Classic and Contemporary Buddhist Works

Foundations of Dharmakīrti's Philosophy - Introduction


Buddhist philosophers often speak of beginninglessness. It is claimed that the minds of living beings, for example, have no beginning, and that our current universe is only one in a beginningless cycle of expansion and decay. Some Buddhist thinkers would claim that even the most mundane task can have no true beginning. That is, if a beginning occurs, there must be some moment, some “now,” in which it occurs. For the present to exist, however, there must be a past and a future, for what would “now” mean if there were no time other than now? And of course, if there is a past, then how could now be a beginning? Now should instead be the end of the past. Each beginning, in short, must itself have a beginning.In a more concrete sense, this book also starts from beginninglessness, for it arises from a need for a point of departure—a place from which to begin—in my work on the thought of Dharmakīrti, a South Asian Buddhist philosopher of the seventh century (c.e.).That Dharmakīrti is worthy of our attention seems scarcely necessary to justify. Following upon the work of his predecessor Dignāga, Dharmakīrti addressed at length numerous questions that are of central concern to Buddhist thought and practice. The impact of his views on Buddhist theories of perception, inference, and language is difficult to overestimate. Indeed, it would not be outlandish to claim that his ideas are repeated in every Mahāyāna Buddhist philosophical work written after his time in South Asia. To this day, the Tibetan translations of his Sanskrit texts are recited, studied, and debated by Tibetan monastic scholars to such an extent that, in the central monasteries of the largest Tibetan tradition, a lengthy monastic holiday is devoted entirely to debate on the works of Dharmakīrti.The difficulty in beginning a conversation on the work of Dharmakīrti stems from a problem that often plagues systematic philosophy and theology: the elements of the system are so tightly intertwined that the first word of an argument appears to presuppose the system in its entirety.In Dharmakīrti’s case, two circumstances render this hermeneutical circle particularly vexing. First, the systematicity of his thought is matched by its complexity and extreme concision. And second, the Buddhist traditions of South Asia and Tibet, in their reverence for Dharmakīrti, have reappropriated his works through successive generations of commentaries such that we encounter a sometimes dazzling variety of ways to read Dharmakīrti. As a result, we often find a striking lack of consensus on the most basic issues in the contemporary study of Dharmakīrti’s thought.

A lack of consensus is not itself a problem: Dominick LaCapra has noted that one frequently acknowledged sign of a great work is its resistance to definitive interpretation. Nevertheless, in the case of a systematic thinker such as Dharmakīrti, some of our most useful readings must emphasize the tightly woven nature of the web of ideas that constitute his thought, and without a consensus on even his most basic positions, such readings become impossible. Instead, we find ourselves arguing over the details of a particular position—such as his notion of an entity’s nature (svabhāva)—without ever coming to the point where we ask how theories about an entity’s nature relate to other issues, such as the questions of rational justification and authority. The central aim of this book is thus to contribute toward the development of a consensus by presenting the foundations of Dharmakīrti’s philosophy in terms of a consciously constructed starting point.

In speaking of the “foundations” of Dharmakīrti’s philosophy, I mean those issues that repeatedly surface throughout his work: they are the fundamental elements of his conceptual system that, on my view, make all of his arguments possible. I have organized those elements under three broad categories: (1) ontology, (2) the “natural relation” (svabhāvapratibandha) in inference, and (3) the issues of justification and authority, which I place under the rubric of “instrumentality” (prāmāṇya). These broad categories, which structure this book, encompass all the elements that enable one to understand and appreciate any argument made by Dharmakīrti. At the same time, these categories include what is most difficult—and hence, most controversial—in Dharmakīrti’s thought. Thus, somewhat to my surprise, I have written a book that is both an in-depth introduction to Dharmakīrti’s philosophy and a detailed interpretation of certain difficult points in his work.

A Question of Method: A Point of Departure
The central concern of my approach to Dharmakīrti’s thought is my interest in developing interpretations that attend to its systematicity: the manner in which one theoretical position—such as the uniqueness of particulars (svalakṣaṇa)—is mutually constrained and enabled by numerous others, such as the ultimate irreality of universals (sāmānyalakṣaṇa) or the role of habituation in perceptual judgment. Above all, one interpretive practice initially led me to read Dharmakīrti’s work in this fashion: namely, my ineluctable reliance on traditional commentaries. But while I learned this valuable lesson from the style of reasoning employed by traditional commentators, their practices and certain features of their texts likewise posed a set of historical problems that compelled me to construct a starting point. In sum, I will focus on only Dharmakīrti’s earliest texts, the Pramāṇavārttika and Svavṛtti, and I will resort only to the earliest commentators, Devendrabuddhi (ca. 675 C.E.) and Śākyabuddhi (ca. 700 C.E.). To understand my reasons for restricting this study in the aforementioned fashion, I should first explain why I was led to rely on commentaries.

