools of Indian philosophy (e.g. Vaisesika) as well as in the nastika schools (e.g. Jainism). Vaisesika Atomism Kanada’s atomism, set forth in his Vaisesika Sutra, was one of the most influential doctrines. Pace Kanada, every physical object is constituted by a finite number of paramanus—immutable, eternal, self­subsistent, indestructible, and individually distinct atoms. These atoms have a spherical shape and touch each other. Vaisesika proposes a theory of atomic permanents, like the atomism of the Jains, who also believed that eternal and immutable atoms flowed through events.

The Vaisesika doctrine of permanents somewhat resembles the view of Democritus, who submits a reductionist material constitution theory. Dem­ocritus believes that the objectively existing compound bodies to be aggre­gates of atoms, which are eternal, true units, and characterized by solely by primary qualities. Atoms and the empty space they are located exist eteei (in reality), whereas colors and other sensible qualities exist only nomoi (by convention).

The early Vaisesika atomism of Kanada later came to be elaborated by Vatsyayana, a contemporary of Vasubandhu, who speculated that atoms are absolutely imperceptible, i.e. infinitesimal. Composite entities are integral wholes in their own right, inhering in the conglomerate (pinda) of their atoms but having different qualities from them. The discrete atoms serve as the material base of constitution in which the real integral wholes can inhere. The perception of the integral whole (avayavirf) takes place simultaneously with the perception of certain of its parts, e.g. even though one sees a tree in aspect, one can be said to be seeing the whole tree. In Vatsyayana’s mereology, atoms are never found in isolation. Two atoms (or least parts) cohere into a dyad-atom {dvyanuka') or least whole of atomic size, and all further combinations are formed from these dyad-atoms.

This strange construction is motivated by the difficulty that a least part is spatially partless and therefore unable to be spatially adjoined to other least parts. The dyad-atom is held to solve this problem because it has parts and can connect to other dyad-atoms in the usual fashion.

Buddhist Atomisms The Buddhist atomistic doctrines are considered by some to be among the most difficult atomist theories, their many subtleties and contradictions being a source of much polemical debate within Indian philosophy. I will not be able to fully explore all of the aspects and issues related to the various differentiations of atomism among the many Buddhist sects. Within the compass of this thesis, I will therefore restrict myself to a survey of the most basic doctrines and developments.

In the fold of Indian Buddhism, atomist doctrines can be found in the Sarvastivada system of tenets, as well as in the early Sautrantika (in Gelug doxography, the school of “Sautrantika Following Scripture”) and late Sau­trantika (“Sautrantika Following Reasoning”) systems. All three atomist systems of Buddhist thought each in different ways hypothesize that the familiar physical objects of the external world are somehow constituted by ultimate simples. They maintain that all ultimate truths can be classi­fied either as partless simples, as absences, or as cessations, and that all conventional truths reduce to their ultimate constituents.

A distinguishing feature of the Buddhist schools versus non-Buddhist doctrines of atomism can be found in their rejection of the realists’ inherence (samavaya) theory of wholes. In fact, the Buddhist positions reject all theories of constitution with the exception of mereological summation.

In modern analytic philosophy, the strongly reductionist stance that there exist only partless entities is commonly referred to as mereological nihilism. More precisely, the Vaibhasika and early Sautrantika atomisms can be char­acterized as varieties of impure physicalistic mereological nihilism (atoms are physical and interact with each other), whereas the late Sautrantika position can be classified as a pure mereological nihilism.

The position of mereological nihilism maintains that only elementary building blocks without proper parts have real existence—all seemingly discrete and integral wholes such as macroscopic objects in fact exist only as nominally designated aggregates or conglomerates, i.e. no differently than scattered mereological sums of atoms. Interestingly, there is not much controversy among commentators in that the epistemologies of all three types of Abhidharma scholasticism are con­sidered to be different inflections of the same underlying hybrid theme of pluralistic atomism and reductionist realism.

Why and how did Buddhism initially articulate the doctrine of atomism? As we have already pointed out, the earliest strata of Buddhist writings exhibit no evidence of such doctrines. Therefore, the atomist doctrine was almost certainly a later addition and not an actual historical teaching of Gautama Sakyamuni (though Mahavira, the contemporaneous founder of Jainism, almost certainly did teach atomist doctrines).

There are three possible lines of explanation for the development of Buddhist atomism—(1) a physical atomism that was borrowed from non­Buddhist darsana, (2) a phenomenalist origin in minimal sensory magni­tudes, and (3) a nominalist or mereological origin. According to the phe­nomenalist line of explanation (2), which we construe as an epistemological hypothesis about limit-constructs of sensation, the atoms are the sensations at the absolute threshold of vision, hearing, touch, etc., i.e. the “pixels” of a sensory field. While the 19th century psychophysics of Fechner was the first attempt to scientifically measure such thresholds and quantitatively determine functional relationships between stimuli and sensations, it is not unreasonable to assume that the pre-scientific epistemologists of early Abhidharma were well aware of the existence of sensory limits. The nomi­nalist line of explanation (3) is best construed as a mereological hypothesis about limit-constructs of composition where the Buddhist whole-part nomi­nalism (the whole is unreal, its parts are real) comes to be radicalized into a mereological nihilism (the whole is unreal and only its self-subsistent least parts are real).

2.4 The Vaibhasika School

In its living details, the Vaibhasika doctrine of atomism has a much richer and messier philosophical texture than the abstract type we are presenting here. The Vaibhasika stance on perceptual experience may be described as a form of physicalistic sense-datum positivism. They believed that it was possible to separate the objective constituents of perception from the epistemic object and developed a theory of atomic substance (dravya'). Their theory of atoms (paramanuvada) proposes a direct realism operating over an ontology that hybridizes all three types of atomism—physical, phenomenalist, nominalist and physical.

The Vaibhasika account of perception is strongly presentationalist in flavor. Extramental objects present themselves directly to an observing cons­ciousness-moment, without intervening representational synthesis. In other words, when observing a macroscopic object, one actually directly observes its atomic constituents in aggregate. Vaibhasika believes that these percepts are as real as the perception itself, and that the external world is directly grasped in non-conceptual perception. The atomist doctrine of the Vaibhasika sect passed through three major phases of ideo-historical development.

The mature doctrine of atomism as presented in Vasubandhu’s Kosa may be traced back to the Mahavibasa, an encyclopaedic compendium of the Abhidharma of the Sarvastivada schools, dating to the 1st or 2nd century CE. However, the first recorded mentions of a Buddhist doctrine of atomism must be credited to Dharmasn, the author of the Abhidharmasara. This massive synthesis of extant Abhidharmic doctrine was the first commentarial work to forge the Sarvastivada doctrines into a system of scholastic thought. The Abhidharmasara was to remain the paradigmatic model for such compendia even up until the era of Vasubandhu and Sanghabhadra. It has been shown that the systematic presentation and structure of the Abhidharmakosa is derived in large part from the Abhidharmasara.

Vasubandhu’s Kosa and Sanghabhadra’s contemporaneous but consider­ably more voluminous Nyayanusara (a polemic rejoinder to the Kosa in 120,000 verses) form the apex of the classical Abhidharma of the Sravaka- yana, or non-Mahayana schools of Buddhist thought. Further developments of Abhidharma were to take place mostly in the context of Mahayana system­atizations, e.g. the Abhidharmasamuccaya of Vasubandhu’s half-brother, Asariga the Yogacarin.

2.4.1 Time, Space and Atom in Vaibhasika

Although some commentators characterize Vaibhasika as having a theory of atomic permanents, this is somewhat too simplistic. Anacker notes that for Dharmasri, atoms are momentary, though they may form “series,” or causal chains of momentary events. Time is also atomized into a series of discrete instants, but the duration of these moments is so brief that the illusion of continuity is generated, as is the case with the frames of a motion picture following in rapid succession. The atomizing approach has the greatly attractive feature of not having to account for the genidentity of macroscopic physical objects, as under analysis, the appearance of a physical object persisting in time simply explodes out into the causal chain of momentary instants of its atomic decomposition.

The Vaibhasika are known as Sarvastivadins due to their canonical tenet that “everything exists” (sarvam asti), meaning the existence of entities in past, present and future. The epistemological rationale for this principle is the wish to defend a form of externalism—because knowledge is always knowledge of the real, the intentional objects of epistemically valid cognitions must somehow exist.

The Kosa offers up a number of other arguments in favor of the sarvam asti thesis, including causal and moral rationales such as that past entities must exist in order for present effects to be brought about, and that future entities must exist or else actions would not have consequences.

Time As the Sarvastivada Abhidharma continued to evolve, it became increasingly aware of the problem of time. The concept of an immutable entity migrating through the stages of time was no longer acceptable, as eternal substances were held to be contrary to the fundamental Buddhist doctrine of transience (anityata). Due to this doctrinal concern, the Vai­bhasika Abhidharmikas were under pressure to evolve more sophisticated accounts of substance and time.

All elements of experience (.dharmas), including atoms, persist in the three periods of past, present and future. In modern terms, the sarvam asti principle may be described as a “block universe” account of space-time, as it affords equal ontological status to all moment-entities regardless of their time. The Vaibhasika advanced four major accounts of temporality, credited to the encyclopaedists of the Mahavibasa who originated them: Dharmatrata, Ghosaka, Vasumitra, and Buddhadeva.

The Kosa attributes to Dharmatrata the time theory of “change of ex­istence” (bhava-parinama'), turning on a substance-existence account of change. The existence (bhava) of an element is an accident that can arise and cease, whereas the substance (dravya) is eternal. This view is immedi­ately dismissed as a guise of the previously refuted non-Buddhist Samkhya system. Ghosaka is attributed with the view of “change of aspect” (laksana-pari- nama), which asserts that entities have three temporal aspects or qualities (laksanas)—pastness, presentness and futurity. A future entity is merely an entity that is appearing under its future aspect and hiding its past and present aspects. The Kosa rejects this theory, reasoning that it implies a contradictory compresence of past, present and future in the same entity.

Frauwallner observes that the last two theories, the views of Vasumitra and Buddhadeva, account for time exclusively in terms of external connection, most likely in an attempt to avoid the difficulties inherent in Dharmatrata’s and Ghosaka’s substance-oriented theories of time.

Buddhadeva proposes a relational theory of time, where temporality is accounted for as a change of contingency (apeksa-parinama'). Change is governed by a tenseless binary relation between entites (dharmas), such as “r is earlier than y,” that imposes a partial order upon them. An examination of the Kosa’s refutation of the theory of Buddhadeva reveals it to be identical to the classic McTaggart argument against the reality of time. If events are ordered by a two-term relation into what amounts to a McTaggart B-series, then each event comes to have all of the McTaggart A- determinant properties—the present moment is also future because it is later than another moment, and past because it is earlier than a later moment. The Vaibhasika deployment of McTaggart argument against temporal order drops a strong hint about the developmental trajectory of Abhidharmika views on the reality of time.

The theory of Vasumitra is described as “change of position” (avastha-pari- nama). The Kosa’s explanation and critique of the four Vaibhasika theories of time concludes that the theory of Vasumitra is the correct view.

Frauwallner conjectures that the positional theory is a conglomerate of two contradictory theories of temporality—each proposed by a scholar named Vasumitra—which were conflated in later doxography. He points out that the abacus analogy used to show the intent of the theory logically excludes the ideas of efficiency and change.But the assumption that Indian Buddhist doxographers conflated two different Vasumitras is scarcely credible, especially not when it comes to such a central theme in the Vaibhasika philosophy of time. It is more reasonable to hypothesize that the conglomeration of seemingly contradictory views was a synthesis attempted by a single scholar.

The general scheme of Vasumitra’s synthesis as I conceive it is that of strongly reductionist reformulation of the concept of time in terms of change and causality. Temporality itself is effective action in the infinitive, and the tenses past, present and future are the modal positions of this “causal arrow.” An entity is real iff it is effective, which holds for any of the following three cases: its efficiency is actually being discharged (present), if it is possible for its efficiency to be discharged (future), or if it is not possible for its efficiency to be discharged (past). The temporal position of a future or past entity in other words depends on whether it is possible or impossible for the entity to discharge an effect. Moreover, the position is considered to be external to the entity, rather than a property or accident of it—in other words, it is the entity’s modal context. Past and future entities are held to exist de re.

Thus, Vasumitra’s “temporal eliminativism” refines the philosophically naive eternalist account of time taken over from the Vaisesika atomism into a considerably more sophisticated modal account which explains how entities can meaningfully be said to exist in past, present and future without resorting to metaphysicalities such as absolute time and eternal existence. Vasumitra’s account is still straightfowardly realist, but it does away with the dimension of time.

A present entity is a real actuality, past and future entities are possible and impossible real actualities, respectively. The concept of substance is identified with function, and the coming into appearance of an element is explained as the presence of the necessary conditions for its function. In this, we see the idea being reflected that the criterion of reality is efficiency (arthakriyakaritva)—a basic Abhidharmic theme which is agreed upon all the way through to Cittamatra and beyond. The debate on this issue can now be crystallized around the Vaibhasika and Sautrantika defense of an externalist version of this principle.

So far from being a mistaken and inconsistent conflation of separate theories, as Frauwallner uncharitably claims, Vasumitra’s theory should be considered a novel and significant development in the Abhidharmic critique of time.

Nonetheless, there remains a major philosophical tension in the theory of Vasumitra. Why is it that we only ever experience the present, if the past and future are essentially equally real? One way that Sautrantika attempts to resolve this tension is to retain Vasumitra’s modal account of time but abandon modal realism about the past and future, thus arriving at its characteristic doctrine of momentariness (ksanikavada), about which we will later have more to say.

Space The Vaibhasika system asserts that atoms are partless, that there are intervals of empty space (interstices) between atoms, and that the sub- stance-atoms do not touch when aggregated. This is also asserted by some Sautrantikas, e.g. by Samgharaksita. However, the Sautrantika critic Yasomitra maintains that in the Vai­bhasika view, substance-atoms are without the quality of impenetrability or resistance (pratighata), which entails the quality of covering or spa­tial extension (avarna-laksana).5e Obstruction is an emergent quality of the aggregate-atom (samghata-paramanu), as is spatial extension. The samghata-paramanu is argued by the Vaibhasika to be both extended and partless.

The Vaibhasika substance-atom resembles the view of Hume, in that Hume also takes perception to be ultimately constituted of simples or minima sensibilia: “A single visual atom, in itself, has no extension, but combinations of atoms do. The perception of an object composed of two or more atoms is complex, since it is a perception with more than one spatial part.” According to Ayer, Hume contends that the impression of any physical extension must be compounded of a finite number of mathematically point­like sense impressions, which are concrete objects. Because these objects are partless, they are the smallest conceivable sense-data.

Vaibhasika scholasticism defines space (akasa) in purely negative terms as absence of obstructive contact and enumerates it as one of the three uncompounded phenomena. But unlike the other Buddhist tenet systems, Vaibhasika considers space to be existent (sat), a partless, permanent substantial entity, and infinitely divisible. Clearly, the Vaibhasika system proposes the existence of absolute space, though it still lacks an explicit notion of what one would call in the term­inology of modern mathematics an intrinsic geometry or metric, a difficulty which we will return to later on in Chapter 3.

Atom A substance-atom (dravya-paramanu) is an irreducible sensation­unit of a certain sensory type, i.e. it is visual, tactile, auditory, olfactory, etc. These sensibilia are metaphysically real qualities and are objectively present in the world. In this respect, the Vaibhasika atomism can be characterized as phenomenalistic, i.e., derived from the notion of minimal perceivable magnitudes. But the idea of atoms being located in absolute space necessi­tates that they be somehow smaller than physiological minimal perceivable magnitudes.

The visual features of an object, e.g. a tree, should otherwise be indepen­dent of the viewer’s distance to it, which is clearly not the case—the closer we move to the tree, the more visual detail we are able to resolve. But why are the Vaibhasikas willing to sacrifice the purity of a phenomenalistic atomism of sense-minima in order to preserve a notion of absolute space? Perhaps they feel that they are making a reasonable concession to realism. To ensure that “the things we really see when we say we see a tree” are not solipsistic fictions, they need to have an objective existence—and this requires a tran­scendentally objective space (akasa) to serve as an observer-independent frame of reference.

Matilal points out an instructive comparison of Vaibhasika to the view of Ayer, who “recommends that we conceive of perceptible external objects as being literally composed of the ultimate particles of physical theory, these being imperceptible, not in principle, but only empirically, as a consequence of their being so minute.” According to the Vaibhasika, a dravya-paramanu never occurs in isola­tion. It is always a part of an aggregate-atom, or basic cluster of atomic sensory properties (samghata-paramanu). These aggregate-atoms are only perceptible in collection, although they each appear individually to the sense­consciousness.

The Sarvastivadin neo-orthodoxy of Sanghabhadra, as critiqued in the Kosa(A and defended in the Nyayanusara, specifies that each aggregate­atom consists of a total of at least sixteen substance-atoms—each of the four sensory atoms of color, odor, taste and tangibility being accompanied by its own set of four general elemental atoms. These four fundamental elements (mahabhutas) which, like all Buddhist elements, are more like force than substance, function as the intensive magnitudes of their associated aggregate-atom. The earth element repre­sents the qualities of solidity, hardness and repulsion; the water element represents cohesion and attraction; fire corresponds to temperature and air to motion.

Every aggregate-atom contains one atom each of these four. Earth, water, fire and air are always present in equal proportion—the difference lies only in their perceptual intensity. A tiny point of light can shine dimly or brightly, steel needles are more intensely felt than the touch of a brush, etc. Another variant of the Vaibhasika atomic theory is offered by the scholar Dipakara. According to this theory, the aggregate-atom is formed of a minimum of seven substance-atoms. We may hypothesize that one atom is located in the center of the aggregate’s cubic spatial “receptacle” and the six other atoms are positioned at the cube’s vertices. The minimal aggregate­atom comprises one atom each of the fundamental elements (which together constitute the aggregate-atom’s tactile properties), and one substance-atom each of color, odor, and taste.

The most important points to keep in mind about this bewildering multi­plicity of Vaibhasika atomic theories are that the atoms are held to have spatial extent and to be separated by interstices. Also, the shape (samsthana) of a visual atom is held to be an ultimate and can be perceived independently of color.

The Vaibhasika system of tenets is unmistakeably phenomenalistic—all that exist are sensory qualities and their intensities. An eternal substance as is proposed by the Samkhya is denied by the Vaibhasika (with the exception of Dharmatrata). Moreover, Vaibhasika presents a full-blown reductive account of sensation, where the likeness of a macroscopic physical object is simply identified with the spatial arrangement of a vast number of ele­mentary sensation-particulars or microscopic tropes. In other words, this assembly of spatiotemporally determinate, mind-independent micro-tropes is what we actually have commerce with when we think we are perceiving a macroscopic physical object. The world is built up out of the epistem- ically and ontologically ultimate bits and pieces given in immediate sensory experience.

Therefore, although the Vaibhasika are considered by Buddhist doxog- raphy to be a substantialist school, the somewhat paradoxical situation obtains that the Vaibhasika substances are equivalent to the sense-data themselves. Substance is completely identical to trope—a sense-datum requires no further substrate. In this particular respect, the Vaibhasika mosaic theory is closer to the hypothetical “original flavor” of early Buddhist phenomenalism than its critical refinement, the Sautrantika, which moves away from the phenomenalizing idea of the simple as a sensory ultimate in favor of a progressive account of the constitution of physical objects.

Though in many ways a compromise with physical realism, the Vaibhasika system is by deliberate design far too reductionistic to qualify as a truly “naive” physical realism such as Nyaya-Vaisesika, which asserts the real existence of wholes, universals and actions.

The Vaibhasika propose a quintessentially Buddhist solution to the qualia problem. Their atomizing, objectivizing phenomenalist reductionism simply denies that sensations have any intrinsic subjectivity in the ultimate view. To Vaibhasika, “raw feels” such as minute flashes of color, tactile pin-pricks etc, are wholly public, mind-independent and objective. Vaibhasika argues the notion that sensation-atoms are somehow private to the perceiving consciousness to be an egological error.

2.4.2 Critical Review of Vaibhasika Presentationalism

In the debate on time in the fifth chapter of Vasubandhu’s auto-commentary to the Kosa, the Sautrantika takes the Vaibhasika to task over the sarvam asti thesis and the notion of karitra. Among other criticisms, the Sautrantika interlocutor denies the intelligibility of the Vaibhasika’s claim that past and future entities exist in the same sense as present entities. That Vasumitra’s theory should pose many difficulties in this regard is unsurprising, as in the terminology of modern logic, assertion of the existence of past and future entities would entail quantification into modal context. The Sautrantika also argues against the Vaibhasika ideas that future existents come into appearance by entering the scope of the present, and that the time of an entity is something separate from the entity itself. In analogy to A. N. Prior’s ideas about the logic of time, the Sautrantika would simply eliminate the metaphysical distinctions between the temporal status of an entity, its function, and its intrinsic nature.

Later on in the debate, the Sautrantika presses the Vaibhasika to concede that the intentional object (visaya) of a cognition can be a Meinongian absistent (Gegebenheit)'. “Thus it is that both existence and non-existence may be objects of cognition.” If cognitions about irrealia are not admitted, he reasons, many absurdities result—such that if one desires not to hear an unpronounced word, one would be compelled to actually pronounce the word in order to satisfy externalism about intentionality. However, if the Vai­bhasika yields to some version of Brentano’s thesis of intentional inexistence, then it becomes much more difficult for him to resist the Sautrantika’s fictionalism about past and future entities. As against existence in past, present and future, the Sautrantika is advancing a version of Augustine’s view that “the present of things past is in memory; the present of things present is in intuition; the present of things future is in expectation.”

There are problems in reconciling the Vaibhasika of objective sense-data with phenomenalism—it is difficult to conceive of a reasonable account of the existence of unsensed sensibilia that does not make liberal use of counterfactuals. If we lock a blue ball into a lightproof black box, surely it is ridiculous to say that the ball is blue even inside the box where there is no light and nobody can see it—what we mean is something of the sort that, given the same perceptual conditions inside the box as outside, if we were able to see the ball, it would be blue.

Ayer cogently observes that there are no empirical means of settling the question as to whether or not sense-data can exist while they are not being apprehended. Another Ayerian objection to the Vaibhasika position would be that that a phenomenalist who speaks of sensibilia as having a distinct existence in space and time has not fully carried out his program—his so- called sensibilia are merely physical objects in disguise.Pyle raises the concern that a theory of perception predicated upon a mereo­logical nihilism about compound bodies cannot be realist in the strict sense— if the existence of compounds is denied, the explanandum of the atomic theory is lost and the atomic theory of perception must be construed as an “error—theory” about why we hold mistaken views when we see macroscopic objects. However, this objection is not really a problem for the Vaibhasika since he intends to propose an error-theory of perception.

Yasomitra’s critique of the Vaibhasika atom concept in his commentary of the Kosa is an anticipation of Vasubandhu’s critique in the Vimsatika-. “In the Abhidharmakosa and in Yasomitra’s commentary, the Sautrantikas— who refused to accept that conglomerate atoms were without parts—then pressed the Vaibhasikas. They asked them how it could be that if none of the substance atoms were extended in space and if none of them offered any resistance, how could many together do this? How could one put together many atoms that took up no space and obtain something with any extension at all?”

