Theories of Everything with Curt Jaimungal - Why Universal Skepticism Is Philosophy's Greatest Deception

Episode Date: April 19, 2025

Today we are joined by philosopher Jennifer Nagel for a take-no-prisoners look at universal skepticism—philosophy’s greatest deception. We unpack why doubt itself is the ultimate illusion, how knowledge is primitive instant recognition, and what this means for self, free will & consciousness. As a listener of TOE you can get a special 20% off discount to The Economist and all it has to offer! Visit https://www.economist.com/toe Join My New Substack (Personal Writings): https://curtjaimungal.substack.com Timestamps: 00:00 Introduction 01:28 The Nature of Knowledge 10:58 Philosophers and the Skeptical Mindset 16:57 Types of Skepticism 22:27 Exploring Knowledge Attribution 29:51 The Illusion of Knowledge 34:16 Knowing Without Knowing 38:10 Writing About Knowledge 46:10 Analyzing Knowledge 55:08 The Gettier Problem and Its Challenges 1:01:10 The Functionality of Knowledge 1:11:23 Collaborative Understanding of Knowledge 2:10:00 Understanding and Consciousness 2:26:32 Truth and Its Nature 2:32:16 Superposition and Contradictions 2:32:19 Conclusion Listen on Spotify: https://tinyurl.com/SpotifyTOE Become a YouTube Member (Early Access Videos): https://www.youtube.com/channel/UCdWIQh9DGG6uhJk8eyIFl1w/join Links Mentioned: - Knowledge: A Very Short Introduction (book): https://www.amazon.com/Knowledge-Very-Short-Introduction-Introductions/dp/019966126X - Knowledge and its Limits (book): https://www.amazon.ca/Knowledge-its-Limits-Timothy-Williamson/dp/019925656X - Very Short Introductions (series): https://www.google.com/search?q=a+very+short+introduction+to+series&sca_esv=3da4db664be6b3a1&ei=ypX6Z6flHsDniLMP2v2QkQk&ved=0ahUKEwin8oSB9tKMAxXAM2IAHdo-JJIQ4dUDCBA&uact=5&oq=a+very+short+introduction+to+series&gs_lp=Egxnd3Mtd2l6LXNlcnAiI2EgdmVyeSBzaG9ydCBpbnRyb2R1Y3Rpb24gdG8gc2VyaWVzMgUQABiABDILEAAYgAQYhgMYigUyCxAAGIAEGIYDGIoFMgsQABiABBiGAxiKBTIIEAAYogQYiQUyCBAAGIAEGKIEMggQABiABBiiBDIFEAAY7wVIqBRQxAtYwBBwAXgAkAEAmAFZoAGtAqoBATS4AQPIAQD4AQGYAgSgAocCwgIKEAAYsAMY1gQYR8ICDRAuGIAEGLADGEMYigXCAg0QABiABBiwAxhDGIoFwgIPEAAYgAQYQxiKBRhGGPsBwgIbEAAYgAQYQxiKBRhGGPsBGJcFGIwFGN0E2AEBwgIGEAAYBxgemAMAiAYBkAYKugYGCAEQARgTkgcBNKAHph6yBwEzuAf_AQ&sclient=gws-wiz-serp#wgvs=e - Time: A Very Short Introduction (book): https://www.amazon.ca/Time-Short-Introduction-Jenann-Ismael/dp/0198832664 - Laplace meets Godel: https://www.youtube.com/watch?v=ZB3tS7j7nNU - Flexible Goals (paper): https://onlinelibrary.wiley.com/doi/pdf/10.1111/cogs.13195 - The Legend of the Justified True Belief Analysis (paper): https://philpapers.org/archive/DUTTLO-3.pdf - Lay Denial of Knowledge for Justified True Beliefs (paper): https://philpapers.org/archive/NAGLDO - TOE’s Consciousness Iceberg: https://www.youtube.com/watch?v=GDjnEiys98o - Matt Segal on TOE: https://www.youtube.com/watch?v=DeTm4fSXpbM - Curt reads Plato’s Cave: https://www.youtube.com/watch?v=PurNlwnxwfY - David Bentley Hart on TOE: https://www.youtube.com/watch?v=NEAgVvW9i10 - Donald Hoffman on TOE: https://www.youtube.com/watch?v=CmieNQH7Q4w&t=1s - Iain McGilchrist on TOE: https://www.youtube.com/watch?v=M-SgOwc6Pe4&t=6326s&ab_channel=CurtJaimungal - Geoffrey Hinton on TOE: https://www.youtube.com/watch?v=b_DUft-BdIE - John Vervaeke on TOE: https://www.youtube.com/watch?v=GVj1KYGyesI&t=1s - Wolfgang Smith on TOE: https://www.youtube.com/watch?v=vp18_L_y_30 - Polymath’s Ai panel: https://www.youtube.com/watch?v=abzXzPBW4_s - Donald Hoffman and Philip Goff on TOE: https://www.youtube.com/watch?v=MmaIBxkqcT4 - Robert Sapolsky on TOE: https://www.youtube.com/watch?v=z0IqA1hYKY8&pp=ygUUY3VydCByb2JlcnQgc2Fwb2xza3k%3D - Curt debunks the “all possible paths” myth: https://www.youtube.com/watch?v=XcY3ZtgYis0&t=46s Support TOE on Patreon: https://patreon.com/curtjaimungal Twitter: https://twitter.com/TOEwithCurt Discord Invite: https://discord.com/invite/kBcnfNVwqs #science Learn more about your ad choices. Visit megaphone.fm/adchoices

To support sustainable food production, BHP is building one of the world's largest hot ash mines in Canada. Essential resources responsibly produced. It's happening now at BHP, a future resources company. You think you're watching a video, could just be dreaming, is this even really happening? There's all kinds of tactics you can put in play. Start making you feel like, yeah, actually, maybe I don't know anything. And that's very interesting. So there's this weird thing that some philosophers are very drawn to the pattern of intuitions you get when you get really self-conscious about knowing.