In the several years of research that went into this book, two reasons compelled me to resort frequently to commentaries. First, in practical terms, any reader of Dharmakīrti’s Sanskrit texts knows that his elliptical and intricate statements often remain impenetrable without commentarial elucidation. Speaking in general of Dharmakīrti’s style, Richard Hayes refers to “the tortuous writings of this highly complex thinker.”And referring specifically to the Svavṛtti (PVSV), a text that is especially important for my analysis, Hayes and Brendan Gillon together note, “Dharmakīrti’s style is so terse that it is not always immediately clear what philosophical points he intends to make.”I would add that, leave alone the question of its philosophical content, even the straightforward meaning of a sentence sometimes seem utterly obscure in Dharmakīrti’s sparse style. The result is that, unless one wishes to argue from highly conjectural interpretations, one must refer to commentaries, where missing phrases are supplied and the elegantly tortuous relations of Dharmakīrti’s grammar are plausibly restated. Thus, for purely practical reasons, commentaries become an inevitable companion on any foray into Dharmakīrti’s texts.

Beyond practical concerns, however, lies another compelling reason for my reliance on commentaries: my larger aim—one that extends beyond the present work—is not to understand Dharmakīrti’s thought in and of itself, but rather the subsequent use of his thought throughout the history of Mahāyāna Buddhism. Thus, even if one could somehow understand Dharmakīrti’s works in a manner that ignored the history of their interpretation, such an approach would thwart me in my goals. In part, an ahistorical reading would be useless because it is a fantasy masquerading as truth: my assumption here is that my own understanding is historically conditioned, and thus, an ahistorical reading of Dharmakīrti would be at best deluded. But setting aside questions of delusion, one of my central aims in attending to the use of Dharmakīrti’s thought in particular historical moments is to create an awareness of my own location by reflecting on the way others from a different time and place appear from my perspective to be conditioned by their circumstances. I do not envision that such an awareness will awaken me from the nightmare of history so that I might move beyond my own “contingent arrangements.” Instead, by gaining a greater awareness of that contingency, I hope to create a more effective agency therein.Thus, in consulting commentaries, I have not used them as uncontested restatements of Dharmakīrti’s texts. Instead, I have sought to learn and even to employ the style of reasoning that they bring to the interpretation of Dharmakīrti’s thought.In doing so, I am not only able to use this work as a foil to the styles of reasoning available to me in my own milieu, but I also hope to have offered an interpretation of Dharmakīrti’s thought that readily enables one to more deeply appreciate its relevance in the history of Buddhism.

Having chosen to rely on commentaries, I eventually encountered three features of their style of reasoning that are especially relevant here. The first is the systematicity that we have already mentioned. Put simply, for traditional commentators, the best interpretation of the matter at hand is one that allows for the greatest coherence—or at least produces minimal tension—with any and all other issues addressed by Dharmakīrti. One upshot of this systematic approach is that it inadvertently highlights the pieces that do not fit readily into Dharmakīrti’s philosophical puzzle. Among the strategies used to cope with such inconsistencies is the second relevant feature of this style of reasoning, namely, that meaning resides in the author’s intention, not in his texts, and that in most cases the author’s intention remains the same over the entire corpus of his work. Finally, the third feature of the style of reasoning employed by traditional commentators is straightforward: Dharmakīrti, to put it bluntly, can never be wrong.

