You are talking about valid cognitions (pramanas), right? A valid cognition is any means of knowing, of understanding, that can ascertain the actuality or truth of something.
So maybe if we take the most important ways of knowing:
- sense-perception
- inference from sense-perception
- and the word or testimony of reliable others
Nagarjuna famously rejected all means of knowing, arguing for an extreme skepticism of no-thesis, non-knowledge. But Nagarjuna was still a Buddhist, and wrote of the importance of faith in the Buddhist path, ethical living, etc - so despite his extreme skepticism he did apparently accept the testimony of others (in the form of Buddhist scripture, the words of the Buddha, etc). Moreover, the fact that he used reasoning - in the form of tetralemma - to deconstruct any positive statement of knowledge, implies that he did accept some form of inference (or logical reasoning) as a valid form of cognising.
The other notable Buddhist approach to this - that taken by Dignana and Dharmakirti - accepted sense-perception and inference as valid tools of knowing, but rejected the testimony of others.
The testimony of others was rejected by them because no one can ever know for certain that another person - such as the Buddha - is telling the truth. The Buddha’s behaviour - for those who knew him - might give people confidence that he was authentic, genuine, honest, etc, but appearances can be deceptive, a person’s behaviour is not always a guide to their inner condition, and a person can change: someone who one could trust yesterday, may today be completely unreliable; they may be mistaken about what is happening now even though yesterday they were speaking the truth. So verbal testimony is out.
But sense-perception is undeniable. The senses can be confused (by optical illusions, drugs, intoxication, ill health, defect, etc), but the fact, the actuality of sense-perceiving is as close as we can ordinarily get to direct knowledge of the world.
Dignaga contrasted our essentially non-conceptual and instantaneous perception of perception-particulars (svalaksana), with the secondary knowledge we derive from these, in the form of generalised mental concepts or concept-universals (samanya-laksana).
Perception-particulars are always and ever only fleeting, momentary, but they are what is most real, most actual in our experience. Whereas the generalised concept-universals they give rise to in the mind - the abstractions and inferences we draw from these sense-perceptions - is always secondary, unreal (relatively speaking), non-actual, even though concept-universals make up the entirety of what we call scientific knowledge.
So we might say that an actual fire burns and cooks. If I see this actual fire with my eyes, my vision will be presented with an intense perception of colour and movement. If I touch the actual fire with my hand, my hand will get burned. (Btw, as a Yogacara, Dignana was agnostic about whether the actual fire exists, but the sense-perception of “fire” he took to be real, actual).
But if I close my eyes and walk away from the fire, the fire that I have in my mind (i.e. in my memory) does not burn or cook. The image that I have of it in my memory is a secondary echo, a mental generalisation of the actual fire, and so lacks true vividness, particularity, and presentness.
So for Dignana, despite the flaws that sense-perception can have, it is our most fundamental tool of knowing, and is in fact the basis for all other knowledge (i.e. science, logical inference, etc).
Interestingly enough, Dignana and Dharmakirti did accept one further epistemological tool: what they called “yogic perception”. By this they seem to have meant a special, non-ordinary form of perception (i.e. not through the senses, but something which can happen in meditation) that can perceive a thing as it truly is, without any division of a perceiver and a separate perceived.
We can see that there are some similarities here with Krishnamurti’s approach to perception. For Krishnamurti true perception can only take place when the perceiver is the perceived, or - expressed differently - when the perceiver is absent, and there is only perception.
More generally Krishnamurti accepted the place of science (and therefore inference generally), but only as a secondary, functional form of establishing knowledge. The only pramanas he really accepted were sense-perception (the sensitivity of the senses themselves) and awareness, attention, perception and insight (which he distinguished but did not fundamentally separate).
When it came to his own words, he did, it seems, accept - to some extent at least - the testimony of others. Because he said that if one could listen, completely, without any resistance, to the words being spoken from a person who has seen truth, then such listening can be liberating (i.e. constitute true knowing, true insight). But in other contexts he also said to doubt the verbal testimony of others, including himself, so this last pramana seems equivocal to me.
What do you think of all this?