If you see your friend across a crowded room, you almost instantly recognize them. This is an effortless act of knowing. However, if I challenged you by saying, hey, are you sure that it's them? Couldn't it be a look-alike? etc. Then your certainty may evaporate. Professor Jennifer Nagel of the University of Toronto argues that the boundary between knowing and not knowing holds the key to understanding knowledge itself. As much as we love to say in our society that everything is an illusion, that we can't know anything, Professor Nagel's groundbreaking work argues that this universal skepticism itself is the most pestiferous illusion. It's the illusion of illusion,

in a sense. Today we talk about the alternative picture. Knowledge is less like constructing a proof – that's been the philosopher's game since Descartes, and more like instant recognition, a fundamental capacity. Knowledge itself is primitive. This has implications for your conception of self, free will, and consciousness, all of which we talk about today. Enjoy. What's the most significant unsolved problem in philosophy and why is it so consequential?

I think all of our problems are significant. Don't make me pick one. If I had to have some kind of ranking, I would say free will is right up there. Massive problem and it's one that you face at every moment of your life as an agent. My own skills are completely inadequate to solve that one. So I'm focused on something that actually might be related to it. My own work is on the nature of knowledge.

I'm very, very interested in the line between knowing something and failing to know. And I like to think that actually has some connection to the massive unsolved problem of free will. But the connection itself is something quite tenuous and we can get into it later. But I think if we were very clear in our minds about the relationship between knowing and not knowing, free will wouldn't be so daunting philosophically. So what is knowing?

Knowing is a state of mind of a type you can have only to the truth. Somehow I feel like that answer to your question might not have completely satisfied you, but that's the fastest way to describe it. Is this a controversial answer? Yes it is. Okay, so then it requires some more expanding. So why don't you talk about what is the consensus or what was the consensus? Why don't you talk about how did philosophers used to think about knowledge? How do they think of it now and how do you think about it?

So there is no consensus about this and there's no consensus about anything of significance in philosophy. There's some points in which we're at near consensus, but, and of course within logic, which is one branch of philosophy, there are some very well established results, but, so there's no clear consensus. I do think we as human beings are very, very sensitive

to the difference between knowing and not knowing. This is just a line that we're constantly navigating in everyday social exchanges. We have a very deep human familiarity with the difference between knowing and not knowing. You see the face of a, or you're playing tic-tac-toe, right? You have an instant sense of ease if you're in a domain where you're knowledgeable, and challenging confusion where you're not. I think any conversation that you engage in with another human being is really structured by your sense of what's known and what's unknown, right? That when you're speaking to somebody, you constantly

have to decide, am I going to tell or am I going to ask on some point of interest, right? I get to tell you if I think there's something I know that you don't. And when I ask you, what do I ask you? I ask you questions where I think you have knowledge that I lack, right? I mean, I might also maybe ask you pedagogical questions. I'm a teacher, you're a student. I'll ask you a known answer question

to see if you're up to the level. There again, I think I'm looking for, I'm evaluating the quality of your answer based on whether it's knowledgeable or not. And if you look at natural languages, English, French, Russian, Chinese, any language, you'll see a distinction between verbs that attribute knowing and verbs that don't. So we just automatically intuitively hear a clear contrast between a sentence like, Jane thinks that she's being followed, and a sentence like Roger knows that he's being followed.

There's a contrast there. If Roger knows that he's being followed, right? There's a contrast there, right? If Roger knows that he's being followed, it follows from that logically that he actually is being followed. If Jane thinks she's being followed, is she being followed? Maybe, maybe not. That formulation, or Jane believes she's being followed, that formulation marks out an attitude, the attitude of believing, which you could have either to truth or to falsehood. Knowing is an attitude you can have only to the truth. There's a way that philosophers have marshaled this contrast, philosophers and linguists actually. So attitudes of the type you can have only to the truth are called factive attitudes.

Factive attitudes, you can think of those as attitudes you can have only to facts. A fact is just a true proposition, as far as I'm concerned. Non-factive attitudes are attitudes you can have either to truth or to falsehood. So there's a lot of non-factive attitudes out there. There's assuming something. You can assume something that's true or you can assume something that's false. Hoping that something is the case. Strongly suspecting that something is the case. Being absolutely certain that something is the case. That's

a non-factive attitude because you can be absolutely certain that your partner is loyal and in fact they're not. So absolute certainty, that's a subjective state. And it's not something that's constrained only to link you to the truth. On the side of fact of attitudes, we have a great variety there as well. So if you see that the barn is on fire, it follows that the barn is actually on fire, right? You could think that you see something, but if you actually see that something is happening,

the thing has to be happening, right? It's like a success term, right? If you realize that your wallet has been stolen, it's going to follow from that that your wallet actually has been stolen. You could be strongly convinced that your wallet was stolen when in fact it isn't, you know, it's there in the bottom of your bag, you just can't find it. But realizing that something is the case is different from being convinced that something is the case. Realizing that is factive, noticing that is factive.

Those are all factive attitudes. And like, what's the place of knowledge here? Knowledge is the thing that all of these have in common. So I like to call it the mother of all factive attitudes, right? If you see that the barn is on fire, you know that it's on fire. If you realize that your wallet is stolen,

you know that it's stolen. And so on with all fact of attitudes, what they have in common is that they are ways of knowing. And so if you remember that something happened, then that thing actually happened. You can think you remember something that didn't actually happen, but if you really do remember that you had coffee this morning at breakfast, it's going to follow from that that't actually happen. But if you really do remember that, you had coffee this morning at breakfast,

it's going to fall from that, that you actually did have coffee. That's a natural language distinction. I can see you nodding as we're speaking. You recognize that distinction and it's there in all natural languages, that contrast. We're very, very comfortable with it. We're very, very familiar with it. We're very, very familiar with it.