In combination, these three features—systematicity, the appeal to authorial intent, and Dharmakīrti’s inviolable correctness—lead to several concrete practices among traditional commentators. For our purposes, the most relevant practices concern the resolution of inconsistencies—cases where pieces of the puzzle do not perfectly fit. Specifically, if a commentator confronts an inconsistency for which he can formulate a solution, he will feel free to supply arguments that, if necessary, move beyond the text; from the commentators’ standpoint, movement beyond the text is justified because the locus of meaning is not the text but Dharmakīrti’s intention. Nevertheless, even though Dharmakīrti’s commentators feel free to move beyond the text, their arguments will not contradict the problematic or inconsistent passage at hand. Instead, commentators construct arguments in the semantic space that is left open by the text itself. In some cases, a passage’s ambiguity opens it to multiple readings, and this ambiguity constitutes the space where such arguments may be created. In most cases, however, the strict and highly inflected style of Dharmakīrti’s Sanskrit—along with a precisely defined technical vocabulary—leaves little room for such semantic maneuvering. As a result, commentators must often depart from the text altogether to compose, in effect, a statement of the unsaid that supplies the requisite argument. The hermeneutical principle that enables a commentator to supply such addenda is the appeal to Dharmakīrti’s intention: one is simply revealing his intention.

For a contemporary academic interpreter, these various features and practices of traditional commentaries lead to two clear advantages and some notable problems. We have already touched on one advantage, namely, the attention to systematicity, which permits questions that are impossible if one attends only to minutiae.As a second advantage, the commentarial practices offer the contemporary interpreter an opportunity to consider the regnant intellectual problems and contributions in each commentator’s historical era. That is, since the commentators are most concerned with resolving inconsistencies, and since resolutions to old problems lead to new critiques, each generation of commentators is thus responding to a new set of concerns typical of that era. Hence, by providing access to the concerns that are distinctive of each generation, the commentaries can serve as an important tool for the work of intellectual history.

While the traditional commentarial approach offers these advantages to the contemporary interpreter, the very same features and practices also prove problematic in a way that has led me to place strict limits on the present study. The attention to systematicity, inasmuch as it is coupled with an appeal to Dharmakīrti’s intention, permits commentators to move freely among Dharmakīrti’s texts, and since Dharmakīrti composed eight philosophical works,a contemporary interpreter would encounter significant problems if she were to uncritically accept the commentators’ approach. In practical terms, the sheer size of Dharmakīrti’s written corpus would require a contemporary historian to apply the current, highly focused standards of historical interpretation over an unmanageable amount of material. But more important, a failure to attend to the differences in Dharmakīrti’s texts effectively leads one to adopt the same stance as the traditional commentators, namely, that meaning resides in Dharmakīrti’s intention, not in the texts. In other words, if I choose to explain the meaning of a passage in Dharmakīrti’s Pramāṇavārttika, for example, by recourse to any other passage in any other of Dharmakīrti’s texts, I must claim that something beyond that text itself links it to those other texts. For the traditional commentators, that link is provided by the intentions in Dharmakīrti’s mind, and unless I affirm an even more obscure linkage, I too will eventually resort to the notion of authorial intention. Beyond the problems attendant upon any attempt to uncover intention,the main difficulty here is that, in constructing my own version of Dharmakīrti’s mind, I will fail to see the version presented by the commentaries at hand. In other words, as the particularities of the texts themselves fade from consideration, so too will the distinction between my own imagined Dharmakīrti and the commentators’ version.

My response to this problem is not to reject the commentaries in favor of some pure reading of Dharmakīrti, nor to reject altogether the intertextuality of Dharmakīrti’s work. Instead, I have chosen to restrict my focus to Dharmakīrti’s earliest and most extensive works, namely the Pramāṇavārttika and the Svavṛtti, the lengthy prose commentary on the first chapter of the Pramāṇavārttika. In every case, I move beyond these texts only when the commentators themselves prompt me to do so. This technique allows me to avoid both the practical and methodological problems associated with an attempt to speak in terms of Dharmakīrti’s entire philosophical corpus.