Although atoms are said to be directly observable, they are not evident to ordinary perception, like houses, pots and trees. The atom therefore remains a metaphysical construction, and if it is admitted that the existence of atoms is only known by inference (anumana) to the best explanation, then who is to say that there cannot be some other, better explanation for phenomenal appearances? This Mind-Only critique will form a part of Vasubandhu’s argument against the Vaibhasika position later on in Chapter 3. For the idea that atoms are observables only collectively but not individually comes very close to a causal theory of perception.

As Matilal puts the Mind-Only argument—“If the so-called atoms gen­erate perceptions out of their own power, why are they not also grasped in such perceptions?” But as ultra-nominalists, the Vaibhasika have no other choice but to defend a doctrine of homeomeria.™ Vaibhasika encounters the difficulty that it wishes to maintain on the one hand that the atomic sensa­tions are sensed directly, yet on the other hand claims that they are sensed only in aggregate. The Vaibhasikas assert that an isolated minute particle is not an object of a sense-consciousness, nor is it the composite as a whole, as composites are merely imputed collection-generalities (gana-samanya)

In this critique, which will be the critique of Vasubandhu, we find a close affinity to Mach’s rejection of atomism due to his skepticism about im­perceptible entities: “All his life, Mach was never able to make friends with the atomistic corpuscle theory. In discussions, whenever anybody presumed the existence of atoms as a matter of course—and we may assume that this happened quite frequently—he irascibly interjected: ‘Have you ever seen one?’ (‘Hams ans g’sehn?’)”

The purpose of physics, according to Mach, is not to determine the funda­mental ontological constituents of reality, but rather, to explore and describe the functional relationships between the elements given to experience. To Mach, it would be an unforgivable case of metaphysical confusion to conflate sensations (the “given” or sensory core of experience) with atoms (ontological ultimates).

The Vaibhasika account of shape is deeply counterintuitive. Why is it that individual atoms come to be attributed with a substantial shape which is however imperceptible, whereas the perceptible shape of macroscopic objects is held to be merely conceptually imputed? Wouldn’t it be more reasonable to argue that perceptible shapes are real and imperceptible shapes are imputed?

Finally, realistic presentationalism is uncomfortably vulnerable to argu­ment from illusion—how to say, for example, that a stick partly submerged in a glass of water only appears to be bent?

A radical phenomenalist of the Machian stripe argues that there is really no such thing as illusion, or rather, that it is not possible to draw a meta­physical distinction between illusion and veridical perception. For him, the entire issue of illusion is a pseudo-problem of the metaphysical realist— whenever we speak of illusion, we can be sure that something metaphysical is cooking. Similarly, wherever the epistemic reductionist may go, the hob­goblins of emergent properties and behaviors follow not far behind. “Illusion” is unreal—or so we believe. It is an explanation for what is happening when the synthetic activity of sensory perception generates some epistemically ungrounded percept. Except for deceiving us, the illusion is causally inert. “Emergence,” however, is quite real—or so we suspect. It is our account for what is happening when dynamic interaction of a physical system generates some ontologically irreducible property. Here, a radical phenomenalist not committed to realism will argue that illusion and emergence are philosophical enantiomorphs. As soon as we draw a meta­physical divide between the epistemic subject and the object, the problem of illusion rears its ugly head on the side of the subjective constitution of experience and the problem of emergence arises on the side of the objective constitution.

The Vaibhasika will insist that it is all a matter of finding the “predeter­mined breaking point” between the subject and the object. He will argue that there cannot be such a thing as an illusion of a tree, because “tree” is merely a linguistic and conceptual convention imputed to some patches of color that we see. But how to account for the fact that these colored blobs are spatiotemporally correlated with certain tactile sensations of roughness when we attempt to touch the trunk of the tree? The things that we are seeing and the things that we are touching, the Vaibhasika explains, are the same things—our sensory awareness is making direct contact with the visual and tactile sensation-atoms are objectively located in space and exist independently of our awareness. When we close our eyes, where do the col­ored blobs go? They don’t go anywhere, the Vaibhasika avers—they are still out there in space, but our sensory awareness simply loses contact with them. But this account is badly vulnerable to argument from hallucination—where do the pink rats go when one closes one’s eyes?

To preserve external world realism, the Vaibhasika must concede the possibility of illusion and attempt to explain it as an error of perceptual synthesis. Without perceptual synthesis, it is difficult to account even for cases of “veridical” perception. For example, the problem of inferring a visual object’s shape from its two-dimensional projection does not admit to a unique solution. This intractable problem is converted into a solvable one by adding hard-wired assumptions about how the world is usually put together. Sensory illusions arise in cases where these a priori assumptions are violated.

But perceptual synthesis is exactly what the Vaibhasika direct realist account of perception wishes to deny—it argues that in non-conceptual perception, we see things as they actually are, and that we are only making a conceptual imputation (prajhapti) when we think we see a gross physical object.

2.5 The Sautrantika School

An important feature of the philosophical program of the Sautrantika school as we construe it here is to attempt to reconcile the basic principle of empiri­cism (that knowledge arises from experience and is limited to the realm of experience) with the discovery of the synthetic character of empirical knowl­edge, i.e. the realization that its constitution is dependent upon categories and assumptions not present in immediate experience.

In this regard, its representationalist epistemology closely accords with the familiar Kantian program of critical realism, which posits that it is necessary to make a priori posits that are epistemically prior to the content of experience; but that such metaphysics is only possible if it is strictly limited to the deduction of the conditions of possible experience.

Of all the schools of Indian philosophy, the Sautrantika was the only one to take such a Kantian turn. The reason for this unique ideo-historical development can only be that Sautrantika was attempting to address the same philosophical challenge as Kant—the need to critically reformulate empiricist realism in such a way as to avoid surrendering realism about a mind-independent world.

Chatterjee observes that Sautrantika is not so much a new philosophy as the critical analysis of the implications of the Sarvastivada realism. “The Sau­trantika must be understood as Sarvastivada itself, aware of its own logical basis. They are not two schools, but two phases of the same metaphysical pattern.”

Though Sautrantika must also be considered an Abhidharmic school, its critical turn of the Abhidharmic pattern forced some important doctrinal points of difference to the Vaibhasika concerning the understanding of Abhi- dharma. The dogmatic presentation of the original seven texts of Abhid- harma dating approximately to the 2nd century BCE as the revealed word of the Buddha needed to be called into question. The Sautrantika strategy for resolving this dilemma was to argue that while Abhidharma existed, it was in fact diffused throughout the sutras. The Sautrantikas held that the independent set of seven core scriptures known as the Abhidharma could not lay claim to being categorically valid word of the Buddha (buddhavacana) taking precedence over interpretations conforming to the sutras in case of any conflict between the two.

Having liberated themselves in this way from the straitjacket of dogmatic scholasticism, the Sautrantika epistemologists were able to articulate two important doctrinal innovations over the system of the Vaibhasikas—the theory of presentism and the causal theory of perception.

Della Santina remarks that the Sautrantika emphasis of of the role of conceptualization, or discrimination (vikalpa) is connected to their critique of the Vaibhasika category system. Synthesis now begins to fill in for many factors of experience that were previously claimed to be objective realities: “They rejected the independent, objective reality of many of the factors the Vaibhashikas accepted, ascribing these dharmas to the functioning of discrimination or imagination. This goes some way toward the standpoint of the Mind-Only school, which eventually denied the objective reality of all objects and affirmed the sole reality of mind.”

Like Kant, Sautrantika has become aware of the synthetic activity of the mind—however the vikalpa is construed as a synthesis of the unreal rather than the real, in keeping with the Buddhist principle of nominalism.

Generally speaking, the division of the two truths according to the two means of valid knowledge or epistemic grounding (pramanas) recognized by the Sautrantika school relegates direct sensory perception or evidence (pratyaksa) to the level of phenomenal (samvrti) truth and inference or rational demonstration (anumana) to the noumenal (paramartha) truth. In Sautrantika, the real is given only to reason, and the correspondence (sarupya) between internal cognition and external cause is the criterion of truth.

2.5.1 Time, Space and Atom in Early Sautrantika

Time Even though the Sautrantika claims itself to be a reconstruction of the original sutric teachings of the Buddha, the Sautrantika doctrine of momentariness does not figure in earlier Buddhist writings and must be considered a proper innovation of the Sautrantika system. In Sautrantika, momentariness is not confined to phenomenal appearances—even ultimate truths such as substances are held to be in flux.

Von Rospatt traces the conceptual evolution of the instant (ksana) from the momentary interval to the time-atom (irreducible duration) to the completely durationless present in great detail, and summarizes his findings as follows:

The usage of ksana in the sense of momentary entity documents that the change in the conception of the term ksana was brought to its logical conclusion. Starting out with the basic meaning of “very short time,” the ksana came to be understood—reflecting an atomistic conception of time—as “the shortest unit of time,” the length of which came to be equated with the duration of mental entities (or transient entities in general) as the briefest conceivable events.” The Sautrantika view of momentariness as lack of temporal persistence can be traced to Vasubandhu’s definition in the Kosa, as von Rospatt explains:

To be ksanika (i.e. momentary), that is to be endowed with such a ksana, then entails according to Vasubandhu to perish imme­diately after having originated. Rather than defining ksanika as “being of momentary duration,” Vasubandhu in this way spec­ifies the nature of this momentary existence and thus excludes alternative conceptions such as that of the Sarvastivadins.

The idea that a ksana has a certain fundamental thickness differing from a psychic addition may hold true for the Sarvastivada view of momentariness, but is not borne out by Sautrantika sources.

Von Rospatt enumerates five modes of deduction that the Sautrantikas use in support of their doctrine of momentariness. The momentariness of mental entities (1) is one indication. The Sautrantika also attempt (2) to infer the momentariness of all conditioned entities from the momentariness of the mind, and (3) to deduce momentariness from change and (4) destruction. Finally, they adduce (5) the experience of momentariness itself.

Presentism is a doctrine unique to the Sautrantika school (and to Mind- Only, which inherited many Sautrantika doctrines). It is not to be found in classical and medieval Western philosophy, with the exception of the Cyrenaic school.

Since Sautrantika does not admit the existence of past and future en­tities, all that remains of the temporal sequence of past, present, and fu­ture is the principle of continuous, universal flux (samtanai—a dynamic but durationless “momentary now” or “eternal present” where phenomena are continuously in the process of arising or perishing, but never actually persist for any length of time whatsoever. The present moment is a “van­ishing present”—both streaming and standing, like the Jamesian “specious present” or Husserl’s “immanent temporality.” Husserl writes that the subjectively constituted concreta arise from the immanent temporality as the lowermost streaming ground of concreteness. The Sautrantika theory of time may be conceived as a form of “replacement presentism,” as argued convincingly by Grupp."

It is important to emphasize that Sautrantika not only rejects the exis­tence of temporal continuants, but eliminates even instantaneous existence. Phenomena and elements {dharmas) can therefore never be found as com­plete existents in the present—one only ever finds their arising or perishing. Garfield points out that in Madhyamika, even the domain of conventional phenomena cannot be resolved into constantly arising, constantly ceasing momentary phenomena that have inherent existence. In contrast to this Nagarjunian doctrine of total ineffability, the Sautrantika system asserts a considerably more ontological view—that there is individual arising and ceasing of ultimates, although the individual entity is always in flux.

Space Unlike the Vaibhasikas, some Sautrantikas do not regard the ex­periential constituent of phenomenal space {akasa-dhatu) as a substantial entity (dravyasat). The Sautrantika view advanced by Vasubandhu in the Kosa, construes shape as a derivative property of color. As Sellars trenchantly observes, the notion of shape as existing only “by way of idea” is a Berkeleyan one. But in Sautrantika, not all spatial properties are conceptually imputed—e.g. distance is still held to be an ultimate. The account of samsthana as purely a mental conception rather than a substantial and real property in the Kosa may be a possible precursor to Vasubandhu’s argument against atomism, as it shows the beginnings of a critique of space.

Aside from shape, the Sautrantika are also beginning to whittle away at the spatial interval of the atomic interstice: “The Sarvastivadins- Vaibhasikas held the view that there are intervals between the atoms of the conglomeration, while the Sautrantikas maintained that there are no intervals and yet the atoms do not touch each other.” One explanation would be that the Sautrantika held this view chiefly for apologetic reasons. If they had declared outright that the atoms touch each other, they would have been dismissed as having the view of a non­Buddhist school, the Vaisesika. This is the way Vasubandhu seems to deal with this issue in the Kosa. He argues that the Sautrantika themselves admit that it amounts to as much as touching: “Quoting an authority, Bhad- hanta, Vasubandhu says that although the atoms do not touch, when they are situated in the closest, gapless proximity we can say in words, ‘they touched’.”

Matilal advances the explanation that the Sautrantika are claiming a potential or latent tangibility of the atoms which is actualized when they are in conglomeration, just as the latent visibility of individual atoms is actualized when they are perceived in aggregate.

A third explanation would be that Sautrantika held the interstice between two atoms to be mathematically infinitesimal, in other words, smaller than any possible measure. This would account for for their claim that atoms can have the seemingly contrary properties of not touching, yet not being separated.

Atom The Sautrantika understanding of the two truths—the phenome­nal (samvrti and the noumenal (paramartha) truth—bears affinities to the Cartesian dualism of the 19th century, where the atom (res extensa) is the fundamental entity of the external world and sensation (res cogitans) are the fundamental entities of the internal world, i.e. the sensory minima brought about by the impingement of atoms upon the sensory organs. The external mosaic of atoms is connected to the inner world of perception by way of the sensation, the smallest physiological unit, which covers up but at the same time reveals the external world.

In Sautrantika, the phenomenal or conventional truths are sensibles and the ultimate truths are intelligibles, i.e. the noumenal structures that are deduced to objectively back the perceptual flux of experience. The ultimate constituents of external objects are believed to be inferable from the “repre­sented” forms in our awareness. Atoms are momentary (ksanika) and are given only to inference (anumana) as purely logical constructs. We are led to deduce their existence by a causal inference, as the atomic data must have caused the gross appearance (pratibhasa) in consciousness. But though this seems like an a posteriori reasoning at first glance, it is actually an a priori one. One is not drawing conclusions from the empirical to the empirical, but from the empirical to the noumenal. The atom is reasoned to be a necessary precondition to empirical knowledge. Sautrantika thought does not have a concept directly corresponding to the Kantian notion of a priori. But inasmuch as we are justified in reconstruct­ing Sautrantika along Kantian lines, the atom would have to be a wholly transcendental construct deriving from a priori consideration.

One supporting argument for the a priori interpretation is that the Sau­trantika atom is imperceptible in principle. In Sautrantika, ultimate truths are sui generis and reachable only by transcendental reasoning, not by a decompositional analysis of empirical phenomena. The Tibetan doxographer Ngawang Belden offers the example that in Sautrantika, a patch of blue is not established as a composite of external particles separate from the aspect of blue, nor is the aspect of blue a composite of external particles. The Vaibhasika atom, on the other hand, can at least in principle be isolated by empirical decomposition. It is only too minute to be individually registered (though we have already discussed the problems with this view).

Sautrantika epistemology has been compared to the representational realism of Locke. Its atomic theory is also similar in many ways to the theory of Epicurus, who also adverts to a rationally apprehended world of atoms for causal explanations of the phenomena appearing in sensory awareness. Clearly, Sautrantika resonates to a reasonable degree with Kantian critical realism, although the Buddhist and Kantian accounts differ considerably as to their forms of construction and categories.

In the following, we will therefore pursue the Kantian line of interpreta­tion and construe the Sautrantika’s inferential knowledge of the noumenal realm (paramartha-satya) to have the form of necessary a priori posits— even though there are problems with this interpretation due to anumana not having the proper technical meaning. The postulation of a necessary presupposition to knowledge would be arthapatti, which is not among the Sautrantika’s allowable pramanas.

Why does the Sautrantika system need atoms at all? The doctrine of atoms was carried over from Vaibhasika, just as the Vaibhasikas in turn had most likely inherited the doctrine from the Jainas and Vaisesikas. Also, perhaps it was considered that appearances required a substance account of causal origination in order to avert the conclusion that they are being generated ex nihilo. Finally, if the Sautrantika causal theory of perception (that mental representations are the effects of external objects) is to support an externalist reference of the intentional object (visaya), it requires that there be objective causal vectors for effective action.

Mind (citta) in Sautrantika moreover possesses the feature of svasam- vedana (self-reflexive awareness, i.e., transcendental apperception)—but in other key respects, it is not transcendental, as the moment-to-moment arising of mind is governed by empirical causality.

One difficulty with the Kantian analogy is that dravya does not conform to Kant’s notion of substance. For Kant, the schema of substance is persistence through time, and we have already shown that Sautrantika does not admit dravya to possess any temporal persistence whatsoever. The Sautrantika concept of dravya would more accurately correspond to the Kantian notion of noumenon or Ding an sich, i.e. a thing as is apart from sensible or empirical intuition. But this reading, too, immediately breaks down, as the Sautrant­ika atom is in flux, whereas the Ding an sich is atemporal.

Nonetheless, the existence of atoms, i.e. of intelligible simples, is given to reason by a kind of transcendental inference, as a merely empirical analysis would never suffice to establish atoms as the nonempirial objects which they are by definition intended to be. They comprise the flux of momentary atomic stimuli that causally induces tactile, visual, aural etc. sensations—from which, in turn, the mind synthesizes its phenomenal representations by way of a process of conceptual elaboration {.vikalpa). In Sautrantika thought, vikalpa generally assumes the role of synthesis, albeit an imaginatively contrived and therefore “unreal” one. Also, if atoms are purely intelligible but not sensible, Sautrantika must rigorously separate the category of substance into rupa (variously construed as phenomenal materiality, Gestalt, or as the constituent of a perceptual datum which presents itself as objective) and dravya (the transcendentally intuited substance, that which is inferred by transcendental reasoning to be the objective basis for perception). If a dravyasat lacks temporal persistence or sensible materiality, why even call it a substance? For an existent to be intelligible may mean no more and no less than that it is a logical posit, albeit a necessary and true one. But what remains unexplained is why such a logical datum should exist independently of its being posited.

2.5.2 Atomism in the Late Sautrantika Phase

Gelug scholarship distinguishes between the Followers of Scripture and Followers of Reasoning. The late phase of Sautrantika is known in the Gelug doxography as the subschool of “Sautrantika Following Reasoning.” This sub-division of the Sautrantika school following the main system of Dignaga and Dharmakirti postulates as its atomic entity the point-instant or unique, unanalyzable unrepeatable particular (svalaksana).

Chatterjee characterizes the svalaksana as an “attributive atom,” which he holds to be the atom in its truest sense. Grupp translates the svalaksana as “abstract atom.” Dharmakirti’s svalaksana may be compared to Carnap’s concept of elemen­tary experience (Elementarerlebnis) in that they both explode the stream of experience out into unanalyzable particulars. However, the similarity ends there. For Carnap, the Elementarerlebnisse are derived by an abstrac­tive process (the procedure of quasianalysis) and employed as constitutional elements, not asserted as metaphysical realities. Atoms are products of abstraction—the stream of experience is not intrinsically atomized in this way. For Dharmakirti, the svalaksanas are the minimal concrete partic­ulars or “givens” of perception that the real splits itself into. As Goodman points out, a system taking concreta as atomic cannot allow qualities as atomic parts of such concreta. This difficulty is resolved by the theory that phenomenal qualities are merely superimposed on the svalaksanas as subjective constructions. The svalaksanas do not constitute anything, as is the case for the Elementarerlebnisse of Carnap’s system. Moreover, the Sau­trantika Following Reasoning assert these as external (mind-independent) objects, which Carnap does not.

The svalaksana nonetheless signals a final departure from the realism of the “substantive atom,” the latter being much closer to the common-sense view. The Sautrantika system of Dignaga and Dharmakirti postulates partless and extensionless naked particulars in a strictly instrumental fash­ion, as a provisional stepping-stone to the fully-fledged Yogacara view. The “attributive atomism” is later in origin than Vasubandhu’s Vimsatika and is unaffected by his anti-atomist argument, for reasons that will be explained later. However, there is by no means a unanimous consensus that Dharmakirti divides time into infinitesimal or point-like instants. Oetke argues that this view is unproven, and that Dharmakirti’s views on the structure of time (whatever they may be) have no bearing on his formal proofs of momentari­ness in the Sattvanumana.^

2.5.3 Critique of Sautrantika Representationalism

Passing now to criticism, one difficulty with representationalism is that it does not fully accord with radical phenomenalism. The phenomenalizing stance as expressed in the Bahiya Sutta would seem to demand a two-term or even a one-term account of perception. If in the seeing there is only the seen (etc.), then the perceived must be identified either with the representational aspect of perception (as is the case in Vijnaptimatra) or with the aggregate of the compositional constituents of perception (as in Vaibhasika)—it is not possible to have it both ways.

Chatterjee argues that the three-term theory is an unstable account of knowledge. Its correspondence theory of knowledge presupposes a presenta- tive account, which, however, renders the correspondence theory superflu­ous.

Also, the novel Sautrantika doctrines of evanescence and flux fit poorly with the doctrine of atomism inherited from the Sarvastivada tradition. The Sautrantika have some difficulties in reconciling atomicity to flux, as the discrete nature of atoms would seem to break the continuity implied by absolute flux.

If the doctrine of momentariness is made to apply to atoms, as it must be, then the atom ceases to be a static entity and becomes a wholly dynamic stimulus. It flashes into and out of existence, but does not actually persist for even the most infinitesimal moment. To use a mathematical metaphor, the atom’s process of arising and perishing might be compared to a Cauchy sequence which, though converging in the reals, fails to converge in the rationals. Despite approaching an ideal limit (the idealization of the atom qua static existent) there is never in fact any such static entity to be found at any point of time during the atom’s process of real arising and perishing. In this reconstructively purified Sautrantika view, flux and atomism are reconciled in the concept of the atom as a fundamental flux-unit, a discrete pulse of effective action (not unlike the quantum of modern physics). As a key rationale for their radical dynamization of the atomic concept, the Sautrantika offer that a static entity is by definition incapable of acting.

Vasubandhu points out that the dynamic, instantaneous nature of the atom-stimulus poses a severe difficulty for the Sautrantika’s causal account of perception—by the time the percept, i.e. the representation is registered in awareness, the stimuli that have caused it have already evanesced. Under the assumption of an externalist, i.e. a correspondence theory of represen­tation, the percept comes to be a representation of the unreal. However, if one is willing to revert to internalism to preserve the veridicality of the representation, then why not also adopt an internalist account of causation and dispense with the noumenal object? The internalist account of reference and knowledge is simply good empiricism, as Vasubandhu might aver. A coherent account of perception does not require external bases of knowledge (alambanas). This will be the nucleus of his argument against the causal theory of perception in verse 16 of the Vimsatika.

