1415

1   Pramāṇa Theory: Dharmakīrti’s Conceptual Context

IF WE ARE TO ENGAGE with Dharmakīrti’s philosophy in a manner that enables us to think through his style of reasoning, then we must learn to speak Dharmakīrti’s language: that is, we must become skilled in the discourse that makes Dharmakīrti’s philosophical choices possible. Since that philosophical language is highly complex and precisely inflected, some readers may find it helpful to have a primer of sorts. With those readers in mind, I have provided in this chapter a basic overview of Dharmakīrti’s conceptual context.¹ To do so, the chapter emphasizes some significant points of convergence among South Asian philosophers of Dharmakīrti’s era who participated with him in a style of discourse that I call “Pramāṇa Theory.” Thus, in a secondary sense, this chapter will also alert readers to some of my presuppositions, for any attempt at a synoptic overview inevitably reveals at least some of its author’s assumptions.

1.1   The Process of Knowing and Its Instrument

To understand Dharmakīrti’s conceptual context, we must appreciate that his location within the Buddhist tradition is only part of a more complex landscape. Although he clearly owes much to his Buddhist predecessors, his work also draws from other traditions. In some cases, Dharmakīrti appears to adopt others’ theories, but most notably he adopts a particular mode of16 discourse in which subject matter, technical vocabulary, rhetorical style, and approach to reasoning are all shared by numerous philosophers from several traditions. We can refer to this style of discourse as Pramāṇa Theory, or “theory of the instruments of knowledge.”² It is the kind of philosophy practiced by the most important of Dharmakīrti’s principal interlocutors, including the Naiyāyika Uddyotakara, the Vaiśeṣika Praśastapāda, and the Mīmāṃsaka Kumārila.³ The primary concern of Pramāṇa Theory is the determination of what constitutes indubitable or indisputable knowledge and the reliable means of attaining it.⁴ While many South Asian philosophers examine knowledge in a general fashion, Pramāṇa Theorists discuss this issue in great detail through a shared technical vocabulary that permits and encourages dialogue across traditions.

That is, philosophers who focus on the study of pramāṇa deliberately engage with other philosophers—both from their own philosophical lineage (paramparā) as well as other traditions—over specific questions within a larger, shared context. To some extent, this larger context consists of a particular style of Sanskrit verse and prose, but it also stems from incessant attention to an ongoing dialogic context. Hence, these thinkers continually refer not only to previous texts within their own traditions, but also in others’ traditions. In employing such deliberate intertextuality, Pramāṇa Theorists17 do not simply note what had been thought in the past; rather, they attempt to justify a particular interpretation by responding to the criticisms of others, whether within or outside their own traditions. Each generation of philosophers thus represents a new layer of interpretation formed by new criticisms and rebuttals. Already by Dharmakīrti’s time, the debates between various traditions had gone back and forth several times, and his work is thus thoroughly ensconced in the context formed by earlier criticisms and his own attempt to justify what he sees as the Buddhist view. One upshot of all this is that, in some ways, Dharmakīrti shares more with thinkers from other traditions than he does with Buddhists such as Sthiramati or Candrakīrti, who do not engage in pramāṇa discourse.⁵

The general contours of Pramāṇa Theory that delimit Dharmakīrti’s own thought find their first systematic expression in the Nyāyasūtras of Gautama (ca. 150 C.E.).⁶ Even at this early stage, a notable characteristic of Pramāṇa Theory is the development of a technical vocabulary that all later Pramāṇa Theorists inherit and share. A central theme in this vocabulary is the use of what I call the “kāraka system,” a formulaic way of analyzing the “functional elements” or kārakas that contribute to an action (kriyā).⁷ Following Gautama’s lead, Vātsyāyana (ca. 475), the earliest commentator on the Nyāyasūtras, applies the kāraka system to the verb pramā, “to know indubitably.”⁸ Of the possible kārakas or elements in an action, three are particularly relevant to the analysis of the act of knowing: the agent (kartṛ) who acts on an object or “patient” (karman) by means of an instrument (karaṇa). Adding to these three the action (kriyā) itself, Vātsyāyana and all18 subsequent Pramāṇa Theorists apply this kāraka analysis to the verb pramā so as to derive four terms: pramātṛ, pramiti (or pramā), prameya, and pramāṇa.⁹ These terms refer to the agent who knows (pramātṛ), the action of knowing (pramiti or pramā), the object known (prameya), and the instrument used to acquire that knowledge (pramāṇa). Using these four terms, Pramāṇa Theorists developed a fourfold style of analysis to analyze knowledge events. That is, their overall analytical framework assumed that every knowledge event involved the event as an action, an agent engaged in that action, a means for the production of that action, and an object to which that action is principally related. Analyses of the process of knowing through these four terms became standard among Pramāṇa Theorists.¹⁰