Another challenge presented by traditional commentaries is the manner in which they are layered. As noted above, the reappropriation of Dharmakīrti in each commentarial generation makes it possible to appreciate the intellectual interests and contributions of each commentator, but we create that possibility only if we discern clearly the distinctions among commentarial strata. As is already evident, on my view each commentator constructs an imagined Dharmakīrti who replaces the text as the repository of meaning, and in this imagined Dharmakīrti’s mind, all systemic inconsistencies find their resolution. In many cases, commentaries from the same generation largely agree—their Dharmakīrti-s can be treated as one—and those commentaries therefore form a single commentarial stratum. When, however, one moves on to another generation (or to another line of interpretation), a new Dharmakīrti appears. And since Dharmakīrti’s texts are taken by the commentators to be the inviolable account of all that matters in regard to issues such as perception, inferential reasoning, and semantics, the history of Buddhist theories on these issues in South Asia is embodied by a line of imagined Dharmakīrti-s, each corresponding to the interpretation of a particular commentarial stratum.The chief challenge for a contemporary interpreter is the work of separating commentarial strata. In short, commentaries tend to build one upon the other, and they thus develop historical layers, often expressed in terms of the accrued repetition of key phrases or ideas from their predecessors. At least some interpretations—and even many phrases—of Devendrabuddhi, the first commentator on Dharmakīrti’s Pramāṇavārttika, appear to be repeated in all subsequent commentaries on that work, no matter how late those commentaries might be.The next commentator, Śākyabuddhi, naturally repeats Devendrabuddhi’s commentary, since part of his work is a subcommentary on Devendrabuddhi’s text. But Śākyabuddhi also expands upon Devendrabuddhi’s work by adding his own insights. When we then come to later authors such as Śāntarakṣita (725?–788) and Dharmottara (fl. ca. 800), we find that the ideas—and often verbatim phrases—of both Devendrabuddhi and Śākyabuddhi are repeated in their works without any indication as to their origin.This trend continues even in Tibet, where ideas and phrases of South Asian commentators are repeated without identification by Tibetan authors.It is worth reiterating that the layering of commentaries does not come about because commentators, in some slavish adherence to tradition, fail to be original. Indeed, the layering of commentaries indicates exactly the opposite: each stratum represents a new set of innovations and insights brought to the issue at hand by that generation’s commentators in response to the various developments of their times. Consider, for example, the following comment of Śākyabuddhi. Here he summarizes a passage from Dharmakīrti’s Svavṛtti that addresses Dharmakīrti’s philosophy of language:

The idea in this passage is that since expressions take as their objects a conceptual appearance that is excluded from other appearances, they are therefore established to have other-exclusions as their objects.

The basic point here is that an object (artha, viṣaya) of an expression (śabda) such as “cow” is actually a specific type of negation that Dharmakīrti calls an “other-exclusion” (anyāpoha). Words, in short, have negations as their objects.

With this passage in mind, let us turn to the much later commentary of Karṇakagomin (fl. ca. 900). As is so often the case, he repeats verbatim this comment of Śākyabuddhi, but he makes an important change:

The idea in this passage is that since expressions take as their object a conceptual appearance that is excluded from other appearances, they are therefore established to have affirmations as their objects.

For Dharmakīrti, the position that expressions have affirmations (vidhi) as their objects is directly opposed to the claim that expressions take otherexclusions as their objects. Thus, when Karṇakagomin repeats Śākyabuddhi’s comment, he ends it with a conclusion that is exactly opposite to

Śākyabuddhi’s. In altering Śākyabuddhi’s conclusion, Karṇakagomin clearly had a specific problem to address,but we can only become aware of that problem if we notice and take as significant Karṇakagomin’s modification of Śākyabuddhi. And we can only do so if we resist the apparent synchronicity of the commentaries.

Attention to commentarial strata—a kind of textual archaeology—is central to my interpretation of Dharmakīrti. Specifically, I aim to present an interpretation that focuses on the earliest commentarial stratum as formed by the interpretations of Devendrabuddhi and Śākyabuddhi.The particular version of “Dharmakīrti” presented here thus arises in dialogue with their version. It is only by developing a restricted interpretation in this manner that we can begin the painstaking task of separating commentarial layers so as to learn the insights that characterize each commentator’s work.And, as noted above, my aim in doing so is to get some sense of the historical development of Buddhist thought as well as my own historical location. The archaeological metaphor that I have employed, however, can be highly misleading, in that it might suggest an almost naïve objectivity that belies my approach to reading Dharmakīrti.