We recognize how this works in everyday conversation. The thing that's extremely strange is that when you start thinking reflectively about knowing and actually, wait, what is it really to know something? Then your ordinary ease and familiarity with that line between knowing and not knowing melts away very rapidly because it's very swiftly possible to generate in people the impression

that actually they don't know a lot of things that they commonly describe themselves as knowing. And it's a very, very interesting question, why that's so. So I think when you ask me about how do people think about knowing, I think there's several camps here, right? So I think you could focus very much on what happens when we think really, really reflectively about knowing. Ordinarily, you want to say, look,

I know there's a water bottle here. I can see it. You can see it too. That just seems like really fling common sense. If you look at how people use the word in everyday contexts, they're using it very generously. Do you know who got the job? Yeah, I do. I just saw a post on social media. Then the interesting thing is that if you get really reflective,

wait, do you really know who got the job? Let's think twice. Maybe that guy was faking, and a social media post is- It's fake news. Yeah, fake news. Then you're like, well, maybe I don't know. Do you really. It's fake news. Yeah, fake news, right? Then you're like, well, maybe I don't know. Do you really know there's a water bottle here? Well, there's a familiar variety

of skeptical arguments we could pull out at this point. Well, actually, you think you're watching a video, could just be dreaming. Is this even really happening? Is it a water bottle? Is there some other liquid in there? There's all kinds of tactics you can put in play. Is this even really happening? Is it a water bottle? Is there some other liquid in there? There's all kinds of tactics you can put in play or alternate scenarios that you can imagine.

Start making you feel like, yeah, actually, maybe I don't know anything. And that's very interesting. So there's this kind of weird thing that I think some philosophers are very, very drawn to the pattern of intuitions you get when you get really self-conscious about knowing. So, René Descartes is in that category. You know, let's pull up our socks and stop, you know, going along with what we learned in school.

Let's wipe the slate clean, start over again, be extremely self-conscious, try to restrict ourselves to clear and distinct ideas. And I think he ends up with a vision of knowledge that's actually really strange and really out of line with the way we ordinarily use the word. Then there are some philosophers. So those philosophers are, I think, skeptical ultimately in their temperament. Descartes was not ultimately a skeptic. He thought he had a fantastic rational argument

to get you out of skepticism. I'm with a lot of contemporary philosophers and thinking that argument actually doesn't work terrifically well. Another way you could go is to take our everyday ease with the concept for granted. And you could stand with G. Moore, who just says, do I know that the external world exists?

Sure, I do. Here's one hand, here's another, just look at it. And Moore's attitude to the skeptic is an attitude of ridicule. It would be ridiculous to suggest otherwise. Of course, we know there's a water bottle here. It's just absurd to say that we don't know that. And that also seems in a weird way kind of unsatisfactory for very

different reasons, right? Like he's just kind of saying, oh, it's just nonsense to have these, you know, skeptical inclinations. And I think the skeptical inclinations are natural, right? They seem really compelling. They look like arguments that we are very hard pressed to rebuff. So just to say, oh, it's just silly. That seems to me like kind of an evasion of philosophical labor that we should be performing. So in the 20th century, with the turn to ordinary language

and heightened attention to usage, some programs emerged that tried to make sense of both the kind of low standards, come on, we know this water there, and the super high standards, like, actually no one knows anything, with the possible exception of the fact that you exist, maybe. And if you try to make sense of both of these things at the same time, you can do it with some really kind of fancy pants semantic theories, which say, for example, maybe the verb to know works like a gradable adjective like tall, where, you know, what is it for something to be tall?

Takes a lot more height for a skyscraper to be tall than an office chair, right? But you can still meaningfully say that something is a tall office chair. You know, you have kind of implicit comparison class in mind. And maybe the guy in the street and the skeptic are just kind of using different implicit comparison classes when they're evaluating people as knowing and not knowing different things. So the guy in the street has a pretty low standard. The guy in the skepticism seminar has an extremely high standard.

So just as the same guy could be tall for an American man and not tall for an NBA player. The very same instance of somebody looking at the thing and judging that it's a water bottle, that could be knowledge by the standards of the street and not knowledge by the standards of the street and not knowledge by really high standards. So that's what the contextualists say, starting with Gale Stein.

And Gale Stein actually had this really interesting thought that it's an essential feature of the concept of knowledge, that it picks out these different relationships in different contexts. It looked great for a time. And actually, if you pull contemporary philosophers right now, epistemologists, contextualism remains an extremely popular position. It has a lot of technical difficulties, though. It has difficulties explaining what goes on when I evaluate someone else's claim to know

that someone else does or doesn't know something, because if we're all using these different implicit comparison classes, you get into sort of monster problems transporting claims about knowledge across conversational contexts because the whole thing is like, well, which standards are we supposed to be using at this point? And it always seemed like one of the old beauties of the concept of knowledge is, if you know something, then you can take it with you wherever you want to go, right? If you know something in math or physics, you don't have to keep asking, are the standards

high or low over here, right? You want that kind of portability of knowledge. You want that solid result. And so contextualism seems to give up on that. I think it also is not an adequate response to skepticism. So, you know my strategy? You wanna hear my line? Okay, so my strategy is when we can get in there

and try and figure out what knowledge is, an irresistible and attractive and important source of guidance is our everyday sense of what knowledge is. But this sense, which produces kind of like intuitive responses to various cases, like you know, intuitively it seems like we know these things. It produces weird, slightly paradoxical patterns. The way I see it, you've got the skeptics who are taking very seriously

that hyper self-conscious set of- When you're using the word skeptic or skepticism, is that a synonym for just doubting, doubting your experience, doubting your propositions? Good question. There's a bunch of different types of skeptics. The word skeptic just comes from the ancient Greek for questioning. I want to separate out a couple of different types of skeptics. The academic skeptic is the one who says,

humans can't know anything, or they can't know anything of substance, and I can tell you why this is so, right? And those skeptics often have like a positive theory of knowledge and a set of arguments trying to show you why human beings fall short of the standard required for knowledge. Academic skepticism is often pretty easy to refute because you can self-apply the theory, right?

