Pramana

• 1. Introduction

  ›

  Pramana names the reasoning behind a classic example from ancient Indian logic: a man is fat, and he never eats during the day. Logicians reasoned that if both are true, he must be eating at night.

  That leap, from what can be observed to what must be true, is only one route to knowledge. Ancient Indian philosophers debated several other routes, and argued fiercely over which ones could be trusted. Comparing something to a memory, drawing a conclusion from what did not happen, and taking someone's word were all proposed as ways of knowing something.

  Some schools of thought trusted almost none of these methods. Others accepted every one on offer. What was actually required for a claim to count as knowledge? And why could two serious philosophers examine the same methods and land on different answers?

• 2. Naming And Counting The Paths

  ›

  The word pramana literally means proof, built from the Sanskrit root pra, meaning outward or forth, joined to ma, meaning measurement. Combined, the root sense grew into pramā, a word for correct notion, true knowledge, or basis and foundation. Ancient Indian thought treated pramana as one part of a trio, alongside pramatr, the knower, and prameya, the knowable thing itself. Each of the three, the means of knowing, the one who knows, and the thing known, shaped the outcome by its own nature.

  Yukti, a related term, describes the active application of what a person already knows, an innovation or a clever methodological trick. Some Indian texts discuss yukti and pramana side by side, treating yukti as active reasoning against pramana's more passive path through direct observation. The Samkhya, Yoga, Mimamsa, and Advaita Vedanta schools built entire theories of error around this idea. Each asked how a person could recognize a wrong conclusion and correct it.

  Verse 1.2.1 of the Taittiriya Aranyaka, composed sometime between the 9th and 6th centuries BCE, lists only four means of attaining correct knowledge. Those four were smrti, meaning scripture or tradition, pratyaksa or perception, aitihya, meaning expert or historical testimony, and anumana, meaning inference. Later texts pushed the count in other directions: Vedvyasa's writings discuss as many as ten pramanas, while the scholar Krtakoti argued for eight. Over time, six pramanas became the most widely recognized: Pratyaksha, Anumana, Upamana, Arthapatti, Anupalabdhi, and Shabda.

  Each of those six would earn its own body of rules for when, exactly, it could be trusted.

• 3. Seeing And Sensing

  ›

  Pratyaksham, meaning perception, comes in two forms. External perception arises from contact between the five senses and worldly objects. Internal perception is the awareness that arises from the mind itself. According to the scholar Matt Stefan, the real distinction is between direct perception, called anubhava, and remembered perception, called smrti.

  Ancient and medieval Indian texts set four requirements for a perception to count as correct. The first, indriyarthasannikarsa, demands direct contact between a sensory organ and the object being studied. The second, avyapadesya, rules out hearsay, where a person simply accepts or rejects someone else's perception secondhand. The third, avyabhicara, demands consistency, ruling out perception that shifts because a sense organ is drifting, defective, or suspect. The fourth, vyavasayatmika, demands a definite judgment, free of doubt from missed details or from mixing inference into observation.

  Some ancient scholars proposed a further category called unusual perception, treating it as a form of internal perception. This covered pratibha, or intuition; samanyalaksanapratyaksa, reasoning from specific cases to a general rule; and jnanalaksanapratyaksa, perceiving a thing's past state from its present one. Other Indian scholars contested the whole idea.

  Some schools of Hinduism went further still, distinguishing nirnaya, a definite judgment, from anadhyavasaya, an indefinite one still open to doubt.

• 4. Logic Of Smoke And Fire

  ›

  Anumanam means inference in Sanskrit, though modern Indian languages often use the same word loosely to mean guess. Classical philosophy treats it as reaching a new conclusion by applying reason to one or more observations and prior truths. Observing smoke and inferring that a fire must be burning is the standard example given in the Indian texts. Nearly every Hindu philosophy accepts inference as valid, with a single exception among the major schools.

  Indian texts break the inference process into three parts: pratijna, a hypothesis; hetu, a reason; and drshtanta, supporting examples. The hypothesis itself splits further into sadhya, the idea to be proven or disproven, and paksha, the subject that idea is predicated on.

  An inference only counts as conditionally true when sapaksha, positive supporting examples, are present and vipaksha, negative counterexamples, are absent. Rigorous philosophers demanded vyapti as well, the requirement that the reason hold true in every case, across both sapaksha and vipaksha. A hypothesis that survived all of these tests earned the name nigamana, a conclusion, though not every proposed path to knowledge could clear that bar.

• 5. Comparison And Circumstance

  ›

  The scholar Lochtefeld illustrates Upamana, comparison and analogy, with a traveller who has never visited a land with its own unfamiliar wildlife. Someone who has been there describes an animal that looks like a cow and grazes like a cow, yet differs from one in some way. Indian epistemologists treat that kind of comparison as valid conditional knowledge, since it helps the traveller recognize the animal later.

  The subject of a comparison is called upameyam, the object it is compared to is upamanam, and the shared quality is samanya. Monier Williams offers the line 'her face is like the moon in charmingness' as an example. In it, her face is upameyam, the moon is upamanam, and charmingness is samanya. The 7th-century text Bhattikavya, in verses 10.28 through 10.63, catalogs many types of comparison and weighs when each one is reliable. Across ancient and medieval Hindu texts, scholars debate 32 distinct types of Upamana and their value to epistemology.

