How Do We Know What We Know? The Indian Schools of Pramana | Bharat Gyan

You are walking home at dusk and see a snake on the path. You slow down, step closer, and it turns out to be a rope. Who was wrong, and how did you find out?
Indian philosophers built an entire field around that question. They asked how we come to know anything, what makes a belief count as knowledge, and what to do when we get it wrong. The field is called pramana, and it is more argumentative than most summaries suggest.
Most of us hold beliefs we have never traced back to where they came from. Indian thinkers spent centuries asking exactly that: how do you actually know?
What does “pramana” mean?
Pramana means a means of knowledge: a reliable way of finding out that something is so. Each school of Indian philosophy lists the means it accepts and argues about what belongs on the list. That list is the real subject of Indian epistemology, so I will start there instead of with a general claim about “the Indian view”. There is no single Indian view.
Same question, different answers: how many pramanas are there?
The schools disagree on how many means of knowledge they accept. As far as I can tell from a standard reference (Wikipedia: Pramana), the common counts run like this.
| Pramanas accepted | Schools | What they add |
|---|---|---|
| 1 | Charvaka | Perception (pratyaksha) only |
| 2 | Buddhism (including Dignaga and Dharmakirti), Vaisheshika | Perception and inference (anumana) |
| 3 | Samkhya, Yoga, Jainism | Testimony (shabda) is added |
| 4 | Nyaya | Comparison (upamana) is added |
| 5 to 6 | Mimamsa, Advaita Vedanta | Presumption (arthapatti) and, for Advaita, non-apprehension (anupalabdhi) are added |
The counts are a convenient summary, not a settled fact, and some of them (the Mimamsa count in particular) depend on which thinker you follow.
The simplest way to see what is at stake is to ask what you would lose by cutting an item. A school that accepts only perception has to treat the word of a reliable friend, or the fact that smoke means fire, as something other than knowledge. Most schools find that too costly, and that is why the list grows.

How Nyaya reasons: the hill and the fire
Nyaya, the school most associated with inference, writes an argument in five steps. The standard example is a hill.
Proposition (pratijna): there is fire on the hill.
Reason (hetu): because there is smoke on it.
Example (udaharana): wherever there is smoke, there is fire, as in a kitchen.
Application (upanaya): this hill has smoke of that kind.
Conclusion (nigamana): therefore there is fire on the hill.
The load-bearing step is the third. It rests on vyapti, the invariable link between smoke and fire. If that link fails somewhere, the whole inference falls apart, which is why Nyaya also catalogues the ways a reason can go wrong, the hetvabhasa or fallacious reasons (Wikipedia: Nyaya).

Notice what this is not. It is not simply “deduction” or “induction” in the Western textbook sense. It builds the example into the argument, which is one of the interesting differences between the two traditions. I explore that in the differences between Indian and Greek logic.
When knowing goes wrong: the rope and the snake
Back to the path. A theory of knowledge has to explain error as well as success. The snake that turns out to be a rope is the classic example. Different schools explain it differently: as a mix-up of two real things, as the projection of something that is not there, or as a gap in what was perceived. I will not pretend there is one answer. In one familiar pair, Nyaya says a real snake seen elsewhere is wrongly placed on the rope (anyathakhyati), while Advaita says the snake is an appearance that is neither fully real nor fully unreal (anirvachaniya-khyati).
What they share is the idea that a perception can be corrected by a later one. You walk closer, you look again, the first impression is overturned. That mirrors how a careful person checks a first impression today.

Is a cognition trustworthy by itself?
There is a second, deeper argument. When you know something, does the knowing certify itself, or does it need something outside it to confirm it? Mimamsa is usually described as saying a cognition is valid in itself unless something shows otherwise, while Nyaya says validity has to be established by an external check. Philosophers call this svatah versus paratah pramanya. Buddhist and Vedanta thinkers took positions on this question too, and they did not all line up with these two.
It is the same question as “should I trust my first impression, or check it?”, and the tradition has several careful answers.

Four names worth knowing
Vatsyayana wrote the classic commentary on the Nyaya Sutras, the root text of the Nyaya school. His commentary, the Nyaya Bhashya, is usually dated to around the 4th or 5th century CE, though scholars’ dates vary.
Dignaga and Dharmakirti built the Buddhist theory that only perception and inference count. Their work carried into Tibet, which is a better route for the influence on Tibetan thought than the one the old post claimed. In Tibet their influence was deep: Dharmakirti’s works on valid cognition (pramana) were translated into Tibetan and became a core part of the monastic curriculum and its tradition of formal debate, where they are still studied today.
Kumarila Bhatta is a major Mimamsa thinker, usually associated with the wider list of pramanas. He is usually placed around the 7th century CE, and his school is said to accept six pramanas: the three above plus comparison (upamana), presumption (arthapatti) and non-apprehension (anupalabdhi).
Gangesha founded the later school of Navya-Nyaya, which sharpened the language of the debates. He lived around the 12th or 13th century CE, and his Tattvacintamani (“The Jewel of Thought on the Nature of Things”) became the basic reference for all later work in the school.
I have left out the Buddha, Nagarjuna, Shankara and Ramanuja on purpose. They matter to Indian philosophy, but I would not describe their views as “epistemology” without a source in hand. They deserve posts of their own.
Try it this week
Try this: pick one thing you believe strongly and ask which of these means of knowledge it rests on. Did you see it for yourself, work it out, take it from someone you trust, or simply assume it? Be honest about the answer; it is usually more interesting than the belief.
Where to go next
Get the next post in your inbox
One short email when a new post goes live.