You can say, like, how can you as an academic skeptic actually, you know, know the truth of these conclusions that you're arriving at? That's a little awkward for them. They're not unaware of that challenge, but I think it remains a good challenge. Peronian skepticism is actually the much harder kind of skepticism to address. So the Peronian skeptic is the person who just keeps questioning, who's just after a certain kind of suspension of judgment. Anytime that you say something, they will ask you, well, how do you know? Or they'll give you some kind of alternative possibility to consider. The Peronian skeptical tactic is a tactic that can result in you never losing an argument,

right? Because as a Peronian skeptic, you never accept a claim. You never make a claim. You only ever question. And that's dialectically an incredibly powerful technique. Of course, the Peronian skeptic never wins an argument either because there's nothing that they can positively establish. Can you turn it on the Peronian and say,

well, what is it exactly that you mean to know when you're asking me, do I actually know? So that is an A plus move to try with any Peronian skeptic, turn the tables, ask him a question, put him in the position of answering. The smart Peronian skeptic does not answer your question with a declarative. The smart Peronian skeptic answers your question with a question. So we'll say, well, what exactly did you mean by, and they'll pick out some term in the question that you've just asked.

Like, if you think of your job as a proning skeptic is never utter declarative, just always question mark at the end. Doesn't matter if someone else asks you a question, you can just keep turning tables right back. Or you can, or you could always give them some kind of alternative like, well, if I were to think about it this way, if I were to think about it like Aristotle,

I would say this, and if I were to think about it like Plato, I would say that. But who knows which is right? That's another skeptical. So what I would say to them if they then turn the question back to me is I would say, okay, just by you asking that, it sounds like there's something that would satisfy you and something that would not satisfy you. Otherwise, you wouldn't be asking the question. So why don't you give me an idea as to what would satisfy you? What would such an answer look like? Just a moment. Don't go anywhere.

Hey, I see you inching away. Don't be like the economy. Instead, read The Economist. I thought All the Economist was was something that CEOs read to stay up to date on world trends. And that's true, but that's not only true. What I found more than useful for myself personally is their coverage of math, physics, philosophy, and AI.

Especially how something is perceived by other countries and how it may impact markets. For instance, the Economist had an interview with some of the people behind DeepSeek the week DeepSeek was launched. No one else had that. Another example is the Economist has this fantastic article on the recent dark energy data, which surpasses even scientific Americans coverage, in my opinion. They also have the chart of everything. It's like the chart version of this channel. It's something which is a pleasure to scroll through and learn from. Links to all of these will be in the description, of course. Now, The Economist's

commitment to rigorous journalism means that you get a clear picture of the world's most significant developments. I am personally interested in the more scientific ones like this one on extending life via mitochondrial transplants which creates actually a new field of medicine. Something that would make Michael Levin proud. The economist also covers culture, finance and economics, business, international affairs, Britain, Europe, the Middle East, Africa, China, Asia, the Americas, and of course, the USA. Whether it's the latest in scientific innovation or the shifting landscape of global politics,

The Economist provides comprehensive coverage and it goes far beyond just headlines. Look, if you're passionate about expanding your knowledge and gaining a new understanding, a deeper one, of the forces that shape our world, then I highly recommend subscribing to The Economist. I subscribe to them and it's an investment into my, into your intellectual growth. It's one that you won't regret. As a listener of this podcast, you'll get a special 20% off discount. Now you can enjoy The Economist and all it has to offer for less. Head over to their website, www.economist.com slash toe, T-O-E, to get started. Thanks for tuning in, and now let's get back to the exploration of the mysteries of our universe.

Again, that's economist.com slash toe. So why don't you give me an idea as to what would satisfy you? What would such an answer look like? Do you want me to roleplay per sure? What an interesting question. I wonder what satisfaction even is. Do you have ideas about the nature of satisfaction?

I want to actually like pull back for a second. We can give that just guesses. But I felt like I was like halfway through this enormous trajectory towards sort of situating myself in the landscape of epistemologists. So I think there's a way that knowledge looks to us when we get hyper self-conscious about it and there's a way that knowledge looks to us in our ordinary dealings. And we get different and in fact paradoxically weird contradictory impressions of knowledge when we look at it one way and when we look at it the other way. And contextualists try to take on board all of these different contradictory impressions

and make up a theory, very complicated theory, saying knowledge itself is moving all over the map. So contextualists don't think there's any one thing that counts as knowledge. There's a bunch of different relationships that agents can have to propositions, some of them requiring a lot of evidence, some of them requiring very little evidence. There's a whole scale there, and they think you're always just floating up and down the scale, very complicated position. Another way you could go is you could say, let's stop treating epistemic intuitions,

that is, your sense of, wait, do I know this thing? Let's stop treating those intuitive verdicts like they're coming from an oracle or from, let's instead, so I think sort of the skeptics listen to just the intuitions generated from the hyper self-conscious mode and the kind of casual guys like GE Moore listen to just the intuitions generated in the casual mode. Contextualists try to listen to it all and make sense of it all at once. And what I want to say is let's stop treating epistemic intuition as some kind of oracle, or even an oracle that we selectively attend to and disregard in some cases when it's awkward for our theory.