  Arthapatti, postulation or derivation from circumstance, works like circumstantial implication in modern logic. If a person was left on a boat on a river, and the expected time of arrival has passed, the circumstances suggest they have arrived. Many Indian scholars judged this kind of postulation weak, since the boat could just as easily have been delayed or diverted. Proponents pointed instead to deriving the time of a future sunrise or sunset as a case where the method held up reliably.

  Schools that rejected Arthapatti argued its conclusions were either derivable from other pramanas or simply flawed. They preferred to rely on direct perception or sound inference instead.

• 6. Absence And Testimony

  ›

  Anupalabdhi treats non-perception itself as a form of valid knowledge, as in knowing there is no jug in a particular room. If something can be observed, inferred, or otherwise proven to be nonexistent or impossible, a person knows more than they did before. Scholars refined this into four types: non-perception of the cause, of the effect, of the object, and of contradiction. Only two schools of Hinduism ever accepted and developed Anupalabdhi, treating it as valid whenever the other five pramanas failed.

  Abhava, meaning non-existence, is treated by some scholars as identical to Anupalabdhi and by others as a separate idea entirely. The scholar Bartley defines a padartha, or referent of a term, as something existent, knowable, and nameable all at once. Examples of padartha include dravya, substance; guna, quality; karma, activity; samanya or jati, universal or class property; samavaya, inherence; and vishesha, individuality. Ancient scholars described an absence as a 'referent of negative expression.' They pointed to negative numbers, silence used as testimony, and the asatkaryavada theory of causation as evidence that absence itself is real and knowable. The schools that accepted Abhava further divided it into four types. Dhvamsa is the termination of what existed, atyanta-abhava is absolute impossibility, anyonya-abhava is mutual negation, and pragavasa is a prior, antecedent non-existence.

  Shabda means relying on word, specifically the testimony found in the shruti and the Vedas. The scholar Hiriyanna explains it as reliance on reliable expert testimony, since no person has time to verify every fact directly. People instead rely on parents, family, friends, teachers, and ancestors to share knowledge and enrich one another's lives.

  The Charvaka school rejected Shabda outright, arguing that the reliability of any source could never truly be established.

• 7. Which Schools Believed What

  ›

  Charvaka, part of the Sramana tradition, accepted exactly one of the six pramanas as reliable: perception. It dismissed every other method as either invalid outright or too prone to error to trust. The Vaisheshika school allowed two: perception and inference. Jainism went one step further, accepting three: perception, inference, and testimony.

  The Sankhya and Yoga schools, along with the Vishishtadvaita and Dvaita branches of Vedanta, settled on the same three: perception, inference, and testimony. These three are enumerated directly in sutra I.7 of the Yoga Sutras. Sutra I.6 of the same text places pramana as one of five classes of mental modification, alongside indiscrimination, verbal delusion, sleep, and memory.

  Nyaya, whose name literally refers to the science and study of pramanas, accepted four: perception, inference, comparison, and word. The Prabhakara branch of Mimamsa accepted five, adding Arthapatti, postulation, to that same list. Advaita Vedanta, together with the Bhatta branch of Mimamsa linked to the scholar Kumarila Bhatta, accepted all six, adding Anupalabdhi and Abhava.

  Nearly all Vedantins, apart from some Neo-Vedantins, rank shabda testimony as more important to knowledge than anumana, inference.

• 8. Buddhism's Own Logic

  ›

  Buddhism accepted only two of the six pramanas as valid: Pratyaksha, perception, and Anumana, inference. The scholar Rinbochay adds that Buddhism also treats scripture as a third source of valid knowledge. That includes the words of Buddha and other 'valid persons' and 'valid minds,' treated as a form of perception and inference. A 2005 translation of Shantarakshita's The Adornment of the Middle Way, by the Padmakara Translation Group, describes pramana strictly as 'valid cognition.' That same translation ties the term to a tradition of logic and epistemology built mainly by two scholars, Dignaga and Dharmakirti.

  Dignaga's central text, the Pramana-samuccaya, helped shape the discipline of pramanasastra by weaving it together with formal logical argument. Dharmakirti, working under Dignaga's influence, carried those ideas further in his own text, the Pramanavarttika. Dignaga also introduced Apoha. It holds that a mind recognizes a general category not by matching a universal form, but by noting what the thing differs from. A perceived chariot is recognized as a chariot because it differs from things that are not chariots, not because it matches some ideal chariot form.

  The scholar Bhavaviveka, a contemporary of Dignaga writing before Dharmakirti, brought this logical approach into his commentary on Nagarjuna. Candrakirti later criticized that fusion of logic and the Middle Way. He argued the ultimate nature of reality lay beyond thought and concept, outside logic's domain. Bhavaviveka's style became known afterward as the Svatantrika approach, built on independent logical arguments. When Madhyamaka reached Tibet, Shantarakshita built a version of the teaching closer to Bhavaviveka's. He kept developing logical argument as a tool for understanding ultimate truth. In the 14th century, Je Tsongkhapa wrote a new commentary on Madhyamaka that became the standard approach in Tibet. He favored Candrakirti's view but restored the place of logic that Candrakirti himself had downplayed.

  In the 19th century, the scholar Ju Mipham wrote a commentary on Shantarakshita's Madhyamakalankara defending the place of logic within Buddhist practice. He warned that dismissing analytical investigation and the inner science of pramana was, in his words, 'a terrible and evil spell.' The aim of that dismissal, he wrote, was to block a true understanding of the Buddha's own words.