Although the early Sautrantika epistemology is already remarkably ad­vanced, we still find in it some vestigial notions inherited from the Vaibhasika direct realism, such that these discrete pulses must possess definite spatio­temporal locations. As we will later see, all of these notions are to become the object of Vasubandhu’s critique of atomism. However, in early Sautrantika we also find notions proper to critical realism, such that the discrete pulses are momentary stimulus-atoms which are themselves imperceptible, but produce percepts by impinging upon the sensory faculty. In the late Sau­trantika of Dignaga and Dharmakirti, the corpuscular concept is done away with—finally the atomist doctrine, now fully purged of all substantialist vestiges, has matured into “attributive atomism.”

Thus, we can reconstruct the evolution of the atomic concept in the Ab- hidharmika schools along roughly the following lines: In the Vaibhasika presentational theory of knowledge, the atom is a physical sensation-atom or trope, i.e. a discrete, particular instance of a sensory quality. In early Sau­trantika representationalism, although the atom’s noumenal substantiality has not been eliminated entirely, it is hidden from phenomenal perception. What was formerly a sensation-atom or unitary sensum is now a discrete sensory stimulus, the necessary existence of which is deduced by (transcen­dental) inference.

The Abhidharmic reductionism attempts to epistemologize the early Bud­dhist teaching of the phenomenalistic elimination of references to self and subjectivity by reducing the physical world described by ordinary language of covering truth (samvrti-satya) to the purely phenomenal world described by the sense-datum language of ultimate truth (paramartha-satya) But whether this program actually plays out as intended is another question. In a parallel to Quine’s classic critique of reductionism as set forth in Two Dogmas of Empiricism, Oetke strongly argues that this phenomenalizing reduction of ordinary experience to “protocol language” is infeasible.

Moreover, the Abhidharmic theory of synthetic wholes as purely nominal entities raises the possibility of similarly nominalizing the parts. Saying that the whole is a synthesis (vikalpa) superimposed on the parts by purely subjective constructive imagination (kalpana) bears the risk of reverting into idealism, because it might be that the subjectivity is constructive enough to posit the parts as well.

In fact, this will be the thrust of Vasubandhu’s argument in the coming chapter. He will argue that the hypothesis of transcendentally objective atoms is unwarranted and superfluous. This, he will say, is firstly because the gross appearance (pratibhasa) is all that is required for empirical purposes, and secondly, because it is not proven that we constitute perception of gross physical objects from transcendentally objective atoms given in experience.

The example of Sautrantika illustrates an Avenarian point—any empiri­cism made aware of its a priori foundations but insisting on realism in­evitably produces a characteristic diplopia of inner experience and outer reality, which raises the problem of how these two realms interrelate. There are two very closely related modes of thought that attempt to resolve the bifurcation of internal and external world that Avenarius famously problematized as “introjection”—phenomenalizing idealism (the route taken by Vasubandhu) and empiriocriticism—the new turn of Mach and Avenarius after Kantian critical realism which reasons against the metempirical status of the synthetic a priori.

3 The Anti-Atomist Metaphysics of Vijnaptimatra

I once heard the question seriously discussed, "How the perception of a large tree could find room in the little head of a man?" (Ernst Mach) HE TASK CUT OUT for this chapter will be to argue, by conducting a detailed analysis of the relevant arguments in verses 11-16 of the Vimsatika, that Vasubandhu really did intend his refutation of the external object to operate as a critique of transcendental objectivity. In other words, by undercutting the epistemic justification of the metaphysical realist’s existence claims for such putatively external objects, he aims to demonstrate that the existence of objects ontologically independent from our own mental representations is not established.

Some modern revisionists regard Mind-Only epistemology as a phe­nomenological realism that focuses its attention on the flux of experience and chooses not to concern itself with metaphysicalities. It is entirely correct to point out that for Mind-Only, things-in-themselves are philosophically gratuitous. In point of fact, Mind-Only generally refrains from engaging in ontological theorizing. Yet this minimal analysis does not go nearly far enough—for if a philosophical position which is non-materialistic, excludes all extramental elements from the constitution of experience, insists on viewing the world strictly in epistemological terms, and maintains that experience of the given moment is noematically constituted solely by the generative activity of transcendental subjectivity does not deserve the label of idealism, then what does?

Any reading of Vasubandhu’s argument against atomism will inevitably be influenced by one’s expectations about the position that he is attempt­ing to establish. A critical review of some of the more popular interpre­tations of Mind-Only will help to triangulate the epistemological intent of Vasubandhu’s argument and set our prochronistic parallels to empirio- criticism on a firmer footing.

The Vijnaptimatra eliminates the object term of the three-term theory of perception advanced by Sautrantika and returns to a two-term theory of perception. However, it does not quite come out to direct realism, as Mind-Only bars all epistemological means of access to the noumenal realm (paramartha-satya). It confines ordinary perception (pratyaksa) and rea­soning (anumana) to the egological sandbox of the conceptually constructed nature of experience (parikalpita-svabhava).

3.1 Readings of Mind-Only

When discussing the Vijnaptimatra school, it is useful to keep three dis­tinctions in mind—(1) the schools “Following Scripture” and “Following Reasoning,” (2) the Nalanda and the Valabhi schools, and (3) the True and False Aspectarian readings.

Tibetan doxography distinguishes the Yogacara of Asanga and Vasubandhu (“Following Scripture”) from the later school of Dignaga and Dharmakirti. (Following Reasoning). In this exposition, we will unfortunately have to restrict ourselves to a discussion of the doctrine of Vasubandhu—the logico- epistemological idealism of Dignaga and Dharmakirti will be largely kept out of consideration, as will the Sino-Japanese reception history of the Mind-Only school.

Within the schools of interpretation that evolved from the early Yogacara of Asanga and Vasubandhu, we must distinguish the Nalanda school, most prominently represented by Dharmapala, and the Valabhi school, repre­sented by Sthiramati. There are several notable differences in interpre­tation between these two schools, of which we will mention only one—the ontological status of the mental representation in terms of the Yogacara three-nature model. For Sthiramati, the aspect (akara) or noema is relegated to the conceptually constructed (parikalpita) nature of experience, whereas Dharmapala holds it to be a part of the other-dependent (paratantra) nature. In according to the akara a causally effective mental reality that is not imaginary conceptual construction, the Nalanda school comes to be much closer to a Berkeleyan idealism than the Valabhi school, for which the akara is unreal, and ultimately seen to “dissolve” back into the interdependent functional flux of the paratantra nature. The third distinction, the distinction between True and False Aspectarian- ism as drawn by Tibetan doxography, will be covered later when we review the traditional interpretation of Cittamatra.

3.1.1 Revisionist Interpretations of Mind-Only

The revisionist line of interpretation generally insists on reading Mind-Only not as metaphysical idealism. Rather, it is held to be a phenomenologically oriented species of pluralist realism. Instead of being concerned with the way things are qua metaphysical realities, all that matters to Yogacara is the way phenomena are perceptually and mentally constituted. It is in other words an attempt to reformulate the basic program of Buddhist empiricism in terms of a comprehensive phenomenological analysis of the activities of the mind. If it is to be considered idealism at all, it is at best a form of epistemological, but certainly not of ontological idealism.

The most influential modern proponents of this phenomenological inter­pretation are WAYMAN, Anacker, KING and LUSTHAUS. Some revisionist scholars claim that Yogacara accepts an external world and urge a critical realist interpretation (notably WAYMAN and KOCHUMUTTOM).

Wayman Wayman alleges that the proponents of the idealist interpretation support their thesis by a selective reading of the textual evidence, bending statements that indicate Yogacarin belief in objective rupas (forms) and other mind-independent elements. But under the premise that Cittamatra is not an entirely fresh creation, but in many respects an evolution of an earlier critical realist scholastic system of the Sautrantika, it shouldn’t be surprising to find that it still preserves many overtly realist elements that have not yet been reformulated or reinterpreted to fully harmonize with idealism. Given the nature of Mind-Only as a major locus of synthesis for Buddhist philosophical doctrine, it would in any case unreasonable to assume that Cittamatra positions are always univocal. It is only natural for the overarching exegetical theme of idealism to be unraveled into a kaleidoscopic multiplicity of perspectives upon closer and more critical readings of the text. A deconstructive hermeneutic uses exegetical themes as prisms for fracturing the text and opening up its inner interpretive space for many meanings old and new. One has to wonder whether a project like Wayman’s which seems to pursue the hermeneutically rather naive goal of uncovering the hitherto hidden “true meaning” of a text really deserves the compliment of revisionism.

Wayman is not alone in giving a strongly critical realist rendering of Citta­matra—Lindstrom, for example, even goes so far as to make the remarkable assertion that the point of Yogacara “is in fact to assert, not deny, ‘the mind-independence of the material sphere.’ ”

I will attempt to show, by a rational reconstruction of Vasubandhu’s argu­ment against atomism derived from the internal evidence of the text, that quite the opposite is true. Moreover, it is reasonable to construe Mind-Only as a species of idealism on the external evidence of the arguments that its Advaitin and Madhyamika critics have lined up against it. One may of course always argue that the historical critics of Mind-Only have universally misconstrued it, but this amounts to contending that only a proponent of Mind-Only can understand the system well enough to properly critique it. This argument can safely be dismissed.

Lusthaus Lusthaus insists that the mainline ontological idealist reading of Yogacara is predicated on a fundamental misinterpretation of the term vijnapti-matra. He argues that correct interpretation implies no more than a realist phenomenology, the idealist features of which are confined to a strictly epistemological sense—in other words, any ontological sense of the term is precluded:

Why has Yogacara been misinterpreted as idealism? The com­mon way of interpreting matra so as to valorize ‘consciousness’ is striking since those same interpreters never impute such im­plications to matra on the many other occasions it is used by Buddhists or Yogacarins. For instance, the closely allied term prajnapti-matra (“only nominally real”) has never led a modern interpreter to speculate that Language is the metaphysical reality behind the world of experience; on the contrary, those prone to idealist interpretations tend to privilege ineffability and yearn for a realm beyond language and conceptions.

Naturally, since anybody would agree that the term “only nominally real” doesn’t indicate that language is the metaphysical reality behind the world of experience, Lusthaus wants us to agree, mutatis mutandis, that vijnapti- matra (in his words, “nothing but noetic constitution”) doesn’t indicate noetic constitution to be a metaphysical reality behind the world of experience. By this analogy, he wants us to believe that an ontological reading of the term vijnapti-matra is wildly implausible. But in order for this analogy to carry his argument, Lusthaus relies on a deliberately misleading implication. The fact that unless we are specifically discussing nominalism about propositions, we typically don’t take the term “only nominally real” to be making a metaphysical proposition about language has nothing to do with the fact that we may very well believe it to be making a metaphysical statement about the relationship of language and object. If we say, for example, that a tree is only nominally real, we are obviously talking about the referent of the term “tree,” not about the term “tree.” Not to belabor the issue ad nauseam, but given that the desire to minimize or eliminate ontological commitments is such a crucial philosophical motivation of nominalism, if we all of a sudden make out “nominalism talk” to be about something other than ontology, then there isn’t much left for it to be talking about!

Lusthaus is right to point out the importance of epistemological idealism. But is he really justified in ruling out ontological readings of vijnapti-matra? Perhaps he is—if, somewhere, he offers a better argument for it than the one we have just discussed.

One can certainly credit Mind-Only epistemology with the intent to be non-metaphysical. Carnap makes the perspicacious observation that realist, idealist and phenomenalist epistemologies coincide to the extent that their constitution theories of experience are ontologically neutral. This obser­vation can help lay to rest a large portion of the perennial vexation about whether Mind-Only is realism or idealism or phenomenalism—to the extent that it happens to be non-metaphysical, all three terms equally apply. Therefore, we have no doubt that Lusthaus is on to something important about Mind-Only. Also, we are not disputing anything that Lusthaus has to say about its Sino-Japanese history of reception. But what if Vasubandhu’s position in the Vimsatika, does, after all, lapse at some point into meta­physics? This is in fact what Cittamatra’s Indian and Tibetan Madhyamika critics oppugn it for. Should we ever catch such a lapse, it would come as no surprise to find that it is an idealist one. If epistemological idealism refuses to let its metaphysics in through the front door, it may simply enter through the back.

Some scholars conjecture that the early phase of Yogacara was mainly phenomenological in orientation, whereas its later developments came to be more ontological. Nagao, for example, notes that the early Yogacara of the Asanga-Vasubandhu era considered the question of an external reality to be a problem of the realist’s ontology, and thus, a problem which was not their concern. However, it would be a mistake to conclude that they therefore desisted from external world skepticism. Lusthaus is more circumspect about the issue of the external world than Kochumuttom. He hedges his bets—while he does not actually say outright that he believes Mind-Only to be critical realism, he conveys this opinion obliquely: “To the extent that epistemological idealists can also be critical realists, Yogacara may be deemed a type of epistemological idealism ...

” Against the view of Lusthaus, I am arguing that the intent of Vasubandhu’s refutation of atomism is not compatible with critical realism—since if the transcendental object is bracketed out of the constitution of experience, critical realism is left behind for good. But as Kant recognized, this is a dangerous philosophical move to make. An epistemological idealism without an objective world cannot hold the middle ground and is inexorably pulled towards an ontological idealism. Once we lose the solid transcendental sea­bed on which to drop our “reality anchors,” our world of experience becomes driftwood on the deep of idealism—or to use a more empiricist figure of speech, we become sailors on Neurath’s raft.

But the Vimsatika has several levels of meaning. Vasubandhu’s soterio- logical subtext is that once we wake up to the fact that we are empirical explorers floating on Kon Tiki, we should steer the raft towards the far­ther shore. The anagogic reason why the Vimsatika is force-feeding the realist with the epistemological idealism of Cartesian epoche is to lead him even further into skepticism—not because it wants to take a last stand on any kind of idealism. However, if the indication of being prone to idealist interpretations of Cittamatra is to “privilege ineffability and yearn for a realm beyond language and conceptions,” as Lusthaus alleges, then surely, Vasubandhu the mystic stands guilty as accused.

Kochumuttom In Kochumuttom, we find a very systematic and generally quite convincing Kantian revisionist reading of Vasubandhu. He makes the argument that Vasubandhu does not deny the noumenon or Ding an sich outright, but merely claims that it can never be directly given in experience unless in the condition of enlightenment. Generally speaking, this is a per­fectly accurate interpretation, but what this really boils down to upon closer examination is that Vasubandhu is a critical realist only in the soteriological sense. There can be no doubt that Vasubandhu believes in an ineffable noumenal nature (anabhilapya-atma) beyond phenomenal experience—the true nature of minds known to the enlightened ones. But first of all, if this is Vasubandhu’s noumenon, it is nothing like a Kantian idea of pure reason. The only rationale for calling it a noumenon is that

Vasubandhu avers it to be beyond ordinary experience. Because it is beyond knowing as we know it, it does not supply any kind of ontological or epistemic grounding, nor can it be rationally intuited. Vasubandhu already shows his true colors in verse 1 of the Vimsatika, where he likens the perception of external objects to seeing cataract-hairs. He is trying to communicate that all the while, we are staring into the face of the ineffable noumenon, but simply fail to realize it because we are ignorantly fixated on the obscurations of ordinary experience. This example goes to show in the context of a Mahayana mysticism in full bloom, Vasubandhu’s belief in an ineffable noumenon is insufficient to establish that he is a critical realist when it comes to the epistemology of ordinary experience—in fact, it establishes the very opposite. Let us examine what Kochumuttom has to say about Vasubandhu’s argu­ment against atomism:

Nowhere during the discussion does he say that there is no extra-mental world. Instead he has thrice said that “an atom is not obtained”. The term translated here as ‘is obtained’ is sidhyati. To be sure, this term does not mean ‘to exist’ (asti). Therefore, to translate the above sentence as “an atom does not exist” would be a gross mistake. The usual meanings of the term sidhyati are ‘to be obtained (in experience),’ ‘to be given (in experience)’ or ‘to be proved to be true’ etc. So Vasubandhu’s main criticism against the atomic realism is that the atoms are neither given in experience nor proved. Therefore he does not really say that there are no atoms at all, although he is not prepared to admit that things-in-themselves which are ineffable, could be conceived in terms of atoms.

There isn’t much to find fault with in this discussion. It shows that Vasubandhu is rightly very wary of making metaphysical assertions involv­ing the existence or nonexistence of objects beyond phenomenal experience. All he says—and all he needs to say—is that atoms are neither given in phenomenal experience nor proven to constitute experience. However, if it is really true that Vasubandhu may read as accepting non- atomic Dinge an sich in the Kantian sense, something is conspicuously missing from Kochumuttom’s interpretation. Why doesn’t Kochumuttom supply a reasoning that establishes the necessity for Vasubandhu of some­thing transcendentally objective as the condition for the function of the categories (such as is done by Kant in his thesis of the second antinomy of pure reason)? The conclusion to be drawn from this question is that if Kochumuttom doesn’t have any such reasoning to show, then there simply isn’t anything to be shown.

Additionally, there is a systematic difficulty with Kochumuttom’s read­ing of Cittamatra as a variation on Kantian critical realism. If Vijnana- vada and Sautrantika are both critical realisms, then what exactly does Vasubandhu stand to gain by refuting the Sautrantika theory of perception in the Vimsatika!

The fact that Kochumuttom cannot sufficiently substantiate the necessity to Vasubandhu’s epistemology of the Ding an sich despite that Vasubandhu’s reasoning is generally Kantian is a key justification of the empiriocriticist reading we are attempting here. Kochumuttom agrees that Vasubandhu neither affirms nor denies the existence of mind-independent atoms in an ontological sense. This is reason to conclude that Vasubandhu is developing the Vijnaptimatra epistemology in Cartesian epoche.

3.1.2 The Traditional Interpretation of Cittamatra

According to the traditional line of interpretation, Cittamatra is metaphysical subjective idealism. The view that the goal of the Vimsatika is to deny the existence of a mind-independent external world is backed more or less uniformly by the classical Indian and Tibetan commentarial tradi­tions (see e.g. HOPKINS, BLUMENTHAL, or Santaraksita), and by many modern scholars such as WOOD, TOLA and DRAGONETTI, CHATTERJEE, Gupta, Matilal and Nagao.

Arnold points out that Yogacara has reduced the Abhidharmika list of categories to mental events, which may be taken an indication of a Yoga­cara view that the mind is epistemically prior to the psycho-physical complex of experience. If it weren’t for the mind, in other words, there would be no sort of experience whatsoever. Wood, as we have already seen, argues that the Mind Only Principle frames Cittamatra as a species of ontological idealism. Matilal contends that Yogacara idealism sets out to demonstrate the inconsistency of the realist account of the external world.

Tibetan doxography unanimously reads Cittamatra as a critical refinement of Sautrantika which purifies it of the extraneous and self-contradictory category of the external object. As Mipham states: “The Sautrantikas, who assert the theory of the mental aspect, are like the Chittamatrins. The sole difference lies in the assertion or denial of the existence of the external object.” Furthermore, the Mind-Only view resembles Sautrantika in accepting their doctrine of momentariness. On this understanding, the Cittamatrin reveals himself to be a Sautrant­ika who comes to realize that he has been cast out of the phenomenalistic garden of Eden into the physical world for having eaten of the forbidden fruit of metaphysical realism. Vasubandhu’s offensive against the external object fbahyartha) is an attempt to rectify this situation by ejecting the Ding an sich from the epistemological constitution of experience.

King argues that the “phenomenalism” of the Yogacara school is clearly in conflict with the Sautrantika epistemology about the issue of being able to make veridical statements about a mind-independent external world. We will later return to this point of conflict and argue it more strongly—the attack on Sautrantika externalism is in fact one of the centerpieces of the Vimsatika’s philosophical program.

3.1.3 Distinguishing True and False Aspectarianism

In Indian philosophy of mind, the debate between idealism and realism revolves around the issue of whether consciousness is formless or has a form. Realistic theories (such as those of the Nyaya and Mlmamsa schools) generally hold that consciousness is formless, whereas idealistic theories (such as Mind-Only) generally argue that consciousness has a form. This theory of the mind provides justification for their thesis that external objects are forms of consciousness. In an attempt to get a better philosophical grip on the the precise sense in which the Mind-Only view denies external objects, we find this debate between idealism and realism being carried out even within Mind-Only, where it centers on whether the sensible forms or aspects of consciousness are true or false.

The True and False Aspectarian views differ as to how far they press the skeptical assault on ordinary experience. Nonetheless, they are in agreement in maintaining that if one subjects ordinary experience to a transcendental analysis, one will discover what seemed at first to be the external world and its objects to be mere mental representation (vijnapti-matra) lacking extra-mental referents. They both assert (in slighly different ways), that phenomena are mere conceptual imputations (prajftaptisat) lacking sub­stantial existence (dravyasat). However, according to Gelug doxography, both True and False Aspectarian views assert phenomena to be established as the substantial entity of the mind. Tsongkhapa attempts to resolve this discrepancy by arguing that because the Mind-Only sense of the term prajftaptisat is less rarefied than the Madhyamika sense, the Mind-Only view is realized under more subtle Madhyamika analysis to entail substan­tial establishment as mind. We may think of such hermeneutic maneuvers as we wish, but they do underscore the necessity of carefully distinguishing the assertions made by a given philosophical system from meta-philosophical interpretations of these assertions—which is admittedly usually much easier said than done.

True Aspectarianism This interpretation of Mind-Only as “realistic ide­alism” denies external objects transcendentally but accepts them on an empirical level, i.e. as conventional/pragmatic truth. Following the Kantian distinction of transcendental and empirical externality, True Aspectarianism denies the transcendental externality or substantiality (dravyatva) but not the empirical externality of objects (i.e. their rupa), or their spatiotemporal determinacy (niyama). Their supposed externality is illusory in a strictly transcendental sense—but not, however, in an empirical sense. True Aspec­tarianism is roughly what Kant’s transcendental idealism would amount to without the Ding an sich.

The True Aspectarian position entails a rejection only of dravya, not of rupa. The concept of rupa may perhaps be best understood under a phe­nomenological reading as hyle or material Gestalt of perception, whereas dravya is a metaphysical term signifying transcendentally objective sub­stance.

True Aspectarianism is still foundationalist because it holds that state­ments about sense-data (e.g. observations) do not require justification. It adopts an idealist position with regard to transcendental reality but a phe- nomenalist empiricism with regard to empirical reality: “The Yogacara is an idealist only transcendentally; in empirical matters he has no quarrel with the realist. All philosophical issues lie between the conflicting interpretations of the facts and not the facts themselves.”

False Aspectarianism The “anti-realist idealism” of the False Aspectarian reading radicalizes the Cartesian skepticism of the True Aspectarian view in holding the external object to be illusory even in the empirical sense. It surrenders the last remnants of realism about the external world to skeptical doubt in rejecting the typical notion of traditional sense-datum theory that raw sensations are incontrovertible. Even the qualia themselves (e.g. the perception of blue as blue) are held to be polluted with innate ignorance (avidya). The desired rock-bottom certainty is to be reached not in sense-data or evidence, but only in the pure self-reflexive apperception. In contesting even the validity of empirical evidence, False Aspectarianism no longer qualifies as phenomenalism sensu stricto.