Before we continue with our discussion of these four facets of knowing, we must first recognize that readers familiar with the epistemological theories developed in the Euroamerican philosophical traditions may feel that our use of the term “knowledge” here is somewhat irregular. On most Euroamerican accounts, “knowledge” is a belief or attitude that is true (under some set of conditions or truth theory). As a belief or attitude, “knowledge” is dispositional, and it therefore cannot be an act in itself. But on the account of Pramāṇa Theory that we have given above, “knowledge” (pramiti or pramā) is the act (kriyā) of “knowing indubitably” that is constituted by a process involving the interaction of an agent, instrument, and object of knowledge. This model requires that the “action of knowing” (pramā or pramiti) be a cognitive event occurring in a particular person’s mind within a particular set of circumstances. A theory of knowledge must therefore take into account any relevant aspect of those circumstances that, for example, might distort a cognitive event in such a manner that we should not consider it knowledge. In examining distortions that prevent a cognitive event from being a knowledge event, these theorists shared a general conception of the relation between body and mind. Hence, they all think it relevant to discuss at length the way in which physical infirmities such as jaundice or cataracts might distort cognitive events: a person with jaundice will see conch shells as yellow; a person with cataracts thinks that his water-jug is filled with small pieces of hair. They also generally maintain that intense emotions such as intense anger or lust so strongly affect the19 mind that all cognitions occurring with those emotions are necessarily distorted. This way of approaching cognitive distortion—and numerous other such issues—clearly indicates that an account of the cognitive event or act called “knowledge” (pramiti or pramā) is concerned largely with the process of producing that event. And the model that we have cited—involving the interaction of agent, object, and instrument—provides the overall structure for Pramāṇa Theorists’ analysis of that process.¹¹

When Gautama, Vātsyāyana, and subsequent Pramāṇa Theorists used this model to give an account of knowledge-events, their works address especially the pramāṇas or “instruments of knowledge,” and it is for this reason that Matilal and others refer to this genre of philosophical literature as Pramāṇa Theory. But why take an analysis of the instrument as one’s thematic focus? Why not focus instead on the agent, object, or event itself?¹² To answer such questions in a somewhat speculative manner, we might give a historical argument that borrows a principle of Pramāṇa Theory itself: if two persons are to have an argument, they must first share many points of agreement. That is, if any two discussants are to disagree meaningfully on some point, their discussion must be framed within some area of agreement.¹³ When discussing the acquisition of indisputable knowledge, Pramāṇa Theorists generally agree on many basic notions about the instruments of knowledge (pramāṇa), whereas they generally encounter fewer areas of agreement on other aspects of that process. Since they tend to agree more readily on issues related to the instrument or means in the process of knowing, the instrument naturally becomes the focus—the propositional subject—of their discussions. The difficult problem we face in making this type of argument is that we cannot readily explain why it is that these thinkers tended to agree more readily on issues related to the instrument of knowledge. We may suspect that some large pool of common assumptions underlies the emphasis on the instruments of knowledge, or perhaps that an emphasis on the instrument most readily affirms their approach by excluding other styles of discourse. Somewhat ironically, these20 suspicions require us to acknowledge that Pramāṇa Theorists would not explicitly discuss shared assumptions or covert exclusions, since all such issues would be obscured by their very givenness. Hence, due to the relative lack of research in this area, the subtler form of this historical argument can only be suggestive at this point.¹⁴

