Sources of knowledge

What are the sources of knowledge?

The two Buddhist authors that will be discussed in this section are Dignāga (late 5th and early 6th centuries) and his interpreter Dharmakīrti (early 7th century).1 Both of these thinkers took the position that there are exactly two sources of new knowledge (pramāṇa), namely, sensation (pratyakṣa) and inference (anumāna). They both denied that there is a special source of knowledge from non-human sources revealed somehow to human beings. While the two thinkers have a good deal in common, there were differences in their understanding of the two sources of knowledge that they both recognized.

Dignāga on the two sources of new knowledge

Dignāga claimed that the two sources of knowledge were distinct in that sensation deals only with sensible qualities, which are always particular, whereas inference deals only with intellectible properties, which are always general. Universals and other general properties are never directly sensed, according to Dignāga. Rather, the intellect ignores subtle differences in sensible particulars and forms the notion that things that are not remarkably different are similar. This idea of similarity or generality is then attributed to what has been sensed so that one who is unwary may come to believe that he has actually sensed the similarity, rather than having imposed a mental construct upon sensation.

Dharmakīrti on the two sources of new knowledge

Dharmakīrti’s set of criteria for separating sensation from inference is more complex than Dignāga’s. While accepting that sensation receives only sensible properties and inference deals only in intellectible properties, Dharmakīrti adds several further considerations. These important additions are spelled out in the first three verses of the chapter on sensation of his Pramāṇavārttika (abbreviated PV).

PV 2.1: There are two means of knowing, because there are two types of subject matter, depending on whether it has or lacks the potential to realize one’s goal. [Floating] hair and so forth [seen by a person with eye disease] is not a real object, because one has no striving for [it as] a goal.

PV 2.2: And [there are two kinds of subject matter] depending on whether or not there is similarity and depending on whether or not it is the subject matter of language, because an idea may or may not arise when some other cause [than the object] is present.

PV 2.3: Here, that which is capable of realizing one’s goal is called real in the truest sense; the other is called real by common sense. These two [realities] are the particular and the universal [respectively].

In this passage, Dharmakīrti correlates the two types of knowable object to the two levels of truth recognized throughout Buddhist philosophy. Because the sensible particular has the capacity to realize an object (arthakriyāśakti), it is real in the true sense (paramārthasat); because the intellectible universal lacks this capacity, it is only conventionally real (saṃvṛtisat).

Realizing an object


  1. A more complete presentation of the ideas presented in this lecture are to be found in Richard P. Hayes, “Whose Experience Validates What for Dharmakīrti?,” in Relativism, Suffering and Beyond: Essays in Memory of Bimal K. Matilal, ed. Purushottama Bilimoria and J.N. Mohanty, (Delhi: Oxford University Press, 1997). ↩︎

Work cited

  • Hayes, Richard P. “Whose Experience Validates What for Dharmakīrti?” In Relativism, Suffering and Beyond: Essays in Memory of Bimal K. Matilal, ed. Purushottama Bilimoria, and J.N. Mohanty, 105–118. Delhi: Oxford University Press, 1997.

Leave a Reply

Your email address will not be published. Required fields are marked *