Pramāṇa: Where Knowledge Is Allowed To Come From

Among all these three means of knowledge, the knowledge derived from the instructions of the inspired comes first. After this, comes examination, with the aid of Observation and Inference.

Three separate claims are in those lines and all three are worth having.

Use all three. A diagnosis reached from one channel is not faultless, and he says so.

No single channel is complete. This is the strongest statement of the three and the most modern. It is the classical form of what an engineer means by not trusting a single sensor.

Testimony comes first in time. Not first in authority: first in order of use. You read the literature before you examine the patient, because without the literature you do not know what to look for. Charaka puts the argument as a question: what would one who has not been instructed succeed in knowing by examining with observation and inference?

Worked example: the same three in a modern system

Take a monitoring system that decides whether a web service is unhealthy.

Pratyakṣa. It measures response time directly from a probe. That is perception: contact with the object, reporting a particular fact.

Anumāna. It sees the error rate rising and the queue depth growing, and concludes the database connection pool is exhausted, which it cannot observe. That is inference, and it depends on a general connection between exhausted pools and those two symptoms. If the connection does not hold, the inference is wrong however good the measurements were.

Āgama. It consults a runbook written by the team that built the service, which says that this combination of symptoms in this service means the pool. That is testimony, and its reliability is exactly the reliability of the team that wrote it.

Now apply Charaka's three claims. A system using only the probe cannot say why. A system using only the runbook cannot notice something new. And the runbook has to be read before the probe is useful, because it tells you which numbers matter. His ordering is not piety; it is the right order.

What pramāṇa theory does NOT say

It does not say testimony outranks observation. Charaka's word is that instruction comes first, and his own explanation is that it comes first in sequence because it tells you what to examine. Nothing in the passage makes it the strongest evidence.

It does not make the list obviously three. Different schools close the list at different places, and Nyāya Sūtra 1.1.3's commentary enumerates them: the Cārvākas admit only perception; the Vaiśeṣikas and Bauddhas two; the Sāṃkhyas three; the Naiyāyikas four, adding comparison; and later schools add presumption, non-existence, probability and rumour. MU's three are the Sāṃkhya and Āyurvedic set. [The Four Pramāṇas of Nyāya, and Why Charaka Has Three] deals with the difference.

It does not settle what perception is. Nyāya Sūtra 1.1.4 imposes four conditions on perception precisely because "I saw it" is not a definition. The next chapter is about that.

Limits

The scheme has one structural weakness that a modern reader should see at once. It has no mechanism for revising the list. A pramāṇa is admitted or not, and a debate about whether comparison is an independent means of knowledge is conducted by argument, not by measurement. There is nothing in the apparatus resembling a calibration experiment.

The strength is the other side of the same coin: because the list is closed and public, an argument can be checked against it by anyone, which is what makes the debate rules of Module II enforceable.

Quick revision

• Pramāṇa: an admitted means of knowledge. MU names three: pratyakṣa, anumāna, āgama.
• Declaring them first is what makes fallacy detectable: a bad argument appeals to something off the list or fails a stated condition.
• Charaka: use all three; no single one suffices; instruction comes first in order of use.
• "By only one of these means of knowledge, knowledge does not arise of everything that should be known."
• Schools close the list at one, two, three, four, six or eight. Three is the Sāṃkhya and Āyurvedic set; Nyāya has four.

Test yourself

1. Define pramāṇa and name the three the syllabus prints.

A pramāṇa is an admitted valid means of coming to know something. The three printed are pratyakṣa (perception), anumāna (inference) and āgama or śabda (authoritative verbal testimony).

2. Quote or paraphrase Charaka's reason for using all three, and explain its modern form.

He says that with all three one should first fully examine a disease and the diagnosis then becomes faultless, and that by only one of them knowledge does not arise of everything that should be known. The modern form is that no single source of evidence is sufficient, so a reliable system combines direct measurement, inference and documented prior knowledge.

3. Why does Charaka put testimony first, and what does that ordering not mean?

First in order of use, because without prior instruction one does not know what to observe or what to infer. It does not mean testimony is the strongest evidence or that it overrides observation.

4. Give one weakness of pramāṇa theory as a theory of evidence.

It has no procedure for revising its own list. Whether a candidate means of knowledge is independent is settled by argument rather than by any measurement, so the set of admitted channels is fixed by dispute rather than by test.