Putting aside covert notions, one can also point to arguments made by the theorists themselves. Among these are two distinct arguments that explicitly acknowledge an emphasis on the importance of the instrument of knowledge (pramāṇa), rather than the agent (pramātṛ), object (prameya), or the action of knowing itself (pramiti). The first argument is suggested by the comparatively early works of Vātsyāyana and Uddyotakara.¹⁵ This argument amounts to the claim that the emphasis on analysis of the instrument of knowledge derives from its primacy in the process of knowing. To use the analogy of a person cutting a tree with an axe: the person and the tree can be identified as the “cutter” and the “cut object” only when the action of cutting occurs, and that action can only occur when a cutting instrument—the axe—is employed. It is only by changing the type of instrument used that the action then becomes a different action. That is, if we replace the agent with some other person, or if we can direct the axe against some other object, the action is still the action of cutting. In short, neither the agent nor object can change the character of the action. If, however, some other kind of instrument, such as a yardstick, is used, then the agent (“the cutter”), object (“that which is cut”) and action (“cutting”) all take on a different character: they become the “measurer,” the “measured” and the action of “measuring.” Hence, inasmuch as the character of the instrument determines the character of the other three factors, the instrument is primary. This way of understanding the instrument as primary appears to have been widely accepted among Pramāṇa Theorists, including Dharmakīrti.¹⁶

The second set of arguments that explicitly acknowledge the emphasis on the instruments of knowledge are adduced only by Buddhist philosophers, beginning with Dignāga and Dharmakīrti. These philosophers reject the notion of an agent, and on their view, the cognitive event identified as knowledge is ontologically identical to the instrument, which they conceive to be a mental image. In some contexts, they also regard the object or patient as dependent in some sense on the instrument, either because it is not ontologically distinct from the instrument, or because the character of that object is determined by the character of the instrument itself. Hence, on their view, the instrument is clearly primary, since all of the other functional elements in a knowledge event are either unreal or determined by the instrument. In subsequent chapters, we will have an opportunity to examine Dharmakīrti’s views on all these issues in greater detail.

Finally, one can also note that the emphasis on the instruments of knowledge allows (or even requires) Pramāṇa Theorists to discuss at length the place of scripture (āgama) or verbal testimony as such an instrument. In all Pramāṇa Theories, scripture plays a special role, in that it is an instrument (pramāṇa) or means that enables one to obtain knowledge that is otherwise utterly beyond one’s ken. Many claims verifiable only by scripture often bear directly on the soteriological goals of the tradition in question. If we assume that Pramāṇa Theorists took those soteriological goals seriously, we would expect them to be especially concerned with knowledge derived from scripture, since scripture is the means to that soteriologically relevant but otherwise unobtainable knowledge. For this reason as well, Pramāṇa Theorists might be inclined to think that the instrument is the most important aspect in the process.

Regardless of the historical and philosophical reasons, two issues remain clear: first, that Dharmakīrti’s conceptual context is formed by an intensive analysis of the process of knowing as embodied by the aforementioned model; and second, that it is especially a knowledge-event’s instrument—and not its object, agent, or the event itself—that most concerned the theorists that Dharmakīrti directly addresses. As we have noted, it is likely that22 this shared emphasis on the importance of the instrument is encouraged by a host of covert and obscure assumptions. Nevertheless, we can still summarize a rather large number of quite clear and explicit assumptions offered by these theorists themselves.

With this in mind, I begin this sketch of the conceptual context of Dharmakīrti’s thought by discussing the pramāṇas or “instruments of knowledge” so as to highlight the notions that he shared with other Pramāṇa Theorists. I will move on to examine some shared notions concerning instrumental objects (prameya), and after highlighting the importance of purpose, I will conclude with some brief remarks concerning the agent (pramātṛ) and knowledge-event itself (pramā or pramiti).