The False Aspectarian position that all perception—true to the spirit of Brentano’s famous dictum “Wahrnehmung ist Falschnehmung”—is tout court erroneous is closer to the position of the Madhyamikas and Advaitins, who both regard awareness as formless (nirakara) in and of itself. False Aspectarianism is a case in point that the formlessness of consciousness need not entail realism about external objects—it can just as well lead to a radically fictionalist view.

Regarding the nature of the mind, False Aspectarianism is much closer to the Advaita position, as in False Aspectarian interpretation, “raw feels” are mistaken and “consciousness is ultimately clear like a crystal.” The false aspects are extrinsic to self-awareness, which is in itself aspectless, veridical and nondually self-knowing. The False Aspectarian view gravitates even further in the direction of transcendental subjectivity, as it is not the tran­scendent, egoless and pure subjectivity that infects the mind with mistaken qualia, but the egological consciousness.

The False Aspectarian anti-realism about sense-data is partially vindi­cated by findings of contemporary psychophysics. For example, Pinker points out that outdoors, a lump of coal reflects more light than does a snowball does indoors. The absolute luminance of the “black” lump of coal is higher than the “white” snowball. Nonetheless, the coal is seen as black and the snowball is seen as white, because our visual system determines the light­ness of an object by its relative luminance, i.e. it takes into account the luminance of its surroundings. On a similar note, Mach observes that a bright surface seems brighter next to a dark surface than next to one brighter than itself. Thus, although perceived qualities such as color and brightness seem to be absolute and independent, they are actually influenced by their surroundings.

In the exposition Madhyamakalamkara (Ornament of the Middle Way), Santaraksita critiques the False Aspectarian view, arguing that it is fraught with inconsistencies. However, it is admitted that in its assessment that the phenomenal qualia are ungrounded, the False Aspectarian position better approximates Madhyamika than the True Aspectarian interpretation.

According to Tsongkhapa, the True and False Aspectarian positions boil down to being different ways of reading Cittamatra texts, i.e. the Vimsatika and other Mind-Only sastras potentially allow for both a True Aspectarian and a False Aspectarian interpretation. While it is tempting to categorize the Nalanda school as True and the Valabhi school as False Aspectarian, Tibetan doxographers don’t classify them along those lines. In fact, they seem to allow that Nalanda as well as Valabhi school texts can be read both ways. Vasubandhu’s Aspect—True or False? Is the Vasubandhu of the Vimsatika a True or a False Aspectarian? Possibly neither—if our hy­pothesis that the doctrine of Mind-Only is only an interim position is true. In his argument against solipsism in 21, he concludes that since one knows other minds as one knows one’s own—under the influence of avidya—one cannot

have direct knowledge of one’s own mind. This arguably rules out the possibility of the transcendental subjectivity knowing itself in self-reflexive awareness or by any a priori means. Here Vasubandhu is fighting fire with fire. Having closed the escape route to metaphysical realism, Vasubandhu urges that the Cartesian epoche with its unpalatable consequence of solipsism (as either subjective or absolute idealism) can only be broken by an even stronger skeptical epoche. In other words, he is implicitly submitting sunyata, for consideration. If Vasubandhu really intends to complete the Vimsatika on a note of absolute skepticism—and I think he does—then this vanishing point of the text is where its reading must be projected from in order to make sense.

But in order to lead his realist interlocutor to the vanishing point, Vasubandhu must first trap him under the bell jar of Cartesian epoche. This, then, is the philosophical task that Vasubandhu has cut out for himself in his argument against atomism.

3.2 Argument against Atomism

Mereological nihilism claims that wholes themselves do not exist, only their constitutive elementary parts. Vasubandhu now takes the Buddhist ultra­nominalism to its logical conclusion in repudiating even the elementary parts. The argument is a fundamental critique of realist conceptions of the part-whole relation. In his Kosa auto-commentary, Vasubandhu has pressed the Sautrantika critique of least-part substances persisting in time as far as possible. Now, in a change of tack, he brings under fire the notion of substances extended in space.

Schmithausen points out that Vimsatika contains modes of expression which betray an unmistakeable Sautrantika influence. There is therefore good reason to assume a continuity of philosophical agenda with the Sau­trantika substance critique. King’s assessment that Yogacara criticizes atomism merely on experiential grounds is too weak, as we have seen in the previous chapter that only the Sarvastivada realist phenomenalism admits atoms to be within the realm of experience—Sautrantika does not. I shall be using TOLA and DRAGONETTI as the reference translation for the following reconstruction. However, the translations of Anacker, Wood, Kochumuttom and FRAUWALLNER are also taken into consid­eration. Also, I will be making frequent reference to Kapstein’s discussion of mereological considerations in Vasubandhu’s argument, as well as to the insightful analyses of SlDERlTS, CHATTERJEE, and Matilal.

3.2.1 The Scheme of the Argument

The structure of Vasubandhu’s argument is apagogic, as is typical for proofs of idealism. He attempts to disprove the realist interlocutors’ claims that external objects exist and are composed of transcendentally objective atoms given either directly in phenomenal experience or demonstrably constitutive of it. Opponent Lineup Vasubandhu is arguing against three metaphysical real­ist opponents, each of whom propose a slightly different version of atomism. All three maintain mereological atomicity, but only the Vaibhasika and the Sautrantika are mereological nihilists. The Vaisesika maintains that physical entities are real and directly given to experience. He proposes a direct realist account of perception with a foundational ontology of physical atoms in conglomeration.

The Vaibhasika position ventures that physical entities are conceptually imputed to coalesced collections of phenomenalistic atoms. Though the collections are only nominal, the atoms are real quality-particulars directly given to experience. Vaibhasika proposes a presentationalist account of perception based on a foundational ontology of substantial sensation-atoms that appear in collection.

In the Sautrantika view, phenomenal experience is held to be unreal (vikalpa), but has underlying realities that are given to reason as necessary conditions to experience. The Sautrantika representationalism commits to a foundational ontology of momentary stimulus-atoms.

Object of Refutation The object of Vasubandhu’s refutation is not the em­pirically external or outer object (as such objects also also appear in dreams) but the mind-independent object, i.e. the category of the Kantian transcen­dental object (which does not appear in dreams). For the atomistic realist, the existence of such external objects is established by the existence of their atomic constituents.

But Vasubandhu argues that by modus tollens, since the existence of atoms cannot be proven, physical objects (or more technically, the outer ayatanas, i.e. the mind-independently existing sense-fields or logical spaces of sensibilia) likewise fail to be established as truly external.

Vasubandhu’s line of reasoning may be roughly reconstructed as follows: The dream hypothesis of perception adequately accounts for experience and is consistent with empirical evidence. Vasubandhu has already established the internal consistency of his thesis against the realist interlocutor’s objections in verses 2-10. Given that Vasubandhu’s model is basically equivalent in skeptical strength to a Cartesian scenario, it cannot be falsified by any empirical evidence we will ever have. At this point, two closely linked objections may be brought into play:

(1) Williams relates the famous argument of O. K. Bouwsma that an illusion is no longer illusory if there is no conceivable way to distinguish it from reality. An illusion that is utterly undetectable as such by any empirical means can no longer be said to be deceiving us in any practically relevant sense. But under a True Aspectarian interpretation, this objection does not pose a problem to Vasubandhu, as he may argue that experience is illusion only in a transcendental sense.

(2) By Popper’s falsification principle, the dream hypothesis is not an empirical theory because it does not have the capability of being proven false by contradicting empirical evidence. Vasubandhu can, however, safely dismiss this objection as irrelevant because he doesn’t claim that the dream hypothesis is an empirical theory—he is proposing it as a metempirical account of experience.

Method Of Refutation Since Vasubandhu is arguing from the position of Cartesian epoche, he metaphysically brackets the transcendental existence of the external world while realism metaphysically asserts it. Therefore the entire burden of proof in this argument lies on the realist. For realism to be true, the existence of the external world must therefore be established by transcendental reasoning.

What could the realist’s transcendental reasoning establishing the exis­tence of an external world look like? The reasoning must have two steps, analysis and constitution. The analysis (i.e., the completion of Kant’s so- called “regressus of decomposition”) is necessary otherwise the metaphysical realist can’t be certain that he has constituted from ontological ultimates. The step of constitution cross-checks the analysis to ensure that the atoms thus determined can actually account for the appearances of things. Vasubandhu opens the argument by setting up a destructive trilemma in 11. The putatively external physical object must either itself be (1) a unitary, unanalyzable whole, or a composite that can be analytically reduced to (2) a collection of isolated atoms (.samghata), as maintained by the Vai­bhasika, or (3) a conglomerate of atoms (pinda), as held by the Vaisesika. The cases (2) and (3) resolve into a dilemma between spatially differentiated and point-like atoms.

Vasubandhu will go about the argument by showing that if simples are held to be spatially differentiated, then the analysis is incomplete, and if they are claimed to be truly partless (i.e. point-like), then they are inca­pable of constituting phenomena. After (2) and (3) have been eliminated by this reductio, he will close by debunking position (1), the final horn of the trilemma.

3.2.2 Against Spatially Differentiated Simples

The argument in 12 is directed against the Vaisesika realist doctrine of minute but spatially extended spherical atoms that touch each other. Vasubandhu argues that it is impossible for spatially differentiated objects capable of direct contact with each other to be least parts (i.e. truly partless simples). This is because truly partless objects by definition cannot partially touch each other in space—they either touch completely or they do not touch at all. And they touch completely only if they precisely overlap each other, i.e. occupy the exact same location or spatial “receptacle.” Because spatially adjoining objects can only touch in parts, partless atoms cannot come into contact or be in connection (samyoga) with each other.

Mereological Argument This reading exploits the intuition of the meta­physical realist that space is absolute, which implies that objects do not modify the geometrical structure of the space that they occupy. Vasubandhu assumes that even if it were possible to superimpose several objects of exactly identical shape into the same space, that the size of the objects would remain unchanged by this operation. The Vaisesika in fact admits that his binary least whole (dvyanuka), is of exactly atomic size because the constituent atoms cannot touch each other in parts. But according to the Vaisesika, the dvyanuka formed in this manner is now a real parts-possessor—a whole over and above its parts. There are, however, two problems with this view. First of all, two parts are only enough to constitute joining in two spatial dimensions. In order to arrange spheres in a three dimensional cubic lattice (as is probably the Vaisesika proposal), a minimum of six parts is required, as each atom contacts six others. This does not pose a major difficulty for the Vaisesika, as he may easily revise his concept of least whole, and argue that it comprises six least parts rather than two.

Vasubandhu would of course argue that this revision doesn’t do anything solve the underlying issue. This leads us to the second problem, which is that mereological parts can only join by occupying the same spatial receptacle. Least wholes cannot contact each other in parts as this would entail the spatial overlap of the contacting parts—and if this were to occur, the six atoms surrounding the center atom in the cubic lattice would collapse into it. All composite wholes would have the same size as a single atom, which Vasubandhu holds to be absurd. This is the “standard interpretation” of the argument as presented e.g. in Gelugpa doxography.

Here we observe that Vasubandhu is controversially assuming that connec­tion (samyoga) entails overlap, which is not generally the case. Connection is a topological concept, whereas overlap is mereological.51 If the two concepts are taken to be equivalent, then topology amounts to a model of mereology. Vasubandhu, being a Buddhist ultra-nominalist, is arguing just this. The gravamen of his case against the realist is the objection that an individual substance qua ontological ultimate must be construed as a bare mereo­logical concept. Due to his Sautrantika roots, Vasubandhu has a critical concept of space and considers topological primitives such as connection and shape to be vikalpa. To him, these conceptual constructions are unreal and therefore not appropriate to the realm of ontology. Once we have accepted the ultra-nominalist premise that ultimate ontology is bare mereology—as Vasubandhu is urging—then his argument against atomism simply runs its course without further obstacles.

Since his two Buddhist opponents, the Vaibhasika and the Sautrantika are both committed to mereological nihilism, they have no other choice but to go along with the unwelcome consequences of this position that Vasubandhu is about to draw for them.

The Sautrantika position will be scrutinized later—for now, Vasubandhu is going to score points against the Kashmiri Vaibhasikas, who submit in XX that the mereological argument has no impact on their position.

The Vaibhasikas explain that because they do not maintain the existence of connections between the partless substance-atoms that constitute the aggregate-atom (we recall that according to their view, the substance-atoms are separated by interstices). They argue that only these aggregate-atoms (or cohesive “molecular” groups) are connected among themselves, since they have parts. Vasubandhu off-handedly dismisses this theory in 14 with the observation that the aggregate is a nominally designated mereological sum. Therefore, it cannot be a parts-possessor or be admitted to have any physical

properties (such as connection, covering, resistance and obstruction) that its individual atoms lack. Otherwise the aggregate-atom becomes a real conglomerate (pinda), which would be the position of the Vaisesika. Being an ultra-nominalist, the Vai­bhasika is forced to accept this consequence. Vasubandhu emphasizes that the aggregate-atom’s inability to connect is due its purely nominal status, not due to the partlessness of the substance-atoms—this is an issue that he deals separately later on in the argument.

The Vaisesika, on the other hand, is still very much alive and kicking. He is not willing to accept the idea that the atom is a bare mereological least part and retables an amended version of his proposal. Substance-atoms are not quite bare mereological least parts, he argues. They occupy receptacles and therefore have spatial extent. Two atoms join through mereological overlap (filling the same receptacle) to form a binary (dvyanuka).^ Topological connection is a property of integral wholes, not of least parts. The dvyanuka, being an integral whole, can now topologically connect as it pleases.

However, Vasubandhu evidently thinks there is something deeply wrong with the idea that a mereological least part can possess spatial extent. Thus far, we have examined his “mereological argument” against the Vaisesika. But later on in 14 and XXIII, he offers a more general “geometric argument” intended to knock down the notion that an atom extended in space (i.e. occu­pying a spatial receptacle) can be an indivisible unity (ekatva). Vasubandhu unfortunately does not specify the argument in depth. He says only this much—a unity in which there is a “division according to the sections of the space” is not logically possible. In the following, we will examine two possible readings for this argument—the Zenonian and the Kantian model.

Zenonian argument from complete divisibility The argument of complete divisibility (essentially a reformulated version of Zeno’s metrical paradox of extension), was first described by Aristotle, but later attributed to Zeno by Simplicius.

If a body is completely divisible, then it must be divisible into parts that do not have any magnitude (else these parts would have spatial magnitude and therefore be incompletely divided). A complete division leaves either extensionless point-parts or nothing whatsoever. Zeno concludes that if nothing whatsoever remains after the division of a body, then the body is an illusion—ex nihil nihilo fit. And if only extensionless points remain, the sum of their extensions is zero, and therefore the body will be unextended. By consequence, if in reality, the body is no larger than a single point-atom, the appearance of an extended body would be illusory.

Against this, the Vaisesika will convey that the indivisibility of the atom is a merely physical one. It applies only to its physical substance and not to its spatial extent. The thesis being stated here is that for an entity to be a mereological least part, it is necessary and sufficient for it to be physically indivisible. This counter-argument is reasonable enough to derail the Zenonian analysis, and it is not immediately obvious as to what might be wrong with it.

We can observe that by insisting on the necessity of a limit to physical division, the Vaisesika is maintaining a position that resembles Kant’s thesis of the second antinomy of pure reason, which now leads us to examine the second interpretive model for Vasubandhu’s geometric argument.

Kantian antithesis In the thesis of the second antinomy, Kant argues that composites are decomposable, and, more controversially, proposes the ne­cessity of mereological atomicity—decomposition must entail reducibility to minimal parts. Every real composite must be reducible to a bottom level of self-subsisting simples. For if the operation of decomposition were to possess the property of closure (i.e. every composite entity yields only further com­posite entities under decomposition), nothing would remain if all composite entities were to be analyzed away, which Kant claims to be absurd.

The antithesis of Kant’s second antinomy attempts to show that composites cannot have minimal parts. Kant’s case is based on the following premises— (1) composition is an external relation of substances and is possible only in space, (2) every part of a composite must occupy a sub-receptacle of the composite’s receptacle, (3) these sub-receptacles are themselves spaces. Therefore the receptacles of the simples are spatial. However, according to Kant, because space is external relation of substances, or rather, an inner mode of representation in which certain perceptions are connected to each other, a receptacle cannot be a space unless it contains a manifold or composite of substances in external relation to each other. But if the receptacle of a simple substance contains a composite, this implies the absurdity that the simple is a substantial composite. Furthermore, Kant thinks that it is inconsistent for one and the same space to have two distinct limit-concepts—the extensionless geometrical point and additionally, the physically indivisible but geometrically space­filling physical point.

Perhaps the Vaisesika position that the atom is “absolutely imperceptible” may be seen as an attempt to get a grip on this discrepancy by letting the physical extent of the atom converge against the spatial limit-concept. Supposing the atom is held to be geometrically divisible but physically indivisible, one may ask whether its shape is physical or geometrical. If it is both, then absurdly, the atom’s shape is divisible and indivisible. Finally, if the atom’s shape is merely geometrical, then the atom has turned out to be a synthesis of a substance with a geometric shape—and a simple, by definition, may not be a synthetic whole. Kant remarks that the antitheses of the antinomies are typical of empiri­cist views. If Vasubandhu’s argument is anything like the antithesis of the second antinomy—and I think it is—then it should be fair enough to characterize his position as a kind of empiricism.

We will not be able to establish here whether Kant’s attempt to prove that a substance cannot have both spatial extent and simplicity satisfactorily accomplishes the work cut out for it. Vasubandhu probably intuits—for reasons much similar to those of Kant—that there can be no such thing as a mereological least part that occupies a receptacle. Would he then also agree with the thesis of the second antinomy? On Kapstein’s reading, Vasubandhu argues both the thesis and the antithesis of Kant’s second antinomy: “Vasubandhu, on the other hand, aims to demonstrate that the atomic theory is both false and necessary.”

But Kapstein’s explanation of what Vasubandhu is up to here is muddled. Vasubandhu is not aiming to establish the necessity of the atomic theory—he is merely attempting to demonstrate that the realist’s atomic theory leads to antinomy. The “necessity” of the atomic theory is the realist’s position, not Vasubandhu’s. What of the possibility of “exotic space?” The realist need not assume that space is continuous—he may argue that it is discrete (as was first proposed by Zuse). Or he could argue that the metrical structure of three-dimensional Euclidean space is a property that only emerges at the scale of macroscopic objects. Continuity and metric could be macroscopic epiphenomena of some more exotic type of space at a microscopic level, e.g. a space that comes in discrete chunks, that has more than three dimensions, etc.—as is often assumed in modern theoretical physics. However, if the realist were to travel down this route, he would effec­tively be arguing that space has hidden properties not accessible to sensible intuition—a very slippery slope for a direct realist like the Vaisesika.

3.2.3 Against Point-like Atoms

The treatment of point-like atoms as a possible case is a hermeneutic interpo­lation to make the proof easier to follow. The terms “point” or “point-like” are not explicitly mentioned by Vasubandhu—the karikas and auto-commentary only discuss the atom qua “unity” (ekatva). This assumption is supported by the interpretation of Siderits, who also ventures that Vasubandhu is examining the case of point-like atoms.

Vasubandhu’s arguments in 14 are directed against the idea that phenom­enal appearances cannot be constituted from point-like atoms. He argues that atoms must be spatially differentiated in order to account for certain visual phenomena such as light and shadow or extended surfaces. Since Vasubandhu has already established that true atoms cannot not have spatial differentiation, the argument that point-like atoms cannot account for appearances completes the reductio against atoms.

Shadow Terminator Argument This argument is against the Vaisesika opponent, who wishes to dodge Vasubandhu’s arguments against spatially extended atoms by proposing point-like atoms in lieu of his original theory that atoms are infinitesimally sized spheres. Vasubandhu’s objection is predicated on the Vaisesika’s physicalistic belief that light is a kind of in­corporeal substance or luminous fluidum (unlike the understanding of the phenomenalizing Buddhist schools, who analyze light strictly as intensity of visual sensation or stimulus). The Vaisesika define shadow as absence of illumination.

Vasubandhu contends that for a matter-atom to be able to cast shadow on other objects, it must obstruct the fluidum, which it can only do if it is bisected by a shadow terminator (a line dividing the illuminated from the shadowed section of the atom). But a truly point-like matter-atom cannot have a light and a shadow side. It is incapable of obstructing the luminous fluidum, any more than a dam can hold back water without an upstream and a downstream face. The realist protests that even if individual atoms are uniformly light or shadow, this does not logically exclude the possibility that a conglomerate of atoms could be bisected with a shadow terminator, dividing uniformly light from uniformly dark atoms.

In response, Vasubandhu inquires whether is it being admitted that an agglomeration (pinda) is something other than the atoms themselves. Is the realist in other words willing to admit conglomerates of atoms into his foundational ontology? The realist objector is forced to answer that they can’t be admitted—to do so would defeat the whole point of having a foundational ontology. However, if the conglomerate is not admitted as an entity in its own right, then the realist’s objection is scuttled. How is it that the conglomerate as a whole can obstruct light if none of its atoms are able to? One would expect it to be totally transparent. Of course, this line of reasoning does not take into account the modern physical theory of point-like particles that can act at a distance by way of field effects. But this poses no problem for Vasubandhu because none of his realist opponents adhere to such a view.

Displacement Argument Here Vasubandhu argues that if atoms are inca­pable of obstructing each other, they lose the ability to spatially displace each other and collapse into the same spatial location when connected. Possibly, Vasubandhu is playing on the geometrical intuition that an extensionless (i.e. point-like) atom with no internal structure can make direct contact with another such atom if and only if they both occupy the exact same geometrical point. In this particular edge case, topological connection coincides with mereological overlap.

The displacement argument applies equally to the Vaisesika and the Vai­bhasika atomisms. The Vaibhasika’s aggregate-atoms cannot connect, nor can their non-obstructing partless atoms. The Vaibhasika of course agrees that atoms do not connect to each other. But if atoms are unconnected, individually partless and non-obstructing (i.e. point-like), the Vaibhasika is unable to account for how they might constitute physically resistant or visu­ally opaque surfaces in sum. He is reduced to abiding by the unconvincing hypothesis that individual substance-atoms have latent properties that are actualized only when they are perceived in collection. This is unsatisfactory because it is not admitted that an individual atom may be perceived in isolation—hence, the cases of latent properties and “insensible sensibilia” coincide. Talk of latent properties does nothing to alleviate the real problem at hand.

3.2.4 Against Integral Wholes

After Vasubandhu has thoroughly laid waste to the idea that atomism is compatible with a two-term theory of perception, the Vaisesika is back with a new proposal predicated instead on the integral whole (auayavin). He objects that since Vasubandhu’s reasoning against atoms and conglomerates has left the phenomenal characteristics (such as form-color) of the external sense-fields untouched, why not simply admit percepts (i.e., integral wholes) as external?

The Vaisesika contends that the whole is an object of valid knowledge. If the whole is known by any means of valid knowledge, it is useless to consider the alternatives as to how it exists in the parts. He pleads the common-sense realist standpoint that abstract speculation cannot override valid experience of unities. The perception of the unitary whole does not depend on the perception of atoms, he argues, although their existence can be inferred. We can already surmise what Vasubandhu thinks of this. In his auto-commentary to the Kosa, he has gone so far as to disparage the integral whole (avayaviri) theory of the Vaisesika as “infantile.”

