In the previous essay, we discussed the study of knowledge (“epistemology”) in general terms. The central question of epistemology is: how do we know things? What does it mean, to know something? Within the context of classical Indian philosophy, according to both Buddhists and non-Buddhists, the answer is that we know things by means of an “instrument of knowledge” or “epistemic instrument” (pramāṇa).

Direct Knowlege: Perception

Thus, for example, it is generally understood by everyone—both ancient Indian philosophers, and ordinary people today!—that we know the objects of sensory experience by means of “direct perception” (pratyakṣa). In this way, direct perception is universally accepted as an instrument of knowledge: “the epistemic instrument of direct perception” (pratyakṣa-pramāṇa). In fact, direct perception is the only epistemic instrument that all the different intellectual traditions of India, both Buddhist and non-Buddhist, can accept as an epistemic instrument.

One school in particular, the atheist-physicalist Cārvākas, insists that there are no instruments of knowledge other than direct perception. In other words, much like the perspective of many atheist-physicalists today, the Cārvākas argue that we can only truly be said to “know” that which we directly experience for ourselves.

Universals and Particulars

For Buddhists, however, this position is unacceptable. In particular, the two most famous and important Indian Buddhist epistemologists, Dignāga (ca. 480-560) and Dharmakīrti (ca. 600-660), assert to the contrary that there are two main instruments of knowledge: direct perception (pratyakṣa), and inference (anumāna). (In Buddhist epistemology, “scripture” or āgama may be considered a third instrument of knowledge under certain very rare and special conditions; this subtlety will be discussed in a future essay). Dignāga and Dharmakīrti insist that there must be two distinct instruments of knowledge, because there are two different objects of knowledge: the universal (sāmānyalakṣaṇa), and the particular (svalakṣaṇa). In fact, they insist that perception and inference are the only two instruments of knowledge, and that any other purported instruments such as “analogy” (upamāna) or “testimony” (śabda) are either invalid or just special cases of inference.

Dignāga (ca. 480-560).

We will discuss the two different types of “knowledge-object” (prameya), i.e., the particular and the universal, at length in future essays. The key point, for now, is that Dignāga and Dharmakīrti considered it very important to argue that a “universal” is a type of knowledge-object, a kind of thing that you can know, that is not known by direct perception. What, then, is a universal? In brief, a universal is a concept: the thought or idea of a thing, as opposed to the thing itself.

Why Does Inference Matter?

For now, the main thing to keep in mind is that, above all else, the Indian science of logic is concerned with practical action in the world. So when, for example, we are cold and interested in acquiring warmth, there is a basic distinction that needs to be made between an actually-existing fire that we directly perceive right in front of us, and a fire that we conceptually know exists somewhere far away from us. It may equally be said in both cases we have certain knowledge that the fire exists. But the “means by which” we have this knowledge—which is to say, the “instrument of knowledge” (pramāṇa) employed—is different. In the first case, when we directly perceive a proximate fire right in front of us, we know this fire exists by means of the instrument of direct perception. In the second case, when (in the classic example) we see smoke on a mountaintop, and we remember that “where there is smoke, there is fire,” we know that the remote fire exists by means of the instrument of inference, because we infer the existence of the fire from the perception of the smoke. In this way, inference is a complex cognitive operation involving the perception of a “sign” or “evidence” (liṅga), as well as the “recollection” (smṛti) of the necessary causal relationship that exists between this evidence (i.e., the smoke) and the object of inference (i.e., the fire).

Dharmakīrti (ca. 600-660).

Here, the most important thing to understand is the motivation behind Dignāga and Dharmakīrti’s insistence that inference is indeed an instrument of knowledge. Essentially, their motivation is to answer the objection of Cārvākas and others like them, to the effect that we can only ever know what we directly perceive. The problem with this view, from a Buddhist perspective, is that as ordinary beings afflicted by ignorance, we do not yet directly perceive the reality of the Four Noble Truths, the efficacy of the Noble Eightfold Path, and so on. That is to say, as ordinary beings we are not yet capable of “seeing reality as it truly is” (yathābhūtadarśana), and so we are incapable of independently verifying the truth of the Buddha’s teaching for ourselves, by means of our own direct perception. Hence, we must at first rely upon inference to attain certainty in the truth of the Buddha’s teaching, until we have progressed on the path to the point that we are able to verify its truth by means of our own experience.