One of the key debates in Indian philosophy is what counts as a pramāṇa: an instrument of knowledge, a “reliable warrant”, a means of knowledge reliable enough that one can be reasonably confident to take its conclusions as true. What counts as a pramāṇa? Many Indian philosophers will provide a numbered list of them.

In the empiricist tradition that remains popular in the West, boosted by the discoveries of natural science, only experience is admitted as a pramāṇa: to a full-blown empiricist, nothing counts as knowledge if it doesn’t ultimately have its roots in experience, based in some sort of direct perception. (Ken Wilber’s thought has come to take this position more and more over the years, to its detriment.) The debate over pramāṇas in modern Western philosophy is often framed as one between empiricism and rationalism. That is, where empiricists admit only experience as a pramāṇa, rationalists also allow reasoning an independent validity: some things can be rationally known a priori, independently of sense experience.

Some Indian philosophers have agreed with these views. It has been argued (controversially I think) that the long-dead Cārvāka materialist school, like the empiricists, accepted only direct perception (pratyakṣa) as a pramāṇa. Dignāga, the early Buddhist logician, admitted logical inference (anumāna), making him more comparable to rationalism as usually construed.

In India, the debate has gone well beyond one or two pramāṇas; some views argue for as many as six. But the most common formulation, first developed by the Sāṃkhya thinkers, says that there are three.

The third commonly stated pramāṇa, beyond perception and inference, is śabda, which literally means “word”. I take this to be important and true: we do get a significant portion of our knowledge from words. And not merely the words we use to make inferences, but – the meaning that śabda takes on in Indian thought – words we trust.

The mainstream of modern Western philosophy, beginning with Descartes but even more so with the empiricists, has avoided the idea of śabda as a means of knowledge. It has not ignored this idea, however; rather it has looked upon it with suspicion. For śabda would have kept it too close to the old medieval philosophies it was trying to get away from. In India as in the West, the paradigm of śabda was an old text that had come to be revered and held above others, “sacred”: the Veda in India, the Bible in Christian Europe. If not a text, then the word of people established as traditional authorities, the Church or learned paṇḍits.

And for the early moderns, this wouldn’t do. Hume saw in accepted “religious” views a catalogue of error, leading human beings to the “monkish” asceticism he thought self-defeating and counterproductive. Descartes was not so hostile to past tradition, but lived through Catholics and Protestants fighting destructive wars over what was, essentially, an epistemological debate. His hope was to avoid such debates entirely by establishing knowledge independently of authority.

Did Descartes and Hume succeed? In a word, no. Descartes’s attempted foundation does not stand; still less so the philosophical structure he attempted to build on it. His view, and Hume’s even more so, has proved enormously influential in part because of the rise of natural science and technology – finding new and effective knowledge derived from experience.

But hold on a minute. To what extent is scientific knowledge derived from experience – from human perception guided by instruments? That is clearly a part of it, but it is far from the only part. There is far too much scientific knowledge out there for any individual to ever have a chance of replicating it. Some of that knowledge must simply be taken on faith in scientists’ śabda.

Hume and his followers would point out that scientists’ śabda is valid because it is presumed to rest on experience. But it is important that in many cases we are not believing them because of our experience; we are believing them on a faith that the knowledge they promulgate comes out of their experience. The assumption is that theoretically we could go back and replicate those experiences themselves – but of course we don’t, and in many cases can’t. Even if I had the knowledge to interpret the results properly, I could not simply walk into CERN and replicate the experiments that led to the discovery of the Higgs boson or antihydrogen; I would need the connections and credentials to get in there. I certainly could not conduct those experiments without CERN’s equipment. And even those physicists who do have access to CERN will not have access to similarly sophisticated equipment in biology or astronomy. Knowledge outside their fields, they must take on faith.

Yes, that faith is at some level a faith in experience, a faith that scientists have had the experiences they say they have. But faith it nevertheless is, and that point is important. It is a fundamental aspect of human knowledge that it always depends at least in part on reliable authority – not merely pramāṇa but śabda.

The point is vital to note because there are so many realms of knowledge that do not and cannot rest on experience. In ethics, experience can tell us that some actions more than others are conducive to our happiness; but it cannot tell us that our happiness is that which we should be seeking in life. More fundamentally – I have mentioned this point several times before but continually turn back to it because I think it is the most glaring and fundamental philosophical error of our times – one cannot use experience to prove the validity of experience. There is a vicious circularity in this approach, one that usually proves self-contradictory. When one is trying to decide what counts as a pramāṇa, one will make a fool of oneself if one relies only on experience. One needs to rely on śabda: the words of others that become our own prejudices, our own “intuitions”, our own “common sense”. Of course the authority one thought reliable can turn out to be in error; Descartes and Hume were absolutely right to fear that. (I’ve certainly spent my share of time on Love of All Wisdom criticizing a reliance on common sense.) But what is important there is to use experience and reasoning as correctives to the errors of reliable authority, not to reject it entirely.