Although he almost can’t be bothered with a counter-argument, Vasubandhu expands into a one-or-many reductio and immediately dis­misses the many case as having already been refuted. As for the case that phenomenal appearances themselves are integral wholes, Vasubandhu ar­gues that a real, self-subsisting unity ought to have uniform, non-conflicting properties. He then points out some of the absurdities that result from this idea. The core of the argument is contained in his axiomatic proposition: “For a concurrent apprehension and non-apprehension of the same thing isn’t logical.”

Here we need to keep in mind that Vasubandhu is speaking of the atom qua perceptual datum. By definition, an atomic perceptual datum must either be given in its entirety or not at all. If it is possible to see only part of an object (for example, an aspect of a tree), the object cannot be an atomic datum. But all gross physical phenomena, Vasubandhu points out, are only ever given in aspect. While we may say that we see a brown table, we will have to admit that there are parts of the table that we do not see. Chatterjee conjectures that Vasubandhu is making an argument for the ne­cessity of phenomenal pluralism to subjective epistemic knowing, as absolute unity would entail an absolute uniformity of experience without change or succession. Kochumuttom has it that Vasubandhu is attempting to refute Parmenidean monism.

While there is something to be said for these points, the following examples provided by Vasubandhu illustrate that his argument is mainly mereological in thrust. On the reading being proposed here, after already having argued against the substantial partless part, Vasubandhu is now demolishing the idea of the substantial partless whole. In the first example, Vasubandhu raises the objection that partless wholes contradict the experiential reality of incremental movement. If the earth, for instance, were an indivisible unity, it should not be possible to walk, because the ground would have no parts that could be traversed by the successive steps. For motion to be possible, there must be distance, which in turn entails spatial extent and directional parts.

Next, if entities are partless wholes, it should not be possible to grasp a whole in some parts but not in others. Yet clearly, we are touching the elephant even if we are only touching its trunk and not its tail.

Moreover, if it is supposed that partless wholes are differentiated only by their phenomenal characteristics and not by their spatial location, then it should not be possible to distinguish horses from elephants, or several horses from one. A separation of entities could not be accepted since everything would be spatially indistinct, i.e. occupying the same location. In the final example, which involves microscopic aquatic animals, Vasubandhu argues that entities are distinguished by size. Else if phe­nomenal characteristics were held to be independent of physical size, we should be able to see micro/-organisms with the naked eye. The distinction of phenomenal characteristics (i.e. description of sensory experience in a purely phenomenal language) is in and of itself insufficient to distinguish unities—non-phenomenal factors such as quantity, distance, size and time must be brought in. Such concepts are not reducible to phenomenal language. The phenomenal wholes given to experience have parts, Vasubandhu rea­sons, and if they are to be substantial, they must necessarily have least parts. But they do not—therefore he concludes that phenomena are insub­stantial, and that there is no more to a phenomenon than our awareness of it.

3.2.5 Against the Causal Theory of Perception

In 16, the argument is against the Sautrantika, who espouses a stripped- down version of the Vaibhasika externalism of mental intentionality. A men­tal representation is epistemically grounded by the collection of stimulus­atoms that causally generated it, the Sautrantika wishes to claim. He is proposing something of the sort that momentary (ksanika) point-like stimulus-atoms act in aggregate upon the sense bases to causally generate non-point-like sensations that assemble into representations. The time-gap argument for representationalism is premised on the idea that because there can be a delay between an observed event and one’s observation of it, one actually observes a representation of the event. But the time-gap argument only supports representationalism if physical objects are assumed to exist. If the external object is the probandum of the argument—as is the case for the Sautrantika strain of realism—it operates against representationalism instead of confirming it.

Vasubandhu points out that under a strict presentist account of atoms, externalism about intentional states breaks down completely. Because Sau­trantika doesn’t admit to external temporal continuants, by the time a mental representation arises, the stimuli that caused it have evanesced. By consequence, all representations of external objects come to be representa­tions of absistents—neither do they offer a solid ground of inference about reality, nor do they require the epistemic grounding that such an inference might provide. This line of reasoning echoes a Vaibhasika critique of the Sautrantika representationalism, to the effect that that inference is based on logical pervasion (vyapti) of the mark of inference and the inferred property. This pervasion being the ground of inference, it cannot itself be derived from inference but must be given in perception. However, since the Sautrantika denies direct perception of external objects, they can never be objects of inference either due to a lack of the necessary ground of inference.

Since the Sautrantika’s causal theory of perception invalidates his own thesis while establishing the position of Vasubandhu, the Sautrantika might as well dispense with his gratuitous metaphysical hypothesis of external objects. Russell once famously argued that naive realism, if true, is false; therefore it is false. Vasubandhu is driving home an analogous point about the Sautrantika position.

3.2.6 Resumee of the Argument

If there are no atoms, then ought Mind-Only not to maintain that phenomena are in principle infinitely divisible? Why is it is taken for granted by his realist opponents that infinite divisibility is a problem? Leibniz, for example, believed the idea that atoms mark an end to the subdivision of matter to be a dangerous error. David Lewis and others have speculated that the hypothesis that physical objects are “atomless gunk” ought not to be ruled out a priori. It is also conceivable that some (but not all) composite objects have a complete atomic decomposition.

It has been argued by some modern commentators that Vasubandhu’s demonstration shows not that phenomena are not compounded, but that they are compounded to an infinite degree, i.e. that phenomena are gunk. Following the principle of ex nihilo nihilo fit, since infinitely divisible phenom­ena lack a compositional base of elemental substances, they are insubstantial and therefore illusory. Vasubandhu’s line of reasoning may be construed as an inversion of Epicurus’ principle that indivisible minima are necessary if macroscopic bodies are not to be mere appearances. Since indivisible minima are logically impossible, Vasubandhu concludes that macroscopic bodies are in fact no more than appearances. Must phenomena have a compositional base in order to sustain realism? Or restating the question in broader terms, need a foundationalist realism entail reductionism and externalism? The Indian realists of Vasubandhu’s day seemed to think so. Would they even recognize a coherentist, non­reductionist, internalist epistemology as realism? To hazard a guess, perhaps they might recognize such a realism as Vijnanavada!

The argument against atomism requires a number of unstated assump­tions to function as intended, many of which are quite possibly false. The notion of physical divisibility, for example, may pose theoretical difficulties to Vasubandhu’s analysis, as it requires a concrete physical operation, as op­posed to a purely abstract or mental operation. The required energy to split a composite particle must be supplied by some particle interaction which physically destroys the original composite particle. This process cannot be compared to analytic decomposition, as once a composite particle has been physically split into its component parts, the composite that they were a part of no longer exists. Therefore, they can no longer be properly considered parts of that composite.

To give a blanket refutation of atomism, Vasubandhu would have to demon­strate that, generally speaking, all atomistic mereologies are either logically inconsistent or empirically useless, but of course his argument accomplishes no such thing. Also, it does not defeat gunk theory and theories such as those of modern particle physics involving point-particles with interactional force fields, nor does it damage ontologically non-absolute atomisms such as constitution from sensory minima, attributive atomism, instrumentalist atomism or logical atomism.

But Vasubandhu’s argument against atomism is nonetheless strong enough to handily dispatch any realist account which proposes that on­tologically absolute atoms are metaphysically constitutive of phenomenal experience. It pulls out the rug from the Vaisesika claim that their particles lack proper parts, and beats back the Vaibhasika notion of point-particle sensations that aggregate to form extended sensation-clusters that join with other clusters to constitute macroscopic phenomena. And, as I will argue in the following, the Vijnaptimatra viewpoint does not intend to damage onto­logically relativist forms of atomism—on the contrary, it is fully compatible with it.

From the Mind-Only point of view, parts and wholes are conceptual fictions (vikalpa), yet within within the dream-like epistemological fiction of ordinary experience, they are true. For unlike Cincinnatus C., the unfortunate protagonist of Nabokov’s Invitation to a Beheading, who is imprisoned and sentenced to death for the unutterable crime of “gnostic turpitude,” we generally do not care to entertain the idea that the world we are inhabiting is an unreal fiction (abhuta-parikalpa) generated by the dualizing mind. During the course of our discussion of the argument against atomism, we have seen how Vasubandhu parlays the Buddhist’s nominalist skepticism about part-whole realism into a full-scale assault on the Ding an sich. Little should we be surprised if the gnostic turpitude that Vasubandhu is so blatantly inciting reveals itself upon closer examination to be nothing other than good, solid empiricism.

3.3 Visions of Vijnaptimatra in Empiriocriticist Eyes

Swapping our philosophical horses in midstream, we now pass to an investi­gation of empiriocriticist views on the ontological status of the atom and the external object. To be sure, the aim of comparing Mind-Only with empirio- criticism is not to completely gloss over the differences between the two. Mach and Avenarius are not subjective idealists, and their critique of the synthetic a priori does not have an equivalent in Mind-Only (though there is a great deal of category critique to be found). The radical presentism and the strong mystical subtext of Mind-Only are absent in empiriocriticism. Nevertheless, a search for epistemological common ground won’t take long to strike paydirt.

We must be careful not to push the comparison with Mind-Only too far— Mach attempts to steer clear of the extreme of mentalism (ontological ideal­ism), whereas Vasubandhu arguably embraces it. But although Vasubandhu is planted more firmly in the idealist camp, his anti-atomistic epistemology is remarkably close to the sensation pluralism of Mach, a passionate critic of atomism. He is opposed to the Kantian Dinge an sich as well as to the metaphysical atomism of the physical theory of his day for very much the same reasons as Vasubandhu rejected the atom and the external object.

Mach attempts to gain the unstable epistemological middle ground be­tween idealism and realism. Sommer points out a small but important detail about Mach’s position: sensations are not intrinsically subjective, as is commonly assumed. To Mach, the elements of experience (the explananda) are the epistemically prior raw materials of the explanans, i.e. the con­struction of subjectivity and objectivity as they are given rise to by naive realism.

In Mach’s view, as opposed to the metaphysical Vaibhasika notion of the aggregate-atom (samghata-paramanu), sensation-nuclei are purely correl­ative conceptual constructs. The atom is a mental shorthand (Gedanken- symbol) for the covariance of functional relationships between the sensa­tional elements of different psychophysical spaces, i.e. the space of vision, of hearing and of tactile sensation. The notions of metaphysically absolute space, time and substance have no place in Mach’s epistemology. Substances and atoms are higher order “tensorial abstractions.” They are pragmatic ideas that help us to economically conceptualize the complexes, relations and connections of sensations. In the same vein, Mach maintains that scientific explanations, including the natural laws of physics, are merely economical descriptions of functional relation.

As Musil observes, Mach does not see any difference in kind between ex­planation and description, nor does he believe the veridicality of descriptions to involve any kind of metaphysical correspondence to reality. His theory of metaphysics and epistemology is grounded in evolutionary psychology. Descriptions acquire their truth-value in pragmatic use. Conceptualization does not stand apart from the process of functional relation, but rather, participates in it, by assisting the organism in selectively adapting to its environment.

Turning to Avenarius, Mach’s philosophical brother-in-arms, we find that he mobilizes a similarly instrumentalist view of atomism, framing the principle of economy as the cognitive rationale for the atomic concept. This principle is the cornerstone of Avenarius’ radically empiricist reformulation of Kantianism, and may be characterized as a philosophical generalization of the principle of least action in physics. The requirement for underlying unities that hold experience together is not an a priori in the Kantian sense, but a matter of descriptive parsimony: Due to the principle of economy, we have a need to think in terms of limiting concepts such as the atom, despite that its existence is not proven or transcendentally established.

Because analytical decomposition satisfies the need for compre­hension, the process of decomposition is mentally continued even where empirical evidence no longer supplies solid confirmation but only indicates the direction of analysis. This actually infinite process of mental analysis can only be brought to a conclusion by interrupting it at a certain point and declaring it to be com­pleted by some new idea at a more remote point left in the dark.

Aside from giving the impression of comprehension, these com­pleting concepts offer intellectual support from two sides, on the one hand, by positing that their referent is no longer composite but simple, thereby putting an end to the mental process of divi­sion, and on the other hand, by suggesting that the indivisible thing is immutable, thus pacifying the demand for a continual change of content. The simple is now the immutable, even though its accidental connections are subject to absolute change, such a fictitious terminating idea is embodied in the concept of the atom, inasmuch as the atom is considered to be unanalyzable and immutable.

According to Avenarius, the regressus of decomposition gratifies the de­mand for ontological closure, i.e. the cognitive need for fixed objects that are regulated by the structuring concepts of experience. But he is under no illusion that this procedure can ever truly be brought to to completion or be made to yield Dinge an sich—rather, it is simply cut off at some point to produce a conveniently fictitious ontology.

However, Avenarius is concerned that this approach might lead to reduc- tionism, which he eschews as being an unacceptable reification of atomist instrumentalism: “Thus this so-to-say atomistic individualization of sen­sations will remain an indispensable aid in creating the impression of com­prehension with regard to our life of sensations and concepts, but it must be warned against viewing the world as a kind of kaleidoscope composed of such mosaic pieces of sensation.”

Avenarius is justifiably suspicious of reductionism. Even if a reductionist program could be carried out, he would be left holding the bag of meta­physical commitment to phenomenalistic atoms, with all the pitfalls that this position entails. Like Vasubandhu, he does not believe that there are atoms given to pre-theoretical, common-sense intuition, or that the whole of experience naturally fractures into atoms along pre-determined breaking points. He characterizes the atom as a mere “Hiilfsbegriff”—a provisional concept of theory that satisfies the pragmatic need for intellectual comprehension, but he dares not urge a firm ontological commitment to it, lest his pure phenomenalism revert back to a pre-critical metaphysics. The Avenarian substance-atoms are theoretical constructs, much like the constitutional elements of Carnap’s Aufbau-. “Constitution theory and transcendental idealism agree that all objects of knowledge are constituted (in idealistic language, ‘generated by thought’), and that this ultimately applies to the basic elements of the constitution system.” For both Avenarius and Carnap, it is irrelevant as to whether or not material composition or epistemological constitution are grounded in a metaphysically absolute ontology. To their point of view, foundational on­tology is just as much a logical construct as the world that is built up from it. The similarity to the epistemological idealism of Mind-Only is striking: “Thus, the thesis of the intentionality of citta becomes displaced in the emerg­ing Yogacara philosophy by an emphasis upon the ‘phenomenalistic’ nature of objects. Objects are really dharma-constructs and representations (vijnapti), dependent upon the complex processes of citta for their appearance.”

The Avenarian empiriocriticism contends that atomism is valuable as a strictly instrumentalist hypothesis, whereas Mach believes even an instru­mentalist concept of the atom to be unnecessary to science. Avenarius’ rationale for the dichotomy of appearances and things-in-themselves is also remarkably similar to the Mind-Only doctrine that the dualizing mind (vijnana) is responsible for the superimposition of the conceptually con­structed (parikalpita) nature onto the flux of experience:

The development of thought eventually had to reach the point where it came to awareness that the so-called properties of the thing were in fact sensations of the sensing subject; thus, the prop­erties ceased to belong to the thing itself—as their designation would indicate—they departed their relationship of dependence from the thing and entered into a more idealistic relationship with the sensing subject. As a result, a difference had to emerge between the thing as it was for the sensing subject and the thing as it was in itself

In Mind-Only, the percept is held to be a noema generated by transcen­dental subjectivity. Thus we find the Avenarian “introjection,” or bifurcation between the strictly private percept and the public object, being carried to its most radical conclusion. The strategy of Mind-Only for eliminat­ing introjection is to wholly absorb the objective world into the subjective perception. The epistemological trajectory that begins with naive realism and progresses through critical realism (where the organizing forms of ex­perience are constituted by metempirical generative acts of the mind) is brought to maturity in the view that experience springs entirely from the transcendental subjectivity. Mach’s psycho-physical parallelism, which attempts to resolve the Ave­narian problem of introjection, is often cast as a paradigmatic example of neutral monism. But while Avenarius’ sensation monism is an explicit one, the picture with Mach is slightly more complex. He argues dualism against monism and vice versa in a Madhyamika-esque play to dialectically hold the elusive middle ground beyond both extremes.

Avenarius feels at home with the idea that epistemological construction is a pragmatic process of evolutionary adaptation to experience. In a metaphor, we build our sandcastles on shifting sand because we have to live somewhere and have no other building materials at hand. Avenarius argues that the conceptualizing of change and process requires that we posit fixed nuclei or absolutes as counterpoints to experiential flux, but regards these nuclei or fixed points as mere idealizations. In this respect, it is not unreasonable to characterize the Avenarian view as resembling a ontologically relativized Sautrantika epistemology.

Avenarius’ theory of the atom as an instrumental posit may be directly compared to Quine’s view: “Positing does not stop with macroscopic physical objects. Objects at the atomic level and beyond are posited to make the laws of macroscopic objects, and ultimately the laws of experience, simpler and more manageable... ” Though Quine has shed the sensationalist and phenomenalist trappings of empiriocriticism and early logical empiricism, in other respects, he is still holding to the course of the via positivistica. The empiriocriticist core of epistemological instrumentalism and fictionalism shines through, extending even to his views on physical objects. In a line that might have been taken straight from Vasubandhu’s playbook, Quine states that “The myth of physi­cal objects is epistemologically superior to most in that it has proved more efficacious than other myths as a device for working a manageable structure into the flux of experience.”

Admittedly, Quine is not making a direct argument against the transcen­dental object. But in cutting physical objects down to size as quasi-mythical mirages conjured by the theory-ladenness of common-sense realism, he is quite vigorously attacking the foundations of the metaphysical realist belief in objectivity. And if Quine is any judge, Vasubandhu’s anti-atomism—far from being an antiquated relic of early medieval Indian philosophy—is of considerable relevance to contemporary philosophy.4 The End of the End of Epistemology

I have a dream. (Martin Luther King)

Are THERE ANY PHILOSOPHICAL or ideo-historical conclusions of value that we can draw from the narrative sweep presented thus far? Possibly, our retracing of the evolutionary trajectory of Bud­dhist epistemology, beginning with phenomenalistic realism and leading to idealist and skeptical positions of increasing critical refinement and rarefi­cation may may add a mosaic piece or two to our picture of the historical development of occidental empiricism. To state the matter more boldly, the idea suggests itself that Buddhism affords to the modern mind the possibility of re-envisioning empiricism from the ground up, in the cloth of an alternate history of ideas. Anticipating the grand themes of the Western empiricist tradition by more than a millenium, the “empiricism of the East” presents us with a unique opportunity to re-experience our own philosophical genesis in a new and perhaps more humble light.

Against the realist’s objections, Vasubandhu proposes in 3 that the spatio­temporal determination of experience is as in a dream. Striking a chord across centuries and cultures, Mach dismisses the question of whether the world is real or whether we merely dream it as being devoid of scientific meaning. Epistemology is just as much an abstract inquiry into the condi­tions and possibilities of knowledge as a historical quest for philosophical beginnings—a trail that may lead us as readily to Mach’s 19th century Vienna as to the 4th century Purusapura of Vasubandhu, if we are prepared to follow it.

However, unlike the quest of Husserl in his Cartesian Meditations to regain the ground of first philosophy, a phenomenological beginning in transcen­dental intersubjectivity, the journey of Vasubandhu and Mach is more than anything else a voyage to epistemology’s end. Each in their own way, they both attempt to recast empirical knowledge in light of its ultimate meaning. To Mach, the end of epistemology is its pragmatic purpose as the critical voice of scientific inquiry. To Vasubandhu, the end is a mystical and soteriological one—the cessation of innate ignorance (avidya) sought for by “the path of practice leading to the cessation of the cosmos.”

4.1 Anti-Cartesian Meditations

Since we have not yet managed to touch upon any of the multifarous critiques that the Mind-Only standpoint of the Vimsatika has generated within Indian philosophy, I will round out the investigation with a brief recapitulation of some of the more interesting objections. Aside from the fairly predictable counter-arguments offered by the realist darsanas such as Nyaya-Vaisesika and Mimamsa, we find a number of critiques put forth by skeptical schools of thought such as Madhyamika or by other idealisms, notably Advaita. It is illuminating to take notice of the fact that in the pre-scientific Indo-Tibetan philosophical tradition, the most devastating critiques of Mind-Only are the skeptical attacks on transcendental subjectivity while the realist defenses of objectivity are comparatively weak. Ironically, the situation in the West is exactly the other way around, as it would seem that scientific progress has catapulted realism into the pole position.

Advaita The Advaita Vedanta epistemology of Samkara may be classi­fied as a species of absolute monism with realistic tendencies. Although it quite evidently qualifies as a strong strain of idealism, Advaita insists on a separation of the subject and object within the domain of phenomenal experience. Sinha observes that Samkara’s criticism of Mind-Only is not very thor­ough. Strangely, Samkara does not offer any direct counter-argument to the Mind-Only thesis that external objects cannot exist due to the impossi­bility of atoms, despite that he was almost certainly aware of Vasubandhu’s argument against atomism. This circumstance severely reduces the force of Samkara’s defense of objectivity.

The keystone of Samkara’s critique is the “phenomenological argument.” He points out that sensory perception establishes the existence of the exter­nal world, as we perceive external objects such as a table or a tree. This seems like Samkara is attempting a Moorean proof of the external object by appeal to common-sense evidence. But what is preventing us from having intentional states directed at some impossible object, e.g. a “square circle?” Given that it is entirely possible to think about absistents, the externality of a percept could very well be a sort of absistent property, such as the “squareness” of a circular patch of color. From the presence of an external world to intentional awareness, we cannot deduce that it exists, any more than we ontologically commits ourselves to the existence of a hare’s horns simply because we are able to think about them. In a Meinongian sense, it is perfectly reasonable to speak of the non-externality of a perceived object.

In another argument, Samkara attempts to turn the tables on the idealist, by asking: if one admits consciousness, why not external objects? The Mind- Only proponent’s assertion of the priority of inner over outer experience seems arbitrary, he argues.

While it is plausible to admit that one can represent in inner experience (i.e. dreams or hallucinations) objects that do not actually exist, it is not equally plausible to admit the converse hypothesis that there exist real objects not representable in inner experience. The Cittamatrin may therefore argue that the former simply expresses a familiar empirical fact, whereas the latter proposes a wildly metaphysical theory—the two cases are very different in nature and far from being symmetrically interchangeable, as Samkara would suggest.

Samkara trots out a number of further arguments against Mind-Only, none of which are especially memorable. He does, however, unfold a cri­tique of self-reflexive awareness (svasamvedana) very much resembling the Madhyamika argument put forth by Candrakirti: Just as the light shining from a lamp can illuminate other objects, but not itself, or the blade of a sword can only cut other objects, a self-reflexive awareness would entail the absurdity of something acting upon itself. Although Samkara’s critique of the Mind-Only position on objectivity fails to be convincing, his criticisms of transcendental subjectivity come a great deal closer to striking at the Achilles heel.

