There’s something strange about the verb “to know”. It is one of the most common verbs—in English and in other languages—and we are ordinarily very generous in our use of it. We freely describe ourselves and others as knowing all sorts of things on the basis of casual gossip and quick glances (just search your own social media posts or emails for the word “know…” to see how naturally generous you are).
Here’s the strange thing: with a tiny nudge, we can become very reluctant to say that anyone knows anything. Do you even know that there is milk in your fridge right now? Perhaps someone in your household has finished it up behind your back, or dumped out the rest on discovering it was spoiled. Or perhaps that unopened carton you bought yesterday actually just contains water, thanks to a manufacturing error at the dairy plant (this actually happened to me once). You might think that you could settle the matter once and for all by pouring yourself a glass and tasting it, but even here, you could wonder: Could this be a dream? Maybe the dream of someone who has actually run out of milk? And, even supposing you are now awake and tasting milk, do you actually know that you are? What does any of us really know for sure?
Philosophers have felt the attraction of skepticism for centuries, and across continents: there were skeptics in Ancient Greece, in China, and in the Classical Indian tradition as well. The sense that no one really knows anything is also easily triggered in modern populations. But why?
This year, I have been studying the mechanisms that govern our instinctive sense of who knows what. These mechanisms play an important guiding role in conversation. If we want to avoid telling people things they already know, or asking them questions they can’t answer, we need to have a strong sense of who knows what. Sociologists working in the field of conversational analysis have developed a theory of “epistemic territory” according to which we instinctively tend to see others as knowledgeable over a predictable field of topics. If someone is talking about something ‘close to home’—his job, his dog, what just happened to him, where he is now headed—then we will take what he says on trust, seeing him as knowing, unless we have a special reason to think that he might be mistaken this time. Recognizing other people as local experts helps us share the work of figuring out what is going on in our world: we don’t automatically trust everyone on every topic, but we can ask the person who has just come into the building whether it is raining again outside.
If we have default settings for assigning epistemic territory, we can also override these settings, in both directions. People can expand their territory by using arguments. Sometimes you know something that lies outside the range of what others would instinctively see you as knowing, and you can still share this knowledge with others by persuading them, on the basis of an argument that starts from premises they will take on trust.
Epistemic territory can also shrink: you will usually see me as an expert on what I am drinking right now, but if you happen to notice that someone has tampered with my glass while my back was turned for a moment, dumping the milk and pouring in white paint as a practical joke, then you would know more than I do on this close-to-home point.
But the skeptic who asks you whether you are sure that this is milk in your glass isn’t like the friend who has seen that a practical joke has been played. That friend really knows more than you do, while the skeptic just asks how much you really know, and not because the skeptic knows better than you do what liquid is in the glass. Perhaps in the hope of understanding knowledge itself, the skeptic decides to challenge a very close-to-home fact, deep inside the territory that most people would grant you instinctively, and asking you to prove what you know. You will be instinctively driven to argue, but you won’t be able to construct an argument that will satisfy the skeptic, because he can just keep challenging any premise you reach for.
Usually, human beings seek to pool their knowledge in conversation, cooperatively telling each other about the rainstorm outside or their plans for dinner, until everyone is on the same page. The sociologist John Heritage has argued that conversation naturally continues until the side that was seen as lacking knowledge on a point admits satisfaction: the questioner or unknowing audience has to grant that they have learned something (“oh, OK”). The skeptic’s great trick is to hold onto the conversational role of questioner no matter what happens, refusing to grant that he has ever learned anything from what you are saying. Like any other cooperative conversational partner, the skeptic is trying to get everyone onto the same page; the catch is, it’s a blank page. The challenge for philosophers is to sort out what this pattern tells us about knowledge itself, and what it tells us about the conversational and argumentative strategies we use to share knowledge socially.