Let's instead break open the black box and try to figure out what is going on in us to generate these epistemic intuitions. I think there's actually something parallel that we've done in science, right, that your perceptual systems were not optimized for a scientific investigation of reality. Your perceptual systems were optimized for the survival of mammals in our lineage, right? And your perceptual systems have deliverances that you cannot take naively at face value, right? And your perceptual systems have deliverances that you cannot take naively at face value, right? Conflicting deliverances, like in the waterfall illusion,

you can have something that simultaneously seems to be standing still and moving, right? That can't be happening. And we don't try to just sort of arbitrarily ignore our senses some of the time. And we better not try to just, of arbitrarily ignore our senses some of the time. And we better not try to just, oh, let's have a scientific theory that takes all of these deliverances at equal value. What we start to do is we start to try to achieve a better understanding of our perceptual

systems. And that enables us to selectively interpret them. So our perceptual systems actually register a big difference between what in the ancient times you might have called motion and rest, right? But we now don't think, like relative to you, right? But we now don't think that there's necessarily such a fundamental distinction there, and that required kind of getting over ourselves and figuring out like how are we situated in the world and how are our perceptual systems working and having a more critical understanding

of those systems. So I think a lot of what my work tries to do is it tries to look at how are our intuitive impressions of knowledge and its absence being generated within us. And if we can, both in just everyday conversations, we're asking each other questions, we're playing games of skill or games of chance, we're meeting friends, we're meeting strangers. We have all these, these senses of,

oh, I know that guy, no, I don't know that guy, right? How is that happening? And how is it happening also when we're being more self-conscious, right? The more you understand about how that machine is working, the smarter you can be in sort of interpreting its deliverances and making sense of what knowledge actually is. So that's my strategy is to have this sort of like self-conscious understanding of the nature of epistemic intuition.

So earlier when you said that John knows he's being followed versus Sally thinks she's being followed and John realized that he left his watch at home or what have you. That sounds to me like a narrator in a book, like a third person can say this from a God's eye point of view, but John can't actually say or he can, but he could be false or he could be mistaken, like we mentioned about the skeptics, that he knows that he's being followed. He can think that he knows he's being followed, but whether he knows he's being followed. We would say that if he actually is being followed, there's a necessary condition, not a sufficient one.

Yeah, you got that. Sounds good. It better actually be the case that he's being followed if he's going to know that he's being followed. So when people are inquiring about what is knowledge, it sounds like they want to know how can a third person say that he knows for sure. Whereas it sounds like what you're doing is John feels like he knows he's being followed and you're investigating the feeling route? So I am absolutely interested in the objective question of what it is for John to know that he's being followed. I'm absolutely interested in that. That is the central thing that I'm interested in. That first thing I said, that knowing is

a mental state of the type you can have only to the truth. That's a characterization, not of how things look from John's perspective, but that's characterization, quote unquote, from the point of view of the universe. That's an objective characterization. I am very interested in the question of how knowledge detection, knowledge attribution systems work, because I think it's only by understanding the strengths and weaknesses of those systems that we can start to sort out the paradoxical bizarreness and figure out what knowledge, like how to understand knowledge itself.

One of the things we want to do is we want to explain away, for example, the intuitions that drive us to skepticism. So I've done some work collaboratively with psychologists where we have people in the lab and they read short vignettes describing a situation and then they have to say whether the key character in the situation either does or doesn't know some simple fact like the fact that the table is red. And we manipulate different details of the story

and we watch them switch down from saying, yeah, John knows the table is red, to saying, oh, he thinks he knows it, and maybe he doesn't really know, right? We're very, very interested in what's driving those changes, and that's something you can investigate psychologically. Skepticism in the sense of becoming very reluctant to describe someone as knowing something,

it's not, as Hume suspected, a disease of the learned that you only get if you read a lot of books or something like that. Skepticism is a natural human impulse. You can produce it in any undergraduate in 30 seconds in the lab, right? Just lay out some scenarios and they'll be like, oh yeah, well, I don't know anything, right? And that's a very interesting phenomenon. And I think if you're going to try to explain what knowledge is and what place it has

in the life of intelligent biological creatures, you have to be able to explain what the hell is happening in those, like, apparently naturally generated skeptical judgments that people will make, right? You have to explain, like, why does that happen? Why do we get that impression? I think it's an illusion, right? Like, so I think we know there's a water bottle here. And if I run through some tricks that get you into a place where you start feeling like, damn, because I don't know that, or if you're watching a video and you start thinking, maybe I don't know that I'm watching a video because this could all be a dream, right? I'm messing with you.

I'm presenting an illusion. It's a cognitive illusion. It's not a perceptual illusion. And what you need as self-defense is you need actually a deeper understanding of the mechanisms that generate impressions of knowledge to appreciate the nature of this illusion, to sort of see for yourself, oh, okay, that's how it's getting generated. And I can put it in its place. And that's, you know, that's something that it takes a certain amount of work,

actually right at the border between philosophy and psychology, to sort out. So it sounds like you're flipping this illusion, this tendency in our culture to say that so-and-so is an illusion, like we'll say the self is an illusion or that water bottle is an illusion on its head. Where there are some things that if you classify them as an illusion, you're saying, well, that itself is the most pernicious kind of illusion. Illusions are natural phenomena, and we can explain them, right? There's lots of illusions that we understand pretty well. Some of them are pretty difficult and mysterious, but they're not beyond the scope of scientific understanding. Can you know something but not know that you know it?

Absolutely. Yes. In fact, you'd better be able to know some stuff without knowing that you know it. Because we're off to vicious regress land if knowing always requires knowing that you know. Because then, of course, knowing that you know, it's going to require knowing that you know that you know. So 100 percent is possible to know without knowing that you know, because then of course, knowing that you know is going to require knowing that you know that you know. So 100 percent is possible to know without knowing that you know.

There's a number of excellent arguments on this point in a temporary philosophy. Can you believe something without knowing that you believe it? Yes, I think so too. That's actually a little bit more controversial, but I think absolutely. Yeah. So I'm following Oxford philosopher Timothy Williamson here.