Madhyamika Tsongkhapa contends that while Vasubandhu’s argument against atomism is valid, it misses the mark. Essence or own-nature (sva- bhava) is the category that is in need of refutation, not substance {dravya). The refutation of atoms fails to knock out this vastly more crucial and elusive target, because atoms are only superficial cognitive misconceptions generated by metaphysical doctrines, whereas svabhava is an innate misconception deeply entrenched in pre-philosophical common-sense intuition. Since the concept of partless simples does not enter into our ordinary experience, which is after all informed by the pre-philosophical intuitions of common-sense realism, Tsongkhapa argues that while Vasubandhu’s refu­tation of atomism is all very well and good, its skeptical force is not strong enough to damage the object of the Madhyamika critique of substantialism. In modern terminology, perhaps Tsongkhapa’s distinction of dravya and sva­bhava may be read along similar lines to the distinction of the existence and the independence dimensions of realism. Evidently, he believes the indepen­dence dimension to be the more challenging target, and the soteriologically more rewarding one. We may note that nowhere in the post-Vasubandhu philosophy of Maha­yana are there any novel defenses of atomism to be found (with the possible exception of the late Sautrantika atomism), nor do we find any novel anti- atomist arguments. This conspicuous absence of further development in the issue might indicate that Vasubandhu’s victory over atomist realism was considered by Mahayanists to be so overwhelming that nothing needed to be added to it.

In the sixth chapter of the Madhyamakavatara, Candraklrti unleashes a scathing barrage of Madhyamika arguments against the the Mind-Only position. These may be roughly classified into four main groups: (1) the refutation of the nonexternality of sense-objects, (2) the failure of mental potentials to account for sense-experiences, (3) counter-arguments to the dream theory of experience and the ultimate existence of the mind, and (4) the refutation of apperception. This classic critique of Candraklrti is one of the philosophical crown jewels of the Tibetan Sarma traditions. It serves as a case in point for how the pure criticism of Madhyamika builds on the Mind-Only position, as the full development of its skeptical force requires the deployment of a sufficiently similar position as an argumentative foil. While the polemic phraseology of the debate evokes the impression that the two positions are very remote from each other, on closer examination, the intimate resemblance of the Mind-Only position to the Madhyamika view is what actually makes for the power of Candraklrti’s argument.

In the Tarkajvala (Blaze of Reasoning), the auto-commentary of his Madhyamakahrdayakarikas (Heart Verses of the Middle Way) the Madhya­mika scholar Bhavaviveka proposes an intriguing argument against the Mind-Only strategy of accepting self-reflexive apperception as a provisional antidote to substantialism. Bhavaviveka’s objection, which happens not to be among those presented by Candraklrti, runs as follows:

If you think that external objects actually do not exist, why con­sider them part of consciousness? If you think that someone first treats them as part of consciousness, then uses another argument, other than [the argument] that they are part of consciousness, to refute [the idea that they are part of consciousness], it would be better to stay away from the mud and not to touch it than to wash it away. ... It is as if a certain fool were to leave a clean road and enter an unclean, muddy river. Others might then ask him, “Why did you leave the clean road and enter the mud?” If he said, “So that I can wash it off,” the others would say, “You fool! If you have to wash it off, you should stay away from the mud and not touch it in the first place.”

Unpacking this argument along Kantian lines, one may inquire: why pos­tulate the self-reflexive awareness of percepts as an antidote (pratipaksa) if one thereby accomplishes no more than to saddle oneself with an un­necessary a priori commitment to transcendental apperception? Kant acknowledges that although the transcendental apperception expressed in the proposition “I think” (“Ich denke”) is an empirical intuition, the concept of “I” fails to be an empirical concept because it is never met with in empirical experience. In this light, the general tendency of the Madhyamika critique of self-reflexive awareness may be outlined as follows: Because the “I” is functionally equivalent to the transcendentality of apperception, it would be misleading to claim the possibility of an ego-less transcendental appercep­tion, as does the Mind-Only. And if self-reflexive awareness is not held to be transcendental, then it would have to be an empirical phenomenon like any other. It would be a contingent aspect of representations rather than the necessary condition of their possibility. But in this case, how can it be employed as an antidote for the reification of empirical phenomena, if the antidote itself is part of the what is being reified? A reified antidote is just what we don’t want. The initial pseudo-problem of objectivity will have merely been exchanged for another one—the problem of transcendental subjectivity.

Bhavaviveka’s argument expresses in a nutshell the general thrust Madhyamika “metacritique” of the Mind-Only epoche of objectivity. As a case of barking up the wrong philosophical tree, retreating up the episte­mological ladder into inner experience is a soteriologically useless maneuver. But perhaps Bhavaviveka’s dismissal is premature—for without an adequate containment strategy for the metaphysical “virulence” of objectivity, scientific realism so resoundingly carries the show that a modern Madhyamika-style critique of subjectivity and essence will never manage to get off the ground. Reified absolutes are poison to the Madhyamika view, yet as we will see, 20th century theoretical physics inescapably suggests them.

4.2 The Empire Strikes Back and the Return of the Empiricists

Briefly turning our attention once more to Western philosophy of science, the meteoric rise to power of Mach-inspired logical empiricism was not always favorably received. It provoked a backlash among many experimental scientists, who preferred the more familiar model of critical realism. Planck may be identified as an influential example of this “reactionary” tendency. He mounted a spirited defense of objectivity and absolutes against the instrumentalist and relativist tendencies of positivism: “That we do not create the external world for reasons of convenience, but rather, that it impresses itself upon us with elementary force is a point which, in our positivistically influenced day and age, may not remain an unspoken matter of course.”

This statement is an obvious gibe against Mach’s instrumentalist views on the status of the external world. Planck points out how the progress of physical science has enabled absolute values to be established for quantities that were previously believed to be definable only as relative differences, notably energy and entropy. In classical mechanics, only energy differences can be measured, and while energy is a conserved quantity, the idea of an absolute value of energy has no significance within the theory. However, Einstein’s theory of special relativity fixes an absolute value for the resting energy of an object (the famous equation E = me2). Similarly, even though the descriptive force of classical thermodynamics is only sufficient to account for entropy differences, the introduction of the quantum hypothesis permits an absolute value to be derived for the entropy of a system S = k log W (where ll r is the number of microstates corresponding to the macrostate of the system), as IE then comes to be discrete.

As Planck delineates, the development of physical theories proceeds from the differential to the integral and from relative quantities to absolutes. The absolutes of Newtonian space and time have been relativized, only to re-establish the absolute on a more fundamental level as the metric of space-time (the speed of light in vacuum c). Planck holds the principle of transcendental objectivity to be the founda­tion and raison d’etre of physics. Even though he concedes that physical theory will never be absolutely complete and certain, he optimistically be­lieves that it is possible to obtain asymptoptically close knowledge of absolute reality. In sum, Planck’s considerations pose a very serious challenge to Mach’s empiricism of radical relativity.

An empiricist counter-attack might be organized along two dimensions— (1) the observational underdetermination and (2) the semantic incomplete­ness of empirical theories. In (1), the Quine-Duhem thesis can be brought to bear against the notion that theories can asymptotically approximate reality. As for (2), Godelian reasoning can be deployed to show that since a theory cannot internally prove its own consistency, there is no absolute sense in which it can be said to have a model. A metatheoretically reflecting empiricism will come, on a metaphysical or epistemological level of reflection, to relativize the objectivity of physical absolutes to the theory they are posited by, just as it relativizes the con­cepts of analyticity, aprioricity (etc.) to some theoretical or metatheoretical language. The model empiricist may venture that the Planckian critical realist notion of absolute objectivity is no more intelligible than the idea of a universal metalanguage or an ultimate level of reflection.

4.3 Concluding Reflections

On any scale, the charting of a possible course from 4th century Indian idealism to contemporary philosophy of science would make for an immensely ambitious undertaking. In an effort to advance through the subject matter at a reasonable pace, I have largely glossed over many noteworthy and distinctive features of Vasubandhu’s thought. The Yogacara theories of karma and causality have received especially short shrift. The nature of the investigation has required a perhaps overly large amount of historical glossing and hermeneutic circling, necessarily at the expense of rigorous analytic reconstruction.

Despite that much of the precision inherent in the Buddhist philosophical terminology of Vasubandhu’s era has been lost in the mists of time, more effort could and should have been invested in clarifying the technical mean­ings of terms. By its very nature, comparative philosophy is a hermeneutic transaction in medias res. To be able to say anything whatsoever, one needs to keep one’s meanings fluid. Interpretation to a certain extent trades in suggestive ambiguities. But if meanings are too fluid, one becomes guilty of fitting the evidence to the theory. The deployment of a semi-formalized philosophical precision language is really the only answer to this dilemma, though in my defense, such a strongly structured approach would not have been appropriate to the intended scale of the investigation.

As far as possible, I have tried to compensate my lack of philological competence in the original languages of Indian philosophy with rational interpolation. When working with classical texts, this is really the only conceivable strength a philosopher can play to vis-a-vis an expert philologist. Regarding the aims and tasks of the investigation, I believe that I have taken a sufficiently broad and deep cross-section of the subject matter for me to be able to hope to have extracted an acceptable fraction of the philosophical import of Vasubandu’s arguments against atomism. Overall, I think that the goal to strike a balance between a presentation of the arguments in historical perspective and a demonstration of their relevance to contemporary epistemology has been accomplished.

Sarva mangalam

Bibliography

Anacker, Stefan, Seven Works of Vasubandhu: The Buddhist Psychological Doctor. New Delhi: Motilal Banarsidass, 1998.
Anuruddha; Bodhi, Bhikkhu, editor, Abhidhammatthasangaha: A comprehensive manual of Abhidhamma. BPS Pariyatti Editions, 2000.
Arnold, Dan, Madhyamaka. in: Internet Encyclopedia of Philosophy. (URL: http: / /www. iep. utm. edu) — accessed on April 12, 2008.
Asanga, Abhidarmasamuccaya. Asian Humanities Press, 2001.
Audi, Robert, editor, The Cambridge Dictionary of Philosophy. Cambridge University Press, 1994.
Avenarius, Richard, Kritik der Reinen Erfahrung. Volume I, Georg 01ms Verlag, 2004.
Avenarius, Richard, Kritik der Reinen Erfahrung. Volume II, Georg 01ms Verlag, 2004.
Avenarius, Richard, Philosophie als Denken der Welt gemaB dem Prinzip des kleinsten KraftmaBes. Edition Classic Verlag Dr. Muller, 2006.
Ayer, A. J., Phenomenalism, in: Philosophical Essays. Macmillan Press, 1954, 125-166.
Ayer, A. J., The Terminology of Sense-Data, in: Philosophical Essays. Macmillan Press, 1954, 66-104.
Ayer, A. J., Hume. Oxford University Press, 1980.
Barrow, John D., The Universe That Discovered Itself. Oxford University Press, 2000.
Bastow, David, The Maha-Vibhasa Arguments for Sarvastivada. Philosophy East and West, 44 1994:3, 489-499.
Becker, Martin, Zum Begriff der Genidentitat - Eine Untersuchung der Wissenschaftstheoretischen Schriften von Kurt Lewin. Master’s thesis, Johann Wolfgang Goethe University, Frankfurt am Main, 1998.
Berzin, Alexander, Basic Features of the Gelug-Chittamatra System. (URL: http://www.berzinarchives.com/sutra/sutra_level_ 5/basic_f eat_gelug_chittamatra_l. html) — accessed on April 11, 2008.
Bhikkhu, Thanissaro, MN 72. Online (URL: http://www. accesstoinsight.org/tipitaka/mn/mn.072.than.html)— accessed on June 23, 2008.
Bhikkhu, Thanissaro, Rohitassa Sutta - AN 4.45 (PTS: A ii 47). Online (URL: http://www.accesstoinsight.org/tipitaka/kn/ud/ud. 1.10. than. html) - accessed on June 21, 2008.
Bhikkhu, Thanissaro, Ud 1.10 Bahiya Sutta. (URL: http://www. accesstoinsight.org/tipitaka/kn/ud/ud.1.10.than.html)— accessed on June 21, 2008.
Bhikkhu, Thanissaro, Vajira Sutta [Please insert “PrerenderUnicode-" into preamble] SN 5.10. Online (URL:
http://www.accesstoinsight.org/tipitaka/sn/sn05/sn05. 010. bodh. html) - accessed on June 21, 2008.
Blumenthal, James, The Ornament of the Middle Way: A study of the Madhyamaka Thought of Santaraksita. Snow Lion Publications, 2004.
Bostrom, Nick, Are You Living in a Computer Simulation? Philosophical Quarterly, 53 2003, 243-255.
Brown, Jason, Microgenesis and Buddhism: The Concept of Momentariness. Philosophy East and West, 49 1999:3, 261-277.
Buescher, John B., Echoes from an Empty Sky - the Origins of the Buddhist Doctrine of the Two Truths. Snow Lion Publications, 2005.
Baumer, Bettina, Upanishaden: Befreiung zum Sein. Wilhelm Heyne Verlag, 1994.
Candrakirti, Introduction to the Middle Way: Candrakirti’s Madhyamakavatara, with commentary by Ju Mipham. Shambhala, 2002.
Carnap, Rudolf, Abriss der Logistik. Julius Springer, 1929.
Carnap, Rudolf, Der Logische Aufbau der Welt. Felix Meiner Verlag, 1998.
Chalmers, David, The Matrix as Metaphysics. (URL: http://consc. net/paper s/mat rix . html) — accessed on May 11, 2008.
Chatterjee, A. K., The Yogacara Idealism. Motilal Banarsidass, 1975.
Chinchore, Mangala R., Aniccata/Anityata - An Analysis of the Buddhist Opposition to Permanence/Stability and Alternative Foundation to Ontology and/or Anthropology. Sri Satguru Publications, 1995.
Cohen, Stewart, Justification and Truth. Philosophical Studies, 46 1984, 279-296.
Deutsch, Eliot, Advaita Vedanta: A Philosophical Reconstruction. University of Hawaii Press, 1969.
Dicker, Georges, Kant’s Refutation of Idealism. Nous, 42:1 2008, 80-108.
Dragonetti, Carmen and Tola, Fernando, Being as Consciousness. New Delhi: Motilal Banarsidass, 2004.
Durland, Karann, Hume’s First Principle, His Missing Shade, and His Distinctions of Reason. Hume Studies, XXII 1996:1, 105-121.
Eckel, Malcolm David, Jnanagarbha’s Commentary on the Distinction between the Two Truths: An Eighth-Century Handbook of Madhyamaka Philosophy. State University of New York Press, 1987.
Essler, Wilhelm K., Analytische Philosophie I. Alfred Kroner Verlag, 1972.
Essler, Wilhelm K.; Preyer, Gerhard, editor, Unser die Welt.
Humanities Online, 2001 (URL:
http://publikationen.ub.uni-frankfurt.de/volItexte/ 2007/4545/pdf/Essler_UDW_ccl .pdf) — accessed on June 26, 2008.
Essler, Wilhelm K., On Absoluteness and Relativity: Modern Philosophy in Ancient Times. Online, March 2007 (URL:
http://www.vfphil.de/pdf/essler-vortrag-delhi.pdf)— accessed on July 3, 2008.
Essler, Wilhelm K., Labude, Joachim and Ucsnay, Stefanie, Theorie und Erfahrung. Alber, 2000.
Essler, Wilhelm K. and Mamat, Ulrich, Die Philosophie des Buddhismus. Wissenschaftliche Buchgesellschaft, 2005.
Fenner, Peter G., Candrakirti’s refutation of Buddhist idealism. Philosophy East and West, 33 1983:3, 251-261.
Findlay, J. N., Kant and the Transcendental Object. Clarendon Press, 1981. Fogelin, Robert J., Pyrrhonian Reflections on Knowledge and Justification. Oxford University Press, 1994.
Frauwallner, Erich, Die Philosophie des Buddhismus. Akademie-Verlag, 1969.
Frauwallner, Erich, Die Entstehung der buddhistischen Systeme. in: Nachrichten der Akademie der Wissenschaften in Gottingen I. Philologisch-Historische Klasse. Vandenhoeck & Ruprecht, 1971, 115-127.
Frauwallner, Erich; Steinkellner, Ernst and Kidd, Sophie Francis, editors, Studies in Abhidharma Literature and the Origins of Buddhist Philosophical Systems. State University of New York Press,
1995.
Garfield, Jay L., The Fundamental Wisdom of the Middle Way: Nagarjuna’s Mulamadhyamakakarika. Oxford University Press, 1995.
Garfield, Jay L., Vasubandhu’s ’Treatise on the Three Natures’ translated from the Tibetan edition with a commentary. Asian Philosophy, 6 1997:2, 133-154.
Garfield, Jay L., Western idealism through Indian eyes: A Cittamatra reading of Berkeley, Kant and Schopenhauer. Sophia, 31 1998:1, 10-41.
Garfield, Jay L., Empty Words: Buddhist Philosophy and Cross-Cultural Interpretation. Oxford University Press, 2001.
Garfield, Jay L., Western Idealism through Indian Eyes: A Cittamatra Reading of berkeley, Kant, and Schopenhauer, in: Garfield, Jay L., editor: Empty Words: Buddhist Philosophy and Cross-Cultural Interpretation. Oxford University Press, 2001, 152-169.
Garfield, Jay L., Reductionism and Fictionalism: Comment on Siderits’ ’’Personal Identity and Buddhist Philosophy”. Online, 2008 (URL: http://www.smith.edu/philosophy/documents/ ReductionismandFictionalism.pdf) — accessed on July 7, 2008.
Goodman, Nelson, Ways of Worldmaking. Hackett Publishing Company, 1978.
Grier, Michelle, Kant’s Critique of Metaphysics, in: Zalta, Edward N., editor: The Stanford Encyclopedia of Philosophy. Spring 2007 (URL: http://plato.Stanford.edu/archives/spr2007/entries/ kant-metaphysics/) - accessed on July 3, 2008.
Grim, Patrick, The Incomplete Universe: Totality, Knowledge, and Truth. MIT Press, 1991.
Grupp, Jeffrey, The R-theory of Time, or Replacement Presentism: The Buddhist Philosophy of Time. Indian International Journal of Buddhist Studies, 6 2005, 51-122.
Grupp, Jeffrey, Mereological nihilism: quantum atomism and the impossibility of material constitution. Axiomathes, 16 2006, 245-386.
Griinbaum, Adolf, Modern Science and Zeno’s Paradoxes. George Allen and Unwin, 1968.
Guenther, Herbert V., Philosophy and Psychology in the Abhidharma. Shambhala Publications, 1976.
Guenther, Herbert V., Mentalism and Beyond in Buddhist Philosophy, in: Guenther, Herbert V., editor: Tibetan Buddhism in Western Perspective: Collected Articles of Herbert V. Guenther. Shambhala Publications, 1977.
Gupta, Bina, Cit Consciousness. Boston: Oxford University Press, 2003.
Guyer, Paul, The Postulates of Empirical Thinking in General and the Refutation of Idealism (A218/B265-A235/B294). in: Mohr, Georg and Willaschek, Marcus, editors: Immanuel Kant, Kritik der Reinen Vernunft. Akademie Verlag, 1998.
Harris, Ian Charles, The Continuity of Madhyamaka and Yogacara in Indian Mahayana Buddhism. E.J. Brill, 1991.
Hermes, Hans, Eine Axiomatisierung der allgemeinen Mechanik. Hirzel, 1938.
Hiriyanna, M., The Essentials of Indian Philosophy. Motilal Banarsidass, 1995.
Hoheisel, Claus, Physik und verwandte Wissenschaften in Robert Musils Roman ’Der Mann ohne Eigenschaften’. (dmoe) Ein Kommentar. Ph. D thesis, Ruhr-Universitat Bochum, 2003, (URL:
https://eldorado.uni-dortmund.de/dspace/bitstream/ 2003/2959/2/ hoheisel2 .pdf).
Hopkins, Jeffrey, Meditation on Emptiness. Boston: Wisdom Publications,
1996.
Hopkins, Jeffrey, Emptiness in the Mind-Only School of Buddhism: Dynamic Responses to Dzong-ka-ba’s The Essence of Eloquence: I. Boston: University of California Press, 1999.
Hopkins, Jeffrey, Maps of the Profound: Jamyang Shayba’s Great Exposition of Buddhist and Non-Buddhist Views on the Nature of Reality. Snow Lion Publications, 2004.
Hopkins, Jeffrey, Absorption in No External World: 170 Issues in Mind Only Buddhism (Dynamic Responses to Dzong-Ka-Ba’s the Essence of Eloquence). Snow Lion Publications, 2006.
Hossenfelder, Malte, Kants Konstitutionstheorie und die Transzendentale Deduktion. Walter de Gruyter, 1978.
Huemer, Michael, Sense-Data, in: Zalta, Edward N., editor: The Stanford Encyclopedia of Philosophy. Fall 2007 (URL: http:
//plato.stanford.edu/archives/fall2007/sense-data/)— accessed on July 6, 2008.
Huggett, Nick, Zeno’s Paradoxes, in: Zalta, Edward N., editor: The Stanford Encyclopedia of Philosophy. Winter 2006 (URL:
http://plato.Stanford.edu/archives/win2006/entries/ paradox-zeno/) - accessed on July 3, 2008.
Husserl, Edmund; Stroker, Elisabeth, editor, Cartesianische Meditationen: Eine Einleitung in die Phanomenologie. Felix Meiner Verlag, 1995.
Ingalls, Daniel H.H., Samkara’s Arguments Against the Buddhist. Philosophy East and West, 3 1954:4, 291-306.
Inwagen, Peter van, Material Beings. Cornell University Press, 1995.
Jain, L.C. and Patni, G.C., Exact Sciences from Jaina Sources Vol. 2: Astronomy and Cosmology. Rajasthan Prakrit Bharti Sansthan, Jaipur; Sitaram Bhartia Institute of Scientific Research, New Delhi, 1983.
Kalupahana, David J., A Buddhist Tract on Empiricism. Philosophy East and West, 19 1969:1, 65-67.
Kant, Immanuel; Weischedel, Wilhelm, editor, Kritik der Reinen Vernunft. Volume 2, Suhrkamp-Taschenbuch Wissenschaft, 1980.
Kapstein, Matthew, Mereological Considerations in Vasubandhu’s “Proof of Idealism”, in: Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought. Oxford University Press, 2001.
Karunadasa, Y., Time and Space: The Abhidhamma Perspective - Professor K. N. Jayatilleke Memorial Lecture 2003. 2003 (URL: http : / / www. lank a library. com/Bud/time . htm) — accessed on April 12, 2008.
King, Richard, Early Yogacara and its Relationship with the Madhyamaka School. Philosophy East & West October 1994.
King, Richard, Vijnaptimatrata and the Abhidharma Context of Early Yogacara. Asian Philosophy March 1998.
Klein, Anne C., Knowledge and Liberation: Tibetan Buddhist Epistemology in Support of Transformative Religious Experience. Snow Lion Publications, 1986.
Klein, Anne C., Knowing, Naming and Negation: A Sourcebook on Tibetan Sautrantika (Translations in Indo-Tibetan Buddhism). Snow Lion Publications, 1991.
Kochumuttom, Thomas A., Buddhist Doctrine of Experience: A New Translation and Interpretation of the Works of Vasubandhu. New Delhi: Motilal Banarsidass, 1989.
Lewin, Kurt, Der Begriff der Genese in Physik, Biologie und Entwicklungsgeschichte. in: Metraux, Alexandre, editor: Werkausgabe. Volume 2, Klett-Cotta, 1983, 47-318.
Lindstrom, Jeff, Imaginations of the Unreal: Modern Interpretations of Yogacara’s ’’Idealism”. 2002 (URL:
http://individual.utoronto.ca/jlindstrom/coursework/ yogacara_essay_2 0 01_12_19 . htm) — accessed on July 10, 2008.
Lusthaus, Dan, The Crux of the Yogacara Project. August 2004 (URL: http://www.bu.edu/religion/faculty/bios/Lusthaus/ yogacara\%20crux .pdf) — accessed on July 3, 2008.
Mabbett, Ian W., Nagarjuna and Deconstruction, in: Buddhisms and Deconstructions. Rowman and Littlefield, 2006.
Mach, Ernst, Die Analyse der Empfindungen und das Verhaltnis des Physischen zum Psychischen. Edition Classic Verlag Dr. Muller, 2006.
Matilal, Bimal K., Error and Truth - Classical Indian Theories. Philosophy East and West, 31 Marz 1981:2, 215-224.
Matilal, Bimal K., Perception: An Essay on Classical Indian Theories of Knowledge. Oxford University Press, 1986.
McDermott, A. Charlene, The Sautrantika Arguments against the Traikalyavada in the Light of the Contemporary Tense Revolution. Philosophy East and West, 24 1974:2, 193-200.
Mejor, Marek, Vasubandhu’s Abhidharmakosa and the Commentaries preserved in the Tanjur. Franz Steiner Verlag, 1991.
Murti, T. R. V., The Central Philosophy of Buddhism: A Study of the Madhyamika System. George Allen and Unwin, 1974.
Murzi, Mauro, Rudolf Carnap (1891-1970). in: Internet Encyclopedia of Philosophy. (URL: http: / /www. iep. utm. edu) - accessed on July 4, 2008.
Musil, Robert, Beitrag zur Beurteilung der Lehren Machs und Studien zur Technik und Psychotechnik. Rohwolt, 1980.
Nagao, Gadjin M.; Kawamura, Leslie and Nagao, Gadjin M., editors, Madhyamika and Yogacara: A Study of Mahayana Philosophies. SUNY Series in Buddhist Studies, 1991.
Narain, Harsh, The Madhyamika Mind. Motilal Banarsidass, 1997.
Oetke, Claus, ”Ich” und das Ich: Analytische Untersuchungen zur buddhistisch-brahmanischen Atmankontroverse. Franz Steiner Verlag, 1988.
Oetke, Claus; Steinkellner, Ernst, editor, Bemerkungen zur
Buddhistischen Doktrin der Momentanheit des Seienden: Dharmaklrtis Sattvanumana. Arbeitskreis fur Tibetische und Buddhistische Studien, 1993.
Olson, Robert F., Candrakirti’s Critique of Vijnanavada. Philosophy East and West, 24 1977:4, 405-411.
Panikkar, Raimon, Gottes Schweigen. Die Antwort des Buddha fur unsere Zeit. Fischer Taschenbuch Verlag, 1996.
Parfit, Derek, Reasons and Persons. Clarendon Press, 1984.
Pinker, Steven, How The Mind Works. W. W. Norton & Company, Inc.,
1997.
Planck, Max, Scheinprobleme der Wissenschaft. in: Vortrage - Reden - Erinnerungen. Springer-Verlag, 2001.
Planck, Max, Vom Relativen zum Absoluten. in: Vortrage - Reden - Erinnerungen. Springer-Verlag, 2001.
Potter, Karl H., editor, Encyclopaedia of Indian Philosophies: Buddhist Philosophy 100-350AD. Volume 8, Motilal Banarsidass, 2002.
Powers, John, Wisdom of Buddha: The Samdhinirmocana Sutra. Dharma Publishing, 1995.
Pyle, Andrew, Atomism and its Critics: From Democritus to Newton. Thoemmes Press, 1997.
Quine, Willard Van Orman, Two Dogmas of Empiricism, in: From a Logical Point of View: Nine Logico-Philosophical Essays. Harvard University Press, 1971, 20-46.
Ricard, Mathieu and Xuan Thuan, Trinh, The Quantum and the Lotus. Three Rivers Press, 2004.
Rosefeldt, Tobias, Das Logische Ich: Kant uber den Gehalt des Begriffs von sich selbst. Philo, 2000.
Rospatt, Alexander von, The Buddhist Doctrine of Momentariness. Franz Steiner Verlag, 1995.
z
Santaraksita and Mipham, Jamgon. The Adornment of the Middle Way.
Shambhala Publications, 2005.
Santina, Peter della, The Philosophy of Mind-Only, in: The Tree of Enlightenment. Online (URL:
http : / / www .ecst.csuchico. edu/ ~ ds ant ina/tree/) — accessed on April 12, 2008.
Sarao, K. T. S., Vasubandhu. in: Internet Encyclopedia of Philosophy. 2006 (URL: http: / /www. iep. utm. edu/) — accessed on April 12, 2008.
Schmauke, Stephan, Wohlthatigste Verirrung: Kants kosmologische Antinomien. Verlag Konigshausen & Neumann, 2002.
Sellars, Wilfrid, Kant’s Transcendental Idealism. Collections of Philosophy, 6 1976, 165-181 (URL:
http : / / www. di text. com/sellars/kti. html) — accessed on July 15, 2008.
Sharma, Chandradhar, A Critical Survey of Indian Philosophy. Motilal Banarsidass, 2000.
Sider, Ted, Van Inwagen and the Possibility of Gunk. Analysis, 53(4) 1993, 285-289.
Siderits, Mark, Nagarjuna as anti-realist. Journal of Indian Philosophy, 18 1988, 311-325.
Siderits, Mark, Replies to Garfield, Taber and Arnold. APA Newsletters 6 2006:1.
Siderits, Mark, Buddhism as Philosophy: An Introduction. Ashgate Publishing Limited, 2007.
Sinha, Jadunath, Indian Realism. Motilal Banarsidass, 1999.
Sommer, Manfred, Husserl und der Fruhe Positivismus. Vittorio Klostermann, 1985.
Sommer, Manfred, Evidenz im Augenblick: Eine Phanomenologie der Reinen Empfindung. Suhrkamp Verlag, 2001.
Stcherbatsky, Theodor, The Central Conception of Buddhism. New Delhi: Motilal Banarsidass, 1988.
Stubenberg, Leopold, Neutral Monism, in: Zalta, Edward N., editor: The Stanford Encyclopedia of Philosophy (Spring 2005 Edition). 2005 (URL: http://plato.stanford.edu/archives/spr2005/ entries/neutral-monism/) — accessed on July 19, 2008.
Tillemans, Tom J.F., Tibetan philosophy, in: Craig, E., editor: Routledge Encyclopedia of Philosophy. Routledge., 1998.
Unger, Peter, I do not Exist, in: MacDonald, G. F., editor: Perception and Identity: Essays Presented to A. J. Ayer with his Replies to them. Macmillan Press, 1979.
Unger, Peter, There are no ordinary things. Synthese, 41 1979:2, 117-154.
Varzi, Achille, Spatial Reasoning and Ontology: Parts, Wholes and Locations, in: Aiello, M., Pratt-Hartmann, I. and Benthem, J. van, editors: Handbook of Spatial Logics. Springer-Verlag, 2007 (URL:
http://www.Columbia.edu/~ av7 2/papers/Space_2 0 07.pdf) — accessed on June 28, 2008. - chapter 1.
Vasubandhu, Vimsatika Vijnaptimatratasiddhih. in: Ruben, Walter, editor: Texte der Indischen Philosophie. Volume 2, Akademie-Verlag, 1956, 366-383.
Von Wright, Georg Henrik, The Tree of Knowledge and Other Essays. Brill Academic Publishers, 1993.
Wallace, B. Allan, Choosing Reality: A Buddhist View of Physics and the Mind. Snow Lion Publications, 1996.
Wayman, Alex, The Yogacara Idealism. Philosophy East and West, 15 1965, 65-73.
Williams, Michael, Problems of Knowledge. Oxford University Press, 2001.
Williams, Paul, On the Abhidharma ontology. Journal of Indian Philosophy, 9 1981, 227-257.
Williams, Paul, The Reflexive Nature of Awareness. Motilal Banarsidass,
1998.
Wittgenstein, Ludwig; McGuinness, Brian and Schulte, Joachim, editors, Logisch-philosophische Abhandlung: kritische Edition = Tractatus logico-philosophicus. Suhrkamp Verlag, 1989.
Wolters, Gereon, Mach I, Mach II, Einstein und die Relativitatstheorie: Eine Falschung und ihre Folgen. Walter de Gruyter, 1987.
Wood, Thomas E., Mind Only: A Philosophical and Doctrinal Analysis of the Vijnanavada. Motilal Banarsidass, 1994.
Wood, Thomas E., Nagarjunian Disputations: A Philosophical Journey through an Indian Looking-Glass. University of Hawaii Press, 1995.
Zimmerman, D, Could Extended Objects Be Made Out Of Simple Parts? An Argument For ’’Atomless Gunk”. Philosophy and Phenomenological Research, 56 1996, 3-29.
Zotz, Volker, Geschichte der buddhistischen Philosophie. Rohwolt, 1996.
Zuse, Konrad, Rechnender Raum. Vieweg, 1969.Index of Subjects and Technical Terms