Two Ubiquitous Instruments: Perception and Inference

When speaking of the instruments of knowledge, the various traditions of South Asian philosophy and the individual philosophers within those traditions disagree considerably on exactly what ways of knowing should be considered instrumental (i.e., instances of pramāṇa), and what forms are spurious or faulty. They also disagree about the criteria through which one can adjudicate whether a particular form of knowledge is instrumental or not. Despite these and other disagreements, they find considerable common ground on a number of other issues.¹⁷ The foremost of these is simply the notion that the instruments of knowledge must be investigated; for most of these philosophers, this need stems from the centrality of knowledge in the search for spiritual freedom or mokṣa. That is, to become free, one must rely upon correct knowledge, but if one is unable to distinguish correct from incorrect knowledge, how could one recognize one’s knowledge as correct?¹⁸

With the renowned but comparatively sparse exception of the Lokāyata or Cārvāka tradition,¹⁹ all Pramāṇa Theorists respond to the need for a means of obtaining indubitable knowledge by positing at least two basic instruments:23 perceptual awareness (pratyakṣa) and inference (anumāna).²⁰ Of course, the virtually ubiquitous acceptance of perception and inference does not prevent these thinkers from disagreeing on exactly how these instruments of knowing operate. Nevertheless, in accepting perception and inference as instruments of knowledge, Pramāṇa Theorists share certain presuppositions and basic doctrines concerning the instruments of knowledge.

Shared Notions Concerning Perceptual Awareness

When speaking of perceptual awareness, Pramāṇa Theorists agree, first of all, that this way of knowing depends directly on the senses. Indeed, the centrality of the senses in this way of knowing is implied by the term pratyakṣa itself, which is often construed etymologically to mean “before the senses.”²¹ We must be careful, however, to recall that in addition to the five senses familiar to Euroamerican traditions, these philosophers also stipulate a sixth sense: the mental faculty (manas). Hence, any instance of “perceptual awareness” may be an awareness of a mental object, rather than a visible form, sound, smell, taste, or tactile object. Pramāṇa Theorists nearly all agree on the stipulation of a sixth sense, and they all agree on the centrality of the senses in perceptual awareness.²²

Another general point of agreement concerns the manner in which perceptual awareness occurs. All Pramāṇa Theorists agree that perceptual awareness necessarily involves the contact (sannikarṣa, sparśa, etc.) of an object (viṣaya, artha, etc.) with a sense faculty (indriya).²³ And except in the case of mental objects, they generally assume it appropriate to consider this contact to involve a relation involving matter (rūpa) or substance24 (dravya). They also agree that physical (i.e., material or substantial) defects in the sense faculties can contribute to certain types of errors in perceptual awareness, as when a person with cataracts apparently sees small hairs or bugs in front of their eyes.²⁴ Another important point of agreement is that perceptual awareness is either the most vivid or the least mediated form of awareness, and that in this sense it takes precedence over other instruments of knowing, such as inference.²⁵ Most of these philosophers also agree that the basic building blocks of matter are irreducible, partless atoms or “infinitesimal particles” (paramāṇu). According to the philosophers who accept this notion, infinitesimal particles are too small to be perceived by ordinary persons; instead, the matter perceived by ordinary persons consists of particles that have somehow been aggregated into an entity of perceptible size.²⁶

Although these points of agreement are certainly significant, it is important to note that Pramāṇa Theorists often disagree upon the precise content of perceptual awareness, either because their ontologies conflict, or because they differ over the degree to which perceptual awareness is determinate.25 We will consider some of these debates when examining Dharmakīrti’s particular theory of perception, but for now, let us turn to an overview of inference (anumāna).

Shared Notions Concerning Inference

Inferential knowledge and the topics related to it are particularly important to Pramāṇa Theorists.²⁷ One can point to three basic reasons for the importance of inference: first, it provides access to entities that are to some degree unavailable to the senses, and such entities are often under dispute. Second, it is closely tied to the understanding of language, an issue that is essential to the success of the South Asian philosophical enterprise.²⁸ And third, it provides the framework for formal disputation, an undeniably crucial aspect of South Asian philosophy.

As Matilal has noted, the earliest theories of inference probably arose out of a concern with the codification of philosophical debate, but properly speaking, what is meant by inference here is not a “syllogism” or some other argument. Rather, an inference produces or constitutes a knowledge-event that knows its object by means of knowledge about another object that is invariably related to that object. A stock example is the inferential cognition that knows fire is present in a particular locus by means of perceptual knowledge of smoke in that same locus. Inference clearly involves some steps, for in providing knowledge of one thing by means of knowing something invariably related to it, the act of inference requires a sequential structure, which we will discuss below. Nevertheless, the central concern for these thinkers is not the formalism of that structure itself; instead, they are most concerned with the way in which that structure supplies the necessary conditions for an inference.