He thinks that our mental states are not transparent to us. And the idea that they were is sort of a confusion. That is, for any nontrivial mental state, by which I mean a mental state that you're sometimes in, sometimes not, it is possible to be in that state without knowing that you're in that state. Okay. Let's go back to the knowing without knowing. There's never knowing without knowing. That would be logically contradictory. You can't know that the key is in the lock

without knowing that the key is in the lock. That's trivial. You can know that the key is in the lock without knowing that you know that the key is in the lock. Right. It's iterations of knowledge, so repetition is doubling up. That's not guaranteed. It's often the case that you not only know, but know that you know. What I'm arguing is, what I believe is, that it is not always the case that every time

you know, you have to know that you know. So give us an example. So I can't actually give you a first person present tense example because counter examples are elusive. Do you know elusive counter examples? So if I were to say, oh, there's this fact that I know, but I don't know that I know it. I would be, I'd be in a really, really strange place, right? Because I'd be trotting out a fact. Here's an example of something that

I know. And it's actually a norm on assertion that anything that I assert to you, I better actually know it to be the case, right? And so I'd be kind of violating that norm even in just saying, here's a fact that I know but I don't know that I know it, right? So first person, present tense, you can't do it. But you can demonstrate that they are there, right? And this is, it's like a non-constructive proof in mathematics where you can demonstrate that there is some X which has these properties even though you can't locate the X.

Third person, however, you can pick out cases pretty easily. This is just drawing on Williamson again, but what you have to do is you have to look for borderline cases of knowledge. Borderline cases of knowledge are going to be cases where you have enough, let's call it epistemic strength, to know, but not quite enough to know that you know on the given point. And I could run through the argument here. It's chapter four of Williamson's 2000 book, Knowledge and Its Limits, called an anti-luminosity argument.

But I think I need a whiteboard. I'm going to have to have a lot of subscripts. And it really trades on issues about intransitivity of indistinguishability. So here's like one of the basic ideas there. Imagine that I show you a series of color patches. And they are so close in color to each other that between patch one and patch two, naked eye, you cannot tell the difference, right?

I offer you a million dollars, which one is darker? You're like, I don't know, right? But they're very, very slightly different in color, right? So one and two are indistinguishable. Two and three also indistinguishable. But you see where this is going, right? I go through a whole bunch of patches and

one versus a hundred clear difference, you can tell them apart, right? So these first two patches are indiscriminable, indistinguishable, but they're not the same. There's not the same state in both. The relationship of indiscriminability isn't transitive. So this can be indiscriminable from that, that can be indiscriminable from this third thing, but these two things might be really discriminable from each other. So the argument is kind of running on that

logic to show you that you do have instances when you get out to the borderline between knowing and not knowing, where you're going to run out of iterations. You're going to be, as you go from knowing to not knowing, you're going to hit that gray zone in the middle, where you know, but you're not in a position to know that you know. Speaking of knowing, people should know that there's a series called A Very Short Introduction To, and it's a philosophical series. It's fantastic to get a short introduction to almost

any topic in philosophy outside of like a Wikipedia article or the Stanford Encyclopedia. By the way, have you read Jen Ann Ismael's A Very Short Introduction To Time? No. That is my favorite in that whole series. But she's coming up on the podcast shortly. Oh, fantastic. She is great. on the podcast early fantastic. She is great Oh, she is really good. I love her work note There's a podcast with Jananne is male coming up shortly

It may already be out so you can either subscribe to get notified or click the link in the description Jananne is an incisive Explicator who excels at explaining and thinking clearly and I think she has Absolutely exquisite level of insight into some very fundamental philosophical problems about, for example, the relationship between us, the models that we're building, and the larger universe that we're trying to reflect, right? She has a sort of very deep insight into the fact that when you build a model of the universe, you shouldn't forget that you're inside there, you're not on the outside.

She's done some really excellent work tracing the philosophical consequences of that because it's so easy to make your toy model of things and kind of reify it and think that you can see it all from the outside. Yeah, there's a lecture she had called The Plasmas Demon Meets Girdle about that. Yeah, yeah, okay. And I'll put a link to that on screen and in the description.

But the reason I was bringing that up is that the very short introduction to knowledge is by Professor Jennifer Nagel. Yes, yes, yes, I wrote that. I wrote that. It was actually quite hard to write. When you write these short introduction books, it's like writing a Hardy Boys series book. There's all these rules, like you write a Hardy Boys series book,

there's all these things that Joe and Frank can and cannot do. So you're scripted. The big thing you cannot do, you're not allowed in those very short introduction books, is you're not allowed to complain about the fact that it is short. You're so tempted as an author to say, I'd love to get into this, but unfortunately with the space pressures on me, I can't.

I 100 percent get why they did that because it would be very disappointing as a reader to think always, oh, there is some bigger truth out there, but you're not getting it because you just bought the light version, right? But as an author, you always have this tantalizing sense of wanting to say so much more. You have to have everything very, very compressed and crystallized and very, very short.

I actually thought, so there's a couple of pages on contextualism in the thing, and I thought that Keith Rose would kill me, like, next time he saw that, like, best living contextualist. But actually, people have been pretty nice about it. I think they understand the pressures that you're under. You know something weird that I discovered when I was writing this, and I don't know if you've had the same experience as well, but when you're trying to explain something in a really short span, maybe to people who don't have a lot of academic background, do you ever find yourself reaching for metaphor?

Yeah, I do too. And I think there's something very interesting about that because if you'd asked me before I wrote this book, I would have thought metaphors are essentially decorative and beautiful to sprinkle them in when you have a little extra space and you can give your reader that bouquet of roses. And I came to think, no, metaphors are extremely functional because I think you're eliciting some kind of model that your audience understands already and bringing it into the service of making a new patch

of territory inhabitable for them. I feel like I'm using metaphors and I'm mixing them too, which is really a catastrophe. Douglas Hofstadter, you know Douglas Hofstadter. Of course, of course, yeah. I mean, not personally, but I know his work. This book is called Surfaces and Essences, I believe. I have not read that one.