To be mores specific, the archaeological metaphor of “commentarial strata” usefully describes some constraints that I have placed on my interpretation of Dharmakīrti’s thought, but this metaphor does not capture the way I have attempted to read and think through the problems therein. That is, by constraining my approach in historical terms, my aim is not to uncover the meaning of the text as if it were an unearthed object. Rather, I hope to create the conditions that enable me to participate in the vibrancy of the Buddhist tradition’s reverence for such an influential thinker. I do not mean that I will bow unconditionally at Dharmakīrti’s feet—certain reservations about his thought prevent me from doing so. But by locating my interpretation within a particular historical reading from a particular style of reasoning, my goal is to enact in imagination the aporias found (if sometimes then obscured) by the systematic approach employed by the commentators on whom I have relied. Those often complex and intricate aporias are precisely the inconsistencies around which the chapters of this book have been organized: problems in ontology, inferential relations, and justification. Throughout I have attempted to employ a hermeneutics of charity that gives the best possible argument from within a historically located style of reasoning. I will not thereby resolve the inconsistencies that we encounter in Dharmakīrti’s thought, but I do hope that I have come to an interpretation that is “good” in that it “reactivates the process of inquiry, opening up new avenues of investigation, criticism, and self-reflection.”

Some Suggestions for the Reader
Because this book has various aims, it also has various audiences. My overall aim is to make the content and style of Dharmakīrti’s reasoning—as interpreted by Devendrabuddhi and Śākyabuddhi—available to all my readers, and this goal thus applies to all audiences of this book. I also hope, however, to speak directly to specialists in an attempt to encourage a historically focused consensus on at least some central issues. Hence, I aim to present an argument that is of sufficient rigor and detail to maintain a specialist’s attention. In doing so, however, I wish to avoid the risk that, in its technicality and minutiae, my presentation will become impenetrable to readers not actively engaged in this field.

Balancing the needs of specialists and more broadly interested readers has its dangers. Those engaged directly in research on Dharmakīrti or Pramāṇa Theory might occasionally ask for even greater abundance of detail and citation, while those not directly engaged in such research will find additional detail superfluous or even overwhelming. To allay some of this risk, I have taken several steps. First, I have placed all introductory material in the first chapter, where I present an overview of the style of discourse focused on Pramāṇa Theory. Second, when we turn in the remaining chapters to a detailed examination of Dharmakīrti’s own views, I offer a sustained but not overly technical argument in the body of the text, while providing greater technical detail in the notes. Third, to provide all readers with the most relevant primary source material, I have included an extensive appendix of translations from key passages in Dharmakīrti’s works. Finally, I have attempted to avoid to the greatest extent possible the use of square brackets in my translations of primary texts. In the academic study of Pramāṇa Theory, the use of [square] brackets has become a standard practice as a means to indicate words or phrases in the translation that, while implied by the source text or supplied by a commentary, are not actually present in the source text. This practice leads to an extremely literalist notion of translation, where the mark of an “accurate” translation is its ability to be read as a simulacrum of the original. By ignoring the way in which translation involves a dialogue with the text, such an approach produces translations that will not enable a reader to think through Dharmakīrti’s works in the way that I hope to encourage. Brackets, moreover, are obviously directed only at specialists, since only a specialist could sort out the philological implications of such insertions. For other readers, brackets are at best a distraction, and at worst they exclude nonspecialists by reminding them that the translation is simply a crib for the source text, rather than a translation per se. A crib—a tool to ease the reading of Sanskrit and Tibetan—may have its uses, but it obviously is irrelevant to those who cannot read these languages. For these reasons, I have decided to eschew brackets, except where the inserted text is unusually lengthy or not clearly implied. My assumption is that, even without brackets, specialists can readily determine which phrases and words have been supplied by context or commentary.

With the above procedures in mind, I am able to offer some practical suggestions on how a reader might best approach this material. For those actively engaged in the study of Dharmakīrti or Pramāṇa Theory, the overview of Pramāṇa Theory may provoke some useful reflections, but if these readers choose to move directly to the discussion of Dharmakīrti’s method and ontology (chapter 2), they will not lack any material essential to my interpretation. Second, I would remind specialists that many notes may be of particular interest to them, since the notes often contain extended, technical arguments. Other readers may also find the notes of considerable interest, but I would suggest that if the annotative technicalia prove tiresome, the argument in the body of the text may remain both intelligible and useful, even if the notes are not consulted. Finally, I remind all readers that the appendix contains lengthy translations of some relevant primary texts. My interpretation succeeds only to the extent that those texts, perhaps initially daunting, become vibrant and intriguing sources of change.


How to cite this document:
© John D. Dunne, Foundations of Dharmakīrti's Philosophy (Wisdom Publications, 2004)

Creative Commons License
Foundations of Dharmakīrti's Philosophy by John D. Dunne is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Based on a work at
Permissions beyond the scope of this license may be available at