Pramāṇa Theorists generally speak of two forms of inference: “inferencefor-oneself” (svārthānumāna) and “inference-for-others” (parārthānumāna).26 The former is simply an inferential cognition: one looks at a smoky room, for example, and (with other conditions in place), one infers that fire is present. In contrast, an inference-for-others is one that is stated verbally so as to induce an inferential cognition in another person. In other words, this latter “inference” (which is actually a series of statements and not an inference) is meant to result in another person having his own inference-for-oneself with regard to the question at hand. In this sense, inference-for-oneself lies at the core of these thinkers’ inferential theory. But ironically, the structural elements that are necessary for one to have an inference-for-oneself are primarily explored in discussions of inference-for-others. To avoid the confusion that this overlap incurs, below I will often speak simply of “inference,” with the understanding that our main focus is the examination of the conditions necessary for a correct (as opposed to a spurious) inferential cognition to occur.

THE BASIC STRUCTURE OF INFERENCE

As one might expect, the aforementioned importance placed on inference prompts considerable disagreement among Pramāṇa Theorists, but their analyses of inference always include the same basic, minimal structure.²⁹ Schematically, I render it as follows:

S is P because E

A typical example of this type of inference is:

The hill (S) is a locus of fire (P) because of the presence of smoke (E).³⁰

Here, S is the “subject,” called the sādhyadharmin or pakṣa in Sanskrit;³¹ P27 is the “predicate,” known as the sādhyadharma; and E is the “evidence,” known as the hetu or liṅga. The first two elements, the subject and predicate, together form the “proposition” (pratijñā or pakṣa), “S is P.”³² Hayes and others have employed an alternative terminology, where the subject is called the “quality-possessor” and the predicate the “quality.”³³ This terminology has the advantage of conveying more literally the sense of the Sanskrit terms (sādhya-)dharmin and (sādhya-)dharma, and it avoids any28 potential misunderstanding concerning the notion of a proposition. Nevertheless, “quality-possessor” is quite cumbersome, and inasmuch as the English term “quality” can also be misleading, “subject” and “predicate” appear to be the best choices.³⁴

THE EVIDENCE-PREDICATE RELATION AND ITS EXEMPLIFICATION

According to Dharmakīrti and his fellow Pramāṇa Theorists, any theory of inference must contain at least the relations implicit in the basic model presented above. The first such relation is generally called the vyāpti or “pervasion”—it is the relation between the evidence (E) and the predicate (P).³⁵ In our example, this is the relation between fire and smoke. Pramāṇa Theorists generally consider this relation to have two aspects: the positive concomitance (anvaya) and the negative concomitance (vyatireka).³⁶ The positive concomitance is a state of affairs such that wherever the evidence (E) is present, the predicate (P) must be present. In our example, this would be stated, “wherever there is smoke, there is necessarily fire.” The negative29 concomitance specifies that the evidence (E) is present only in the presence of the predicate (P) and not in any other circumstances. Dharmakīrti often states the negative concomitance, or “restriction,” in an affirmative statement (i.e., a statement that involves no grammatical negation). In our example, the positive statement would read, “There is smoke only where there is fire.”³⁷ Most Pramāṇa Theorists, however, formulate the negative concomitance or restriction in negative terms; following our example, a negative statement of this concomitance would read, “wherever there is no fire, there is necessarily no smoke.” For Dharmakīrti and Kumārila—and probably also for Uddyotakara and Praśastapāda—the positive and negative concomitance are in contraposition: if smoke is necessarily present when fire is present, then in the absence of fire, smoke is also necessarily absent.³⁸