It's about analogy at the core of cognition. So his thesis is that metaphors or analogies are how our mind works and how we think. That might be going too far, but yeah, I will look for that for sure, for sure. I think it's certainly a lot of thinking is analogical. I don't know if it's a general purpose model. Tell me what else about writing the very short introduction was interesting. Not using inline citation. That's not that interesting.

That line is just for academics out there. Well, I don't mean like the Bible from the very short introduction people that tell you what you must and must not include. I just mean, even in the process of writing it, did you come to some insight that you didn't have prior? Did you realize something, realize there's that word? Oh, that's an interesting question.

I feel like it's a few years back. So I'm not as fresh as I should be on what that process did for me. I do feel like in general, the more I explain things, the better I understand them myself, especially if I have to do a very compressed explanation. And I do think that writing that book helped me concentrate on the problem of epistemic intuition as something that's very much lying at the heart of the difficulty in epistemology. The difficulty in epistemology, the academic study of knowledge, is that we have an embarrassment of riches. We have all these incredible impressions of the presence and absence of knowledge, and you just can't make them all add up, not all at once. Nobody is going to draw a smooth

curve through all those data points. So you've got to get rid of some of them, and you absolutely don't want to do that arbitrarily, right? But what are you going to give up and how? Like you want to have a principled explanation. You don't want to just say, oh, well, here's my favorite theory and everything else is noise, not signal, right? You want to have a theory which says, these particular intuitions are coming from this particular heuristic, which we have good independent evidence is operative in us.

You want to have some story about what you're doing with the data. You certainly don't want to end up overfitting, right? You don't want to end up drawing a line that could cross us through every single data point because it would just be a mess. So if I'm understanding correctly, when you're trying to understand a term, you ideally want a definition of the term. There are a couple ways you're trying to understand a term, you ideally want a definition of the term. There are a couple of ways you can go about generating a definition.

One is in math, it's just necessary and sufficient conditions. Also in philosophy, we want necessary and sufficient conditions if we're defining a term. If you were trying to define knowledge in terms of simpler categories, good luck to you. You could try to give an analysis of it, breaking it down into simpler components. That was the project of analysis which really got going

in the mid-20th century in Anglo-American philosophy. There was the idea that knowledge could be explicated in terms of true belief of a certain kind. So one of the very first things you discover when you look at knowledge, and virtually all epistemologists through history are very clear on this, is that knowledge is only of the true. Something that's very, very widely established is that knowing entails believing. There's a few renegades who disagree with us.

I think they're all making mistakes of one kind or another. So if you go back to the core examples that motivated us, and I've already forgotten the character names. I think one of them is Roger. Roger knows that he's being followed. Jane believes that she's being followed. What's the difference between Roger and Jane? If we can explain that, we're doing so well.

One of the things you can say about Roger, who knows that he's being followed, it is also true that Roger believes that he's being followed. Does he believe this? Yes, he does. Sometimes that point is a little obscured for us because we don't usually describe people as just believing things if they not only believe but also know, because we tend to make them more informative statement rather than the lesser informative or weaker statement. That's for pragmatic reasons. But I think it's pretty clear if you ask, does he believe he's being followed? And he actually knows he's being followed. The only acceptable answer is yes. Yes, he

does. So knowing is believing and what is believed has to be true. And then we've got to have further conditions. So already in ancient Greek philosophy, in Plato's Theotidas, there's the suggestion that you've got to add something else, maybe. They fool around with the idea that knowledge is true belief with an account. You know, the person who just happens to point down, you know, you ask him, which way is it to Larissa? And he just randomly points one way or the other.

If he's pointing you actually in what happens to be the right direction, doesn't necessarily know. If he's the frequent traveler who always goes there and he points, he might point the same way, but there it looks like he does know. What else does that guy have? He's got, so this is actually getting into a different dialogue, which I will now pivot away from and leave that example behind.

He's got to have something else. Sometimes it might look like what the knower has is he has an account. He has something he can tell you. He can give you some reasons or some justification or some explanation. He could say, Larissa's that way. I just came from there myself. I know this way very well, right? Where the person who's just randomly pointing can't honestly say those things. It won't be true of him that he has these things to tell. But if knowledge is true judgment with an account, if an account is just something you could say, even the joker is just pointing randomly,

has something he could say. So it seems like it has to be something more than that. And the theotetus is just a really interesting dialogue in getting into the difficulty of articulating the something more that you have to have beyond true belief in order to constitute knowledge. What keeps happening in that dialogue is the something more has got to be some kind of account and not just something you could say, you know, it'd be really great if it was a knowledgeable account. That would be cool. But of course, that's inadmissible in your definition because you set out to define what is knowledge. Knowledge is true belief with a knowledgeable account. Now you've got a circular definition because you're using the very term that you set out

to define. That's no good. And philosophers just kept trying to come up with reductive analyses of knowledge in terms of true belief plus further factors like justification, evidence, accounts, and so on. A lot of question marks about those further factors, and it's pretty easy to come up with counter examples. So should I talk about the Gettier problem? Because that's a really famous, yeah? Yes.

Famous class yeah? Yes. Famous classic counter examples. Okay, so you might think that what distinguishes the person who knows which way it is to Larissa from the joker who's just happening to point you in the right way, is that the knower has a true belief which is justified. So that's called the JTB, justified true belief, the classical analysis of knowledge. There were several mid-century figures in the 20th century who were advocating that kind of account.

Plato toys with something like that in the Mino, discusses it in the Theotetus, does not endorse it in either place. Julian Dutin has a really great history of the JTB analysis of knowledge, which I will recommend to your readers, your readers, your viewers. But here's a problem with the JTB analysis. It seems to be too weak as an analysis. It lets in examples and counts those knowledge when it really shouldn't. Here's an example of Bertrand Russell's that's very simple.