A
a priori , 57, 104, 106
synthetic, 70
absistent, 53, 101, 113 action
at a distance, 97 moral consequence of, 45
anagoge, 16, 77
antinomy, 24
anti-realism, 83 anumana, 54, 59, 63, 64 apeiron, 20 apoha, 37 apophaticism, 20
aporia, 20
apperception
refutation of, 115
ataraxia, 26
atomic lattice, 90
atomism
attributive, 69
epistemic, 31 explanations for, 42 instrumentalist, 103, 106-108 origins of, 39 phenomenalistic, 49, 87, 107 physical, 39, 87
stimulus theory of, 63
B
bahyartha, 15
Brentano's thesis, 53
C
category, 64
critique, 58, 104
mistake, 21
Cauchy sequence, 68 cessation, 26, 41, 112 classical
electrodynamics, 89 mechanics, 118 thermodynamics, 118 close-packing, 90 cogito, 24, 26 consciousness
appearance in, 63
crystal analogy, 83 lamp analogy, 114
Lichtenberg, 116
mud analogy, 115
privacy of, 52
constitution, 103, 108
noetic, 75
of experience, 71, 77, 80, 88
of space, 95 constitution theory, 76, 108 continuants, 61, 101 counterfactuals, 53
D
diagonalization, 24
Ding an sich, 32, 65, 77, 79, 80, 82, 107
displacement argument, 97 dissolution
of conceptual rigidities, 22
of the ego, 34
dualism
Cartesian, 63
E
E = me2, 118
effective action
causal vectors of, 64
discrete pulses of, 68 infinitive, 47
wet dream example, 8 Elementarerlebnis, 66, 116 elements
a /3 7... A B C... K L M..., 34, 105
four fundamental, 50 emergence, 56 empiricism, 13, 15, 24, 26, 28, 57 empiriocritical assumption, 108 empiriocriticism, 14, 15, 17, 27-29, 70, 72, 79, 104, 108, 110 energy, 118 epoche
Cartesian, 23, 29, 77, 79, 85, 88,
117
of subjectivity, 24 phenomenological, 23 positivistic, 20
evolutionary psychology, 106 experience
as in dreams, 7, 8, 11, 111 elements of, 35, 45 factors of, 58
ineffability of, 20, 61 intersubjective, 12 mosaic theory of, 37, 51, 107 of objectivity, 7
priority of inner, 113 projection of consciousness into,
34
psycho-physical complex of, 80 reduction of, 69
sensory core of, 55 stream of, 63, 66 subjective constitution of, 56 externalism, 44, 53, 81, 101
F
False Aspectarianism, 83, 84 fictionalism, 53
fire simile, 19 forest-dwelling seers, 12 foundationalism, 31
G
Gedankensymbol, 105 genidentity, 44
Gestalt, 65, 82
gnosis, 27
inversion of, 27, 34, 35 gnostic turpitude, 104 gunk, 102, 103
H
hallucination
argument from, 56 doctrine of collective, 12 hell-guardian argument, 9 homeomeria, 54 Hiilfsbegriff, 107 hyle, 82
I
illucutionary negation, 17 illusion
argument from, 55 transcendental, 9 immanent temporality, 61 incompleteness, 118 indetermination
hungry ghosts example, 8 infinitesimal, 62, 94, 96 inherence, 37, 41 internalism, 7, 9, 68 introjection, 70, 109
K
Kantian turn, 14, 57 karma, 26, 106, 119 Korperwelt, 35
L
ladder
of views, 28, 29, 117
Wittgenstein’s simile, 21 levels of reflection, 7, 23, 119 luminance, 84
M
magnitude
intensive, 50
minimal perceivable, 49
McTaggart argument, 46 mentalism, 105 meontology, 18, 22 mereological
argument, 89 atom, 2 overlap, 92 metalanguage, 119 metaphysics, 57, 75, 106 middle ground
of monism and dualism, 109 of realism and idealism, 23, 77, 105
monism
absolute, 31, 112
neutral, 109
Parmenidean, 99
Moorean proof, 9, 113 motion, 100, 118 mysticism, 27, 28
apophatic, 26
immanence, 27
myth
of physical objects, 110
of the given, 108
N
negation
apophatic, 25
tetralemmatic, 21 nihilism
Madhyamika, 21
mereological, 41, 85, 86, 91 nirvana, 19 noema, 72, 109 no-ownership theory, 36 noumenon, 65, 78
O
ontological relativism, 104 ontology
convenient fiction, 107 ultimate, 91
P
paradox
as stylistic device, 16
of metrical extension, 92 paratantra-svabhava, 73 particle physics, 103
point-particle problem, 89 perception
causal theory of, 54, 58, 68 error-theory of, 54
veil of, 32
perceptual datum, 65, 99 phenomenalism, 6, 14, 15, 23, 25,
27-30, 32, 33, 35, 37, 38, 52, 53, 67
philosophical poetry, 16 physical absolutes, 117, 119
physical objects, 30-33, 35, 38, 70 in disguise, 53
myth of, 110
positivism, 5, 21, 27, 42, 117
Prajnaparamita, 22
pratibhasa, 63, 70
pratipaksa, 29, 116
pratyaksa, 35, 59, 72
presentism, 58
replacement, 61
principle
of economy, 106
of Epicurus, 102
of falsification, 88
of least action, 106 pseudo-problem, 20, 56, 117 psycho-physical parallelism, 109 psychophysics, 42, 83
pure reason
antinomies of, 18 idea of, 78
Q
qualia, 27, 52, 83, 84 “quantifying into”, 52 quantum, 68, 118
quasianalysis, 66
Quine-Duhem thesis, 118
R
raft
Neurath’s simile, 77
Sakyamuni’s simile, 21
raw feel, 52
realism
critical, 32, 64, 109 deconstruction of, 30 direct, 57, 72 existence dimension of, 31 metaphysical, 7, 80 modal, 48 physical, 52 representational, 32 reality
absolute, 32, 118
criterion of, 47
mystical, 21
of human existence, 24 transcendent, 27 reality problem, 55 receptacle, 89, 92, 94 regressus, 88, 107 regulative, 8
of sensory intuition, 26 relation
external, 94
functional, 26 relativity theory, 90, 118 renormalization, 89 res extensa, 63
S
s = k log if, ns sarvam asti, 44 Scheinproblem, 21 sensibilia, 33, 35, 48, 49, 53, 98 shadow terminator, 97 shape, 39, 51, 55, 57, 61, 62, 89-91, 95
silence
about absolutes, 18 metaphysical, 20, 21 of the Buddha, 17, 22 skepticism
Cartesian, 13 solipsism, 11, 12, 85 space
absolute, 49, 105 as absence of obstruction, 49 as relation of substances, 94 Carnap scheme, 94
Euclidean, 90 exotic, 95 limit-concepts of, 94
Newtonian, 118 sunyata, 20, 22, 24 svabhava, 114 svalaksana, 66 svasamvedana, 64, 114
T
telepathy, 11 temporal
eliminativism, 47
time
critique of, 38 modal theory of, 48 motion picture analogy, 44 time-gap argument, 101 transcendental
apperception, 25, 64, 116 ego, 23, 25 idealism, 82, 108illusionism, 9, 88
object, 77 objectivity, 10, 15, 32, 71 reasoning, 63, 88 transcending gnosis, 11 True Aspectarianism, 82, 84, 88
U underdetermination, 118
V
via negativa, 24, 28 via positivistica, 28 vikalpa, 58, 65, 91
vyapti, 101
W world
distinction of ego and, 34 inner covering outer, 63 like a dream, 11 limits of, 25 path to the end of, 35Index of Personal Names

A
Anacker, Stefan, 3, 6, 44, 73, 86,
100
Anaximander, 20
Aristotle, 93
Augustine, 53
Avenarius, Richard, 14, 15, 27, 70,
106, 107, 109
Ayer, A.J., 49, 50, 53
B
Bahya Darucirya, 33, 34
Belden, Ngawang, 63
Berkeley, George, 15, 23, 73
Bhavaviveka, 115, 116
Boltzmann, Ludwig, 108
Bouwsma, O. K., 88
Brentano, Franz, 83
Buddhadeva, 45, 46
Bu-ston, 3
C
Candraklrti, 115
Carnap, Rudolf, 23, 27, 55, 66, 68,
76, 94, 108
Chatterjee, A. K., 31, 38, 58, 66, 67,
D
della Santina, Peter, 58
Democritus, 39
Dharmakirti, 37, 66, 67, 69, 72
Dharmapala, 72
Dharmasri, 3, 43, 44
Dharmatrai ta, 3, 45, 46, 51
Dignaiga, 66, 69
Dipakara, 51
E
Einstein, Albert, 90, 118
Epicurus, 64, 102
Essler, W. K., 7, 11
F
Fechner, Gustav Theodor, 42
Frauwallner, Erich, 4, 5, 38, 46, 48,
86
G
Garfield, Jay L., 61
Ghos.aka, 45, 46
Godel, Kurt, 119
Goodman, Nelson, 66
Grupp, Jeffrey, 38, 61, 66
Guenther, Herbert V., 13, 38�
H
Hume, David, 24, 25, 38, 48, 49
Husserl, Edmund, 12, 23, 25, 34,
61, 111
J
James, William, 26, 61
Johnson, Samuel, 9
K
Kan. ada, 39
Kant, Immanuel, 10, 14, 18, 21, 26,
38, 68, 82, 93, 94
Kapstein, Matthew, 36, 37, 86, 95,
100
King, Martin Luther, 111
Kochumuttom, Thomas, 6, 29, 73,
77-79, 86, 99
L
Lamotte, Etienne, 2
Leeuwenhoek, Antonie van, 100
Leibniz, Gottfried, 102
Lewis, David, 102
Lichtenberg, Georg, 25, 26, 116
Locke, John, 64
Lusthaus, Dan, 6, 73-77
M
Mach, Ernst, 9, 11, 14, 25-27, 38,
55, 71, 85, 108, 111
Mahavira, 42
Matilal, B. K., 14, 36, 38, 50, 54, 62,
80, 86, 104
Mauthner, Fritz, 26
McTaggart, J. M. E., 46
Meinong, Alexius, 53
Mipham, Ju, 80, 84
Musil, Robert, 106
N
Nabokov, Vladimir, 104 Nagao, Gadjin M., 4, 76, 80 Naigairjuna, 61, 118
Naigasena, 36
Neurath, Otto, 77
O
Oetke, Claus, 36, 67, 69
Ostwald, Wilhelm, 108
P
Panikkar, Raimon, 20
Paramai rtha, 3
Parfit, Derek, 24
Parmenides, 99
Pinker, Steven, 83
Planck, Max, 117-119
Plato, 36
Popper, Karl, 88
Prior, A. N., 52
Q
Quine, W. V. O., 69, 110
R
Rorty, Richard, 15
Russell, Bertrand, 102
S
{Sakyamuni, 14, 17-21, 24, 33
Samkara, 112-114
Sanghabhadra, 3, 4, 43, 50Santaraksita, 28, 80, 84
Schmithausen, Lambert, 86
Schopenhauer, Arthur, 14
Schrodinger, Erwin, 1
Sellars, Wilfrid, 61, 108
Siderits, Mark, 15, 21, 22, 36, 86, 96
Sinha, Jadunath, 53, 112
Stcherbatsky, Theodor, 38
Sthiramati, 72
T
Tsongkhapa, 81, 84, 114
V
Vasumitra, 45, 46, 48
Vatsyayana, 40
von Rospatt, Alexander, 38, 59, 60
W
Wayman, Alex, 6, 73, 74
Wittgenstein, Ludwig, 21, 30
Y
Yasomitra, 48, 54
Z
Zeno, 92, 93
Zuse, Konrad, 95Appendix

Zusammenfassung in deutscher Sprache

Kapitel 1 Die vorliegende Arbeit behandelt das Argument gegen den Ato- mismus aus den Vimsatika Vijnaptimatratasiddhi (Zwanzig Versen des Nur-Geistes), einer erkenntnistheoretischen Abhandlung des spatantiken indischen Philosophen Vasubandhu. Unter Atomismus wird hierbei durchge- hend die Lehre vom Aufbau der Welt aus kleinsten, nicht weiter zerlegbaren Wirklichkeitsbestandteilen verstanden, wie sie im Altertum in verschiedent- licher Form von griechischen wie auch von indischen Weltanschauungslehren vertreten wurde.

Das Argument Vasubandhus beriihrt insbesondere die atomistischen Leh- ren der Vaibhasika- und Sautrantika-Schulen der fruhbuddhistischen Scho- lastik. Daher wird zunachst ein Abriss der Erkenntnistheorien dieser beiden Schulen des Abhidharma gegeben, wobei besonderes Augenmerk auf die kritische Entwicklungslinie ihrer Kategorienlehren in Bezug auf Raum, Zeit und Substanz gelegt wird. Durch diese vorbereitende Darlegung wer- den sowohl die philosophischen Hintergriinde und Voraussetzungen der Argumentation Vasubandhus aufgedeckt als auch die vorzunehmende In­terpretation seiner anti-atomistischen Wirklichkeitsauffassung im Lich- te der empiriokritizistischen Uberwindung des Dings an sich untermau- ert. Die so gewonnene Sichtweise wird schlieBlich als Ausgangspunkt fur transzendentalphilosophische Betrachtungen im Sinne der buddhistischen Erkenntnistheorie genutzt, wie etwa fur Uberlegungen fiber den Status der transzendentalen Apperzeption, die Konstitution von Erfahrungserkenntnis und die Aussagekraft skeptischer Argumente. In der Einleitung werden zunachst einige Bemerkungen zur Person, zur Lebensgeschichte und zum philosophischen Schaffen Vasubandhus gemacht.