According to these philosophers, in order to have an instance of inferential knowledge one must be aware of the pervasion—the twofold relation consisting of the positive and negative concomitance. So too, the pervasion must be general: it cannot be restricted to a single case, but must pertain to30 all cases of the kind in question. In the dialogical context of inference-forothers, these two requirements—that one have knowledge of the pervasion and that it be general—are reflected by a frequent claim: namely, that an inference-for-others must be accompanied by at least a supporting example (sādharmyadṛṣṭānta). Most philosophers maintain that a counterexample (vaidharmyadṛṣṭānta) may also be necessary, at least in some cases. The supporting example is drawn from the domain of “homologous instances” (sapakṣa)—namely, loci that are similar to the proposition to be proven (S is P) in that they are qualified by P. In an inference of fire from smoke, a kitchen (mahānasa) is the typical example. The aim is to appeal to a noncontroversial case that exemplifies the relationship between the evidence (E) and the predicate (P): one’s past experience of kitchens illustrates the positive concomitance of smoke with fire, in that one’s observations conform to a necessary relation between the presence of smoke and the presence of fire. The counterexample is drawn from the domain of heterogeneous instances—loci that are dissimilar to S in that they lack P. A typical counterexample for the smoke-fire inference is a lake. Here, the point is to show that the presence of the evidence (E, the smoke) is not observed in the absence of the predicate, fire; or alternatively, that smoke is present only when fire is present, and not otherwise.³⁹

In part, the use of examples indicates the psychologism within pramāṇa discourse.⁴⁰ That is, these philosophers are not interested only in the formal aspects of inferential reasoning; rather, they wish to demonstrate the conditions necessary for the occurrence of a knowledge-event that is inferential in form. The distinction here is between the knowledge that smoke is always concomitant with fire, on the one hand, and the knowledge that a smoke-producing fire is present in a particular case, on the other. For Pramāṇa Theorists, the positive concomitance is a relation that must pertain between the evidence (E) and the predicate (P) if we are to infer that the subject (S) is qualified by the predicate (P) because it is qualified by the evidence (E). But for these theorists, it is also crucial that the knowledge of the positive concomitance is a necessary part of the process that leads to an inferential cognition of fire in a locus by way of a perceptual cognition of smoke in that locus.⁴¹

In addition to the psychologism underlying the use of examples, one can also point to certain ontological concerns that are implicit in claims for the32 necessity of examples in an inference-for-others. According to many South Asian philosophers, the twofold relation between evidence and predicate cannot be stated in abstraction from the substances that bear those predicates. When a disputant (let us call him “Devadatta”) attempts to induce another to infer the presence of fire on a mountain from the smoke on that mountain, Devadatta must demonstrate to his interlocutor that the presence of smoke is necessarily concomitant with the presence of fire. But he cannot do so by appealing to the case at hand—the smoke and fire on the mountain—precisely because this case is under dispute. Of course, Devadatta might simply state that relation in abstraction from any given locus or substance, but many Pramāṇa Theorists, especially those from non-Buddhist traditions, resist this approach. This is due in part to the notion that, if the predicates in question are real, they must be instantiated in some substance or locus; and if one cannot appeal to any such undisputed instantiation, then the reality of those predicates remains dubious. Hence, for some Pramāṇa Theorists, one of the reasons for insisting upon examples is that they serve to demonstrate the reality of the entities adduced as predicate and evidence. This ontological requirement also has a certain resonance with an epistemic requirement—namely, that the relation in question must have its final appeal in sense perception itself. In this sense, even if one can logically adduce reasons why some state of affairs must hold true, one’s arguments are generally considered unreliable if one cannot appeal to sensory experience to support that reasoning.⁴²

THE EVIDENCE-SUBJECT RELATION

So far we have discussed the basic form of inference, and we have discussed one of the key relations in this inference: the twofold pervasion (vyāpti) consisting of positive and negative concomitance (anvaya and vyatireka) that pertains between the evidence and the predicate. One other key inferential relation must also be discussed: the relation called upanaya (“application”) or, as Buddhist thinkers tend to call it, pakṣadharmatā (“presence of the quality in the subject”). In some ways, this relation is straightforward: it simply consists of the relation between the evidence and the subject (the dharmin or pakṣa) of the proposition in question. In other words, for pakṣadharmatā to hold true, the evidence must be known to be a quality or predicate (dharma) of the proposition’s subject. In the example of inferring fire on a mountain from the presence of smoke, pakṣadharmatā would simply mean that the smoke used as evidence is present on the mountain. The need for this relation is probably quite obvious; after all, it would make little sense to prove that the mountain is on fire by noting that smoke is present in my pipe. An even more obvious example would be the inference: “Joe is a bachelor because of being unmarried.” Pakṣadharmatā here would simply mean that evidence adduced—the fact of being unmarried—pertains to him, not someone else. Otherwise, we might infer: “Joe is a bachelor because his dog is unmarried.” And this does not make any obvious sense. Thus, the basic point of pakṣadharmatā is that one must readily know that the evidence is a predicate or property of the subject. Some philosophers, such as the Naiyāyika Uddyotakara, claim that this relation must always be known through perception,⁴³ but Dharmakīrti and subsequent Buddhists maintain that this relation may be determined through another inference.⁴⁴

A RESTATEMENT

With the above discussion in mind, let us restate the basic elements of inference according to Pramāṇa Theorists. This restatement combines the elements of both inference-for-oneself and inference-for-others, and it is meant as a heuristic overview of inference, rather than the depiction of any philosopher’s theory:

Proposition (pratijñā, pakṣa): The mountain (S) is a locus of fire (P).

Evidence (hetu, liṅga): Because there is smoke (E).

The Evidence-Predicate Relation (vyāpti, pervasion) and Its Exemplification: Wherever there is smoke (E) there is fire (P), as in a kitchen. And without fire (~P), there is no smoke (~E), as on a lake.

The Evidence-Subject Relation (pakṣadharmatā or upanaya): This mountain (S) is a locus of smoke (E).

All of these elements figure explicitly or implicitly in every Pramāṇa Theorist’s analysis of inference. In the case of inference-for-oneself, the exemplification per se is superfluous, but the principle expressed by that exemplification—that the evidence-predicate relation be generalizable beyond the case at hand—is still required. In a sense, all the elements are also only implicit in an inference-for-oneself, in that they are not explicitly stated, whereas at least some of the elements must be explicitly stated in an inference-for-others. Numerous disagreements arise, however, on the details of inference-for-others. We have already noted, for example, that these philosophers do not agree on the degree or type of exemplification necessary in an inference-for-others. Similar disagreements abound concerning which elements may be dropped as superfluous to a statement of inference, or whether some additional statements are required. But these disagreements focus primarily upon the explicit presentation or repetition of one element or another; the implicit presence of these elements in an inference is not a matter of contention.⁴⁵

We have now covered the most salient views that Pramāṇa Theorists share about the two ubiquitous forms of instruments of knowledge: perception (pratyakṣa) and inference (anumāna). Let us now turn to some basic views concerning the instrumental object (prameya), the object of an instrument of knowledge.

1.2   Prameya: The “Real”

As noted previously, the term prameya refers to the object of the indubitable knowledge derived from an instrument of knowledge or pramāṇa, and to clarify that a prameya is specifically an object of this kind of knowledge, I will generally translate prameya as “instrumental object.”⁴⁶

For Pramāṇa Theorists, an instrumental object is necessarily what we might call “real” in English. I am thinking here especially of the Sanskrit term sat, a participle formed from the verb “to be/exist” (as). The connotations36 of sat converge on the notion of something that is present in a substantial fashion, be it directly or indirectly. Such an object is “real” because only “the real” can be the content of a correct or indubitable knowledge-event: for Pramāṇa Theorists, it makes no sense to speak of an indubitable cognitive event whose object is unreal.⁴⁷ Clearly, this position rests on several assumptions, the most obvious of which is the notion that cognitive events always have objects. This is less trivial than it sounds, for these philosophers maintain that every mental state or form of consciousness is a cognitive event; in short, they espouse an intentional theory of consciousness. That is, all moments of consciousness necessarily have objects, and there are thus no instances of contentless awareness or moments of consciousness without objects. Pramāṇa Theorists thus claim that, even in instances where a cognition is mistaken, one must still account for the presence of an object, even though that object is somehow incorrectly cognized.⁴⁸

In addition to claiming that instrumental objects (prameya) are real, these philosophers also maintain that the “real” is necessarily “knowable” (jñeya), and this is understood to mean that the “real