Es folgt anschlieBend ein Resumee der erkenntnistheoretisch relevanten Themen der Vimsatika, in dem Vasubandhus stark idealistisch eingefarbter AuBenweltskeptizismus in Grundzugen sichtbar gemacht wird. Die Haupt- these Vasubandhus, die er gleich zu Anfang der Vimsatika aufstellt und im Verlauf der Abhandlung gegen realistische Einwande zu verteidigen sucht, ist das Prinzip der Traumartigkeit der Erfahrungserkenntnis. Damit soil im Wesentlichen gesagt werden, dass - trotz des gegenteiligen Anscheins - in der Erfahrung nirgends die Erkenntnis eines erkenntnisunabhangigen Gegenstands moglich ist. Vasubandhu stellt insbesondere die Erkenntnis- moglichkeit einer metaphysisch gegebenen objektiven AuBenwelt in Abrede und schlagt vor, dass eine intersubjektive Konstitution der Erfahrungs- mannigfaltigkeit an dessen Stelle treten solle. Auf die im Resumee herausgestellten Ergebnisse aufbauend lasst sich nun begrunden, weshalb eine nahere Betrachtung der Argumente gegen den Atomismus von geistesgeschichtlichem und erkenntnistheoretischem Interesse ist. Ferner wird die zu verwendende hermeneutische Methode der Untersuchung - der interkulturelle Vergleich zum kritischen Realis- mus Kants und dem Empiriokritizismus seiner Nachfolger Avenarius und Mach - eingefuhrt. Durch den Vergleich sollen Bezugspunkte zur modernen Erkenntnistheorie und Wissenschaftsphilosophie geschaffen und die philo- sophische Relevanz der Ideen Vasubandhus klarer zur Geltung gebracht werden. Es werden dabei allerdings auch die methodischen Probleme eines solchen Deutungsansatzes berucksichtigt.

Um die Analyse der Atomismusargumente mit einer hermeneutischen Hintergrundstruktur zu unterfuttern, wird ein erster Bruckenschlag zwi- schen der buddhistischen und der Mach’schen Erkenntnistheorie gewagt. Das Schweigen des Buddhas Gautama Sakyamuni zu den vierzehn metaphy- sischen Fragen wird zur Mach’schen Auffassung des Scheinproblems in Be- zug gesetzt. Am so gedeuteten Mittelpunktereignis des Buddha-Schweigens wird die Hypothese abgelesen, dass der buddhistischen Erkenntnislehre empiristische und erkenntniskritische Momente zu Grunde liegen. Ganz nach Fapon des logischen Empirismus wird im Zuge des Reflektierens uber die Bedingungen und Moglichkeiten von Erfahrungserkenntnis Ein- sicht dartiber gewonnen, dass uber alles jenseits der Erkenntnisgrenzen Liegende geschwiegen werden muss. Es wird die These entwickelt, dass das Buddha-Schweigen in seinem positivistischen Deutungssinn mit einer Urteilsenthaltung uber das transzendental Objektive einhergeht, also unter die philosophische Figur der „Cartesischen epoche“ gefasst werden kann. Der erkenntnistheoretische Hohengewinn eines die Reflektion von Bedin- gung und Moglichkeit von Erfahrung erlaubenden Standpunktes geht mit einem Abwurf metaphysischen Ballasts einher, bzw. ist nur auf solchem cartesisch-skeptischen Wege zu erlangen.

Allerdings ist dabei noch zu beriicksichtigen, dass das Hauptmoment der buddhistischen Soteriologie in ihrer Lehre vom Nicht-Selbst (anatman) und von der Leerheit der Erfahrungselemente liegt. Auch hier sind Parallelen zu einschlagigen empiristischen Anschauungen (etwa zur Bundeltheorie der Person) gegeben. In seiner mahayanistischen Blute entwickelt sich der der Leerheitsbegriff jedoch zu einer negativen Mystik und verlasst damit das Terrain der Erkenntnistheorie. Die reine via negativa ist nun neben der als gfiltig erachteten positivistischen Deutung die tiefere und eigentlichere Auslegungsmoglichkeit des Buddha-Schweigens, und lasst sich philosophisch gesehen mit einer epoche der transzendentalen Subjektivitat in Deckung bringen.

Uber ein positivistisches Schweigen kann auf hoherer Reflektionsstufe noch nachgedacht und gesprochen werden, wahrend sich das negativ- mystische Schweigen jeder Moglichkeit eines reflektierenden Zugangs ent- zieht. Die epoche des Objekts wird aber dadurch nicht hinfallig - sie gewinnt dadurch im Gegenteil ihren rechten Sinn als Propadeutik. Im Kontext der mahayanistischen Erlbsungslehre wird das vermeintlich absolut existieren- de Erkenntnisobjekt der Alltagserfahrung (wie auch die Erkenntnis selbst und letztlich auch jedes Reflektieren fiber Bedingung und Moglichkeit von Erkenntnis) als instrumentell aufgefasst. Ziel und Ende von Erkenntnis, wie auch das eigentliche Wesen von Erfahrung, sind mystisch und ungreifbar.

Die in den Vimsatika von Vasubandhu entfaltete idealistische Erkenntnis­theorie ist demnach von einem negativ-mystischen Letztsinn geleitet. Dieser Gesichtspunkt muss bei der Deutung seines rational-argumentativen Vor- gehens in ebensolchem MaBe wie die philosophischen Hintergrtinde der abhidharmischen Scholastik Berficksichtigung finden. Obgleich in Grund- ziigen vorhanden, kommt das mystische Moment bei Mach ungleich weniger deutlich zum Vorschein als bei Vasubandhu. Als positivistisch denkender Naturwissenschaftler begniigt er sich damit, den antimetaphysischen Stand- punkt zu gewinnen. Von hier aus soil Physik auf eine Beschreibung des Funktionszusammenhangs der Empfindungen reduziert werden konnen. Dieser Mach’sche Phanomenalismus entspricht aber gerade jenem fruh- buddhistischen Boden, der durch den Realismus der buddhistischen Scho- lastik verloren gegangen war und dessen Zuruckgewinn von Vasubandhu verteidigt wird.

Kapitel 2 Solche weitgreifende Thesen verlangen natiirlich nach einer aus- fuhrlicheren Begrundung, die im Rahmen einer Begriffsklarung erfolgt. Die Kernbegriffe Realismus, Phanomenalismus und Nominalismus wer­den naher ausgeleuchtet. Der Briickenschlag von der buddhistischen zu einer radikal-empiristischen Erkenntnistheorie wird anhand einer Betrach- tung ausgewahlter fruhbuddhistischer Lehrreden weiter vorangetrieben, die Parallelisierung zur Ideenwelt Machs erhartet.

Die noch naher durchzufuhrende Darstellung vom Aufstieg und Fall des buddhistischen Sinnesdaten-Atomismus wird angerissen: Im Zuge einer fortschreitenden Zeitkritik wandelt sich innerhalb der Abhidharma-Schulen die Auffassung vom Objekt. Wo sich in der Vaibhasika-Schule das auBen- weltlich Objektive progressiv als begriffliche Beifugung zum letztlich sei- enden Empfindungsmosaik konstituiert, wird es bei den Sautrantika in regressiver Weise gewonnen, d.h. als apriorisch zu denkender, letztlich sei- ender Kausalgrund von Empfindung (die nurmehr als in bloB begrifflich- beigefugter Weise seiend gilt) gesetzt. Diese philosophische Bewegung soil nun anhand modellhafter Rekonstruktionen der Erkenntnistheorien der Vai­bhasika- und der Sautrantika-Schulen eine weitere Untersuchung erfahren.

Zunachst werden jedoch die historischen Urspriinge des indischen Atomis- mus in der nicht-buddhistischen Vaisesika-Schule gesucht und die Atom- theorie dieser Schule in Grundziigen skizziert - sowohl um eine Verstandnis- grundlage fur die im Folgenden nachzuzeichnenden Entwicklungen der buddhistischen Schulen an die Hand zu geben als auch um die zentrale Kontrastfigur fur Vasubandhus anti-atomistische Argumentation aufzu- stellen. Der Einfluss der realistischen Vaisesika-Metaphysik stellt einen moglichen Grund fur die Entfaltung atomistischer Lehren innerhalb der ur- spriinglich nicht-atomistischen Erkenntnistheorie des Buddhismus dar, aber es sind auch Herleitungen aus phanomenalistischen oder nominalistischen Gesichtspunkten denkbar.

Im Atomismus der Vaibhasika-Schule sind realistische, phanomenalistische und nominalistische Einfliisse in mehr oder minder starkem AusmaBe zur Auspragung gekommen. Lehrmeinung des Vaibhasika ist eine zweigliedrige Wahrnehmungstheorie. Demnach entsteht Erkenntnis im unmittelba- ren Kontakt der Bewusstseinsmomente zu den Empfindungsatomen, die als letztlich seiende Substanzen im absoluten Raum aufgefasst werden. Diese Atomlehre ist im Rahmen enzyklopadischer Systematisierungen des Abhidharma dargestellt worden und teilt sich anhand dessen in drei Entwicklungsphasen auf. Die dritte, abschlieBende Stufe dieser Entwicklung ist in Vasubandhus Abhidharmakosa sowie im Nyayanusara, der Replik seines Lehrers und Kontrahenten Sarighabhadra, gegeben.

Die Vaibhasikas vertraten eine Lehre von der Existenz in Vergangenheit, Gegenwart und Zukunft, und entwickelten vier Zeittheorien, Diese Theorien sowie ihre Kritiken werden etwas ausfuhrlicher dargestellt, wobei Parallelen zur modernen westlichen Philosophie gezogen werden, etwa zur Zeitkritik McTaggarts. Es wird eine Rekonstruktion der Zeittheorie des Vasumitra vorgenommen, welche versucht, Zeitlichkeit als Modus von Aktualitat und Existenz in Vergangenheit und Zukunft mit einem modalen Realismus zu fassen. Die Theorie der Vaibhasika von den Empfindungsatomen wird ausgefuhrt und mit dem Hume’schen Sinnesdaten-Atomismus verglichen. Es folgt eine kritische Betrachtung der dargestellten Lehren, die zum Standpunkt der fruhen Sautrantika-Schule iiberleitet.

Die Anschauungen der Sautrantikas zeichnen sich gegentiber den Theo­rien der Vaibhasikas im Wesentlichen durch zwei Neuerungen aus - den Reprasentationalismus sowie die Lehre von der Momentanheit des Seien- den. Darunter wird der Ineinsfall von Zeitlichkeit und Kausalfunktion mit, dem Wesen des Seienden verstanden. Da die Sautrantika sogar den augen- blicklichen zeitlichen Bestand ablehnen, entsteht damit eine Auffassung von Atomen als augenblicklich-flusshafte Reizimpulse.

Die Sautrantika vertreten eine Korrespondenztheorie der Erkenntnis - eine wahrheitsgemaBe Empfindung ist demnach eine solche, die mit ihrer letztlich seienden, augenblicklich-flusshaften Reizgrundlage ubereinstimmt. Die Reizatome werden den Kant’schen Dingen an sich gegeniibergestellt und dieser Vergleich wird problematisiert. Die Sautrantika-Lehre der irrealen Wahrnehmungssynthese durch begriffliche Einbildungskraft wird beleuchtet, um den Unterscheid zum Synthesebegriff Kants zu unterstreichen. Die Atomlehre der spaten Sautrantika-Phase, in der nur noch von logisch partikularen Augenblicklichkeitspunkten die Rede ist, wird im Ansatz vorge- stellt und mit dem Elementarerlebnis Carnaps verglichen. Die Theorie der Augenblicklichkeitspunkte ist nicht von der Atomkritik Vasubandhus betrof- fen, da sie sich von der Vorstellung der Substanzhaftigkeit und raumlichen Ausdehnung des Atoms gelost hat.

Es folgt eine Zusammenfassung und kritische Wiirdigung der Erkenntnis- theorie der Sautrantika-Schule. Es werden Unstimmigkeiten an ihrer re- prasentationalistischen Erkenntnistheorie aufgezeigt, die auf die Notwen- digkeit einer ganzlichen Aufgabe des auBenweltlichen Objekts (wie dies von der Nur-Geist-Schule schlieBlich vollzogen wird) hindeuten. Kapitel 3 Die idealistische Lesart der Nur-Geist-Lehre wird zu Anfang programmatisch in den Vordergrund gestellt. Die Betrachtung geht zu ei­ner Unterscheidung der wichtigsten historischen Entwicklungslinien der Nur-Geist-Schule uber. Als nachstes wird die revisionistische Linie der mo­demen Rezeptionsgeschichte im Groben dargestellt. Es folgt eine nahere Erlauterung der drei einflussreichsten revisionistischen Lesarten. Diese deuten die Erkenntnistheorie der Nur-Geist-Schule im Wesentlichen als phanomenologische Spielart des kritischen Realismus. Die Stellungnahme zu diesen Lesarten arbeitet die eigene Interpretationslinie klarer heraus und untermauert diese. AnschlieBend wird die traditionelle Interpretation der Nur-Geist-Lehre vorgestellt. Die in der tibetischen Doxographie iibliche Unterscheidung der Wahr- und Falschaspektler-Lesarten wird eingefuhrt und im Ansatz phi- losophisch rekonstruiert. Die Wahraspektler-Lesart wird als „realistische“, die Falschaspektler-Lesart als „antirealistische“ Deklination des Nur-Geist- Standpunktes umschrieben.

Nachdem der Interpretationsraum aufgespannt wurde, wird das Argument gegen den Atomismus schematisch umrissen. Weil sich die Angriffslinie der Zeitkritik am Atombegriff im Sautrantika erschopft hat, versucht Vasub­andhu nun, das Atom und damit den Substanzbegriff auf dem Wege der Raumkritik ganzlich zu eliminieren. Vasubandhu vertritt die These, dass Atome weder in der Erfahrung selbst noch als ihre materiellen Konstitutions- elemente gegeben sind. Die realistische Gegenposition stellt sich in den drei Atomtheorien der Vaisesika-, der Vaibhasika- und der Sautrantika-Schulen dar. Die Argumentation verlauft apagogisch, da die Position Vasubandhus eine epoche ist und deshalb nur die realistische Gegenposition einen Nach- weis schuldet. Gelingt dieser nicht, ist damit der Standpunkt Vasubandhus erwiesen.

Der Realist habe laut Vasubandhu drei Moglichkeiten, das auBenweltliche Objekt nachzuweisen: als einheitliches Ganzes, als loser Atomhaufen oder als zusammenhangendes Atomkonglomerat. Wird wie in den letzten beiden Fallen des Trilemmas eine Atomtheorie angenommen, sind abermals zwei Faile zu unterscheiden: das Substanzatom als raumlich ausgedehntes Teil- chen oder als ausdehnungsloser Punkt. Vasubandhu wird nun zeigen, dass beide Faile unmoglich sind, womit nur noch der Fall des auBen weltlichen Objekts als einheitliches Ganzes zu widerlegen bleibt.

Der Fall des Atoms als raumlich ausgedehntes Teilchen wird mit einer mereologisch ausgerichteten Beweisfuhrung behandelt. Wesentlicher Aspekt der Argumentation ist die implizite Annahme, dass die raumliche Verbin­dung und Anordnung von Atomen nur mit mereotopologischen Mitteln darstellbar sind, Ontologie sich jedoch auf den rein mereologischen Anted beschranken muss. Topologische Eigenschaften sind als begriffliche Ein- bildungen bei der Betrachtung des letztlich Seienden auBer Acht zu lassen. Damit fiihrt Vasubandhu eine von der Sache her analoge Argumentation zur Antithese der zweiten Antinomie der reinen Vernunft bei Kant - er erweist, dass alles, was einen Raum einnimmt, zusammengesetzt ist. Gegen das punktformige Atom tragt Vasubandhu ebenfalls vernichtende Einwande vor, etwa dass sich solche Atome nur beruhren konnen, wenn sie den gleichen Raumort einnehmen. Ausgedehnte Korper, die aus solchen Atomen zusammengesetzt sind, fallen auf PunktgroBe zusammen. Des Wei- teren sind phanomenal wahrnehmbare Eigenschaften wie Verdrangung oder Schattenwurf nur unter Zugrundelegung einer raumlichen Ausdehnung des Atoms konstituierbar. Vasubandhu geht daraufhin zur Betrachtung des Falls tiber, dass das auBenweltliche Objekt als einheitliches Ganzes gegeben sei. Er fuhrt hierzu eine Reihe von Gegenbeispielen an, die wieder in Ahnlich- keit zum Kant’schen Standpunkt zeigen sollen, dass echte Einheiten in der Erfahrungsmannigfaltigkeit nirgends vorzufinden sind.

Im nachsten Argument wird die kausale Wahrnehmungstheorie der Sau­trantika abgehandelt. Vasubandhu weist darauf hin, dass eine Kausa- litatsbeziehung nur zwischen aufeinanderfolgenden Momenten moglich ist. Weil der Anhaltspunkt einer Wahrnehmung zu dem Zeitpunkt, als dar- aus ein Wahrnehmungsobjekt entsteht, bereits entschwunden ist, kann die Korrespondenztheorie der Erkenntnis nicht aufrecht erhalten werden. Damit ist dann auch die Notwendigkeit einer apriorisch zu denkenden auBen- weltlichen Reizgrundlage hinfallig geworden.

Atomtheorien, die als im Vasubandhu’schen Sinne widerlegt gelten durfen, sind solche, die von der erkenntnisunabhangigen Existenz unteilbarer, letzt- lich seiender Substanzen im absoluten Raum ausgehen. Atomismen, die sich nicht auf einen solchen Substanzbegriff festlegen, bleiben von seiner Argumentation unberiihrt, ebenso logische und instrumentelle Atomismen sowie realistische Standpunkte, die von einer unendlichen Teilbarkeit des StofElichen ausgehen.

Den Vergleichsfaden zur empiriokritizistischen Erkenntnistheorie wieder aufgreifend kann der Standpunkt Vasubandhus nun anhand des Mach’schen Antiatomismus neu ausbuchstabiert werden. Mach und Avenarius fiihren wissenschaftliche Erklarung auf Beschreibung, Ontologie auf das Prinzip der Denkokonomie zuriick. In der Mach’schen Erkenntnistheorie gerat das Atom zum bloBen Gedankensymbol, bei Avenarius ist es ein aus Zweck- maBigkeitsgrunden gewahlter Grenzbegriff. Fur die Empiriokritizisten ist die bewusstseinsunabhangige Gegen standlichkeit des Erkenntnisobjekts genau wie fur Vasubandhu eine Fiktion. Ahnliche Standpunkte finden sich bei Carnap und sogar bei Quine, so dass der Idealismus Vasubandhus vom Grundsatz her keineswegs als ganzlich veraltet gelten kann.

Kapitel 4 Die vorangegangenen Erorterungen legen nahe, die Vasuband- hu’sche Sicht als einen Empirismus in fremdem Gewande zu deuten. Der durchgefuhrte Perspektivwechsel zur buddhistischen Erkenntnistheorie ver- schafft dem abendlandischen Philosophierenden einen neuartigen Ausblick auf seine eigene empiristische Denktradition. Die Suche nach Anfangs- griinden und Sinn von Erkenntnis gestaltet sich gleichermaflen im Abstrak- ten als erkenntnistheoretische Reflektion wie auch im Konkreten als geistes- geschichtliche Reise zu den ost-westlichen Urquellen der Erkenntnistheorie.

Um weitere Perspektiven der erfolgten Analyse zu eroffnen, wird die Kritik an der Nur-Geist-Sichtweise vonseiten der Advaita Vedanta- und Madhya- mika-Schulen angerissen. Die Nur-Geist-Lehre dient dabei als notwendige Gegenfigur fur die voile Entfaltung des metakritischen Skeptizismus der Madhyamikas, der sich den Angriff auf die transzendentale Subjektivitat zum Programm macht. Andererseits geht der Madhyamika in seiner De- struktion der Objektivitatskritik der Nur-Geist-Schule deutlich zu weit. Es wird die These aufgestellt, dass Kritik am Objektivitatsbegriff Bedingung der Moglichkeit einer Subjektivitatskritik ist. Hat man namlich keine Mittel zur Hand, sich des absoluten Charakters gewisser physikalischer Grundbegriffe zu erwehren, ist jeder Ansatz einer Wesenslosigkeitsmystik im Sinne des Madhyamika im Keime erstickt.

Planck bringt auf den Punkt, wie aufgrund theoretischer Fortschritte (beispielsweise der speziellen Relativitatstheorie und der Quantenhypothe- se) entgegen den fruhpositivistischen Ansichten Machs tatsachlich auf die Absolutwertigkeit u.a. von Energie und Entropie geschlossen werden kann. Die Mach’schen Prinzipien der Relativitat physikalischer Begriffe und der Instrumentality der AuBenwelthypothese seien durch die gezeigte Bewe- gung vom Relativen zum Absoluten entkraftet. Der Empirismus findet jedoch in der metatheoretischen Reflektion einen Ausweg. Durch absolutwertige Begriffe ist einerseits ein wissenschaftlicher Bodengewinn gegeben, anderer- seits kommt man aus erkenntnistheoretischen Grunden nicht umhin, diese Absolutheit wieder in Bezug auf Theorie und Reflektionsstufe zu relativieren.

AbschlieBend folgt eine kurze Selbstbeurteilung des Geleisteten im Hin- blick auf die Aufgabenstellung der Arbeit und den bearbeiteten Stoff.Lebenslauf Personliche Daten

Name: Bruhacs, Lorand

Anschrift: Eichwaldstr. 56, 60385 Frankfurt a.M.

Telefon: 069 / 46939002

Email-Adresse: nahuaque@gmail.com

Familienstand: ledig, 1 Kind

Staatsangehorigkeit: USA

Geburtsdaten: 21. Oktober 1975 in Wiesbaden

Schulische Ausbildung/Studium

1995 Gutenbergschule / Wiesbaden Abschluss: Allgemeine Hochschulreife

1995— Studium an der Johann Wolfgang Goethe Universitat / Frankfurt a.M.

Fachrichtung: Philosophie (M.A.)

1996— Fachrichtung: Informatik (Dipl.)

Berufliche Erfahrungen

1996 EnterpriseFlorida / Frankfurt a.M. Tatigkeit: Praktikant 1997 Institut fur Neue Medien Tatigkeit: Praktikant 1997—1998 Johann Wolfgang Goethe Universitat / Frankfurt a.M.

Tatigkeit: studentische Hilfskraft 1998—2002 IP23 GmbH /Frankfurt a.M. Tatigkeit: Internet Engineer / Technical Writer 2003— Deutsche Nationalbibliothek / Frankfurt a.M. Tatigkeit: Systemverwaltung Unix/Netzwerke Frankfurt a.M., den 22. Juli 2008 Eidesstattliche Erklarung

Hiermit versichere ich, dass vorliegende Arbeit selbstandig verfaBt und keine anderen als die angegebenen Hilfsmittel benutzt sowie die Stellen der Arbeit, die anderen Werken dem Wortlaut oder dem Sinn nach entnommen sind, dur ch Angabe der Quellen kenntlich gemacht wurden. Ferner versichere ich, dass die Arbeit in gleicher oder ahnlicher Fassung noch nicht Bestandteil einer Studien- oder Prufungsleistung war. Frankfurt a.M., den 22. Juli 2008

Lorand Bruhacs