A guy is, let's say, walking through a train station and he wonders what time it is. And he glances up at the clock and the clock shows that it's five past three. It's like, great, okay, it's five past three. And it's true. It's right. It is five past three. And by low standards, we're going to count this guy as now knowing that it's five past

three, right? And I see you nodding, but if anyone's not satisfied with that nod, there's actually empirical work on this. So some work that I've done with Raymond Moore and Valerie Sanjuan. We took undergraduates and gave them puzzles like that. And yeah, almost everybody counts this guy as knowing that it's five after three. But, and now there's like a plot twist in the story. Let's say that this clock is broken and the hands of the clock have not moved.

It's just been sitting there pointing to five after three for the last 48 hours. If our guy stood there for a while and looked at it, he kind of realized it, but he doesn't. He just glances at it and walks on. Now, when we ask again the question, does this guy know that it's five after three? At least a healthy portion of our experimental participants say no. No, he doesn't know. And it seems like this guy is kind of justified in thinking that it's five after three. I mean, he looked at a clock. That's the typical way that you make up your mind about the time, right? He's not just arbitrarily

saying something the way that the guy who does not know the way that Larissa is just arbitrarily pointing in what happens to be the right direction. He's got some evidence. He's got what you might want to call justified true belief, but it seems he doesn't know. Incidentally, problems of this type were explored in South Asian philosophy centuries before the Anglo-American tradition got to them. So perhaps as early as around 530 of the Common Era, you have Prasashtapada looking at these cases. You get subsequent figures in Sanskrit tradition articulating them in more detail and getting into them. I'm going to give you an example from the 8th century from Darmotira that, oh, and I recommend my colleague, Gennard and Gennari's work on this.

It's great. But I'll give you Darmotira's example because it just takes a second and it's a beautiful example. So you have a desert traveler and he's thirsty and he's on top of a hill and he's looking down into the valley and it seems like in this valley ahead of him there's water, thank God, water at last. And so he forms this judgment like there's water in that valley and runs down into the valley.

Now actually, so there's going to be two plot twists. Plot twist number one is he's hallucinating. It's just a complete mirage that he is experiencing. But if you make it like really, really compelling hallucination, we might want to say he's justified in thinking that there's water there. In some sense, it's justified, although the South Asian tradition had some doubts about that. But, and then plot twist number two is,

coincidentally, there actually does happen to be water in the valley. And you could even have it when the guy runs down to the bottom of the valley, he finds it there and there's a well that's partially covered by a rock, and he can have a drink. And when Dharmata represents this example, he's presenting it actually as a counter example to the theory that knowledge is that state of mind that underpins successful action.

And it's a very effective example, a counter example to that theory, right? Because this guy is acting successfully, he's thirsty, he finds water. He's thirsty. He finds water. He is making a judgment there's water in that valley, which is in fact true. And it's based on what seems to be great evidence. So it's a really interesting question, like, why the guy, when he's standing at the top of the hill, just hallucinating. We don't want to say he counts as knowing, oh yeah, he knows there's water there. He doesn't, right? There's something about the connection. His connection

to the truth doesn't seem to be of the right type, doesn't seem to be of the right kind to count as knowing. And it's such a deep question in philosophy then, like, okay, what is the right kind to count as knowing? What is this guy missing? That's called, I mean, we in Anglo-American philosophy now call that a Gettier case after the philosopher Edmund Gettier, who in 1963 wrote a very short paper just identifying this problem with the classical JTB analysis and raising a couple intuitive counter examples to it. And then a lot of philosophers came along

and tried to patch up the classical analysis by adding a condition to it. So something that was tried in Anglo-American philosophy in 1967 and had been tried in the Sanskrit tradition in like a 14th century is to add a causal condition, right? Like to say, what's wrong with the water seeker who finds that well when he goes down into the valley? Is it wasn't that well that caused him

to have the belief that there was water in the valley. It was a hallucination, right? So some people have said there has to be a causal chain linking the fact to the knower in order to have knowledge. That's what you have to add to justify true belief. That sounds reasonable. Sounds reasonable.

Doesn't work. Doesn't work. Doesn't work. Yeah. Why? Oh, I mean, like, there's six reasons why it doesn't work, some of which are kind of tedious. You can do stuff with like deviant causal chains where,

yeah, there's this causal chain there, but it's not the right kind. Those examples are tedious to point out. Alvin Goldman is the guy who gave analytic philosophy the causal theory of knowing in 1967. Alvin Goldman himself rebelled against it and came up with counter examples in later works. So by 1976, his classic counter example, I don't know how you're going to feel about it.

It may or may not work for you, but I'll tell you. So what Goldman wanted to do was he wanted to have an example of something. The causal theory was going to have to say that is that's knowledge but you intuitively are gonna feel like that's not knowledge okay so here's how the example goes these are called fake barn cases Goldman himself actually did not Goldman himself credits his former colleague, Carl Genet, with coming up with the example. Just give credit where credit is due. So the example's like this. You've got this guy who's driving through the countryside. He's just got his little

kid in the car. And as they're driving along, he's pointing out different, oh, look, there's a cow, look, there's a barn, there's a haystack, you know. So he's like making these judgments out loud as he's driving through. Yeah, and one of the things he knows is, yeah, there's, or one of the things he says is there's a barn, right? And he points at what is in fact a barn, okay? And it looks like, so far it looks like this guy knows that there's a barn there, right? You can stipulate also his perceptual systems are working fine, he's not hallucinating, whatever you want. We're even going to stipulate because we're

fighting against causal theory of knowledge. This guy's belief that there's a barn is caused by the actual presence of a barn on that spot. But now we have to have a plot twist to make it into a get-it-your-case. The plot twist is that this guy is driving through fake barn county. As he's been driving along, there's actually been a whole bunch of things that look like barns, but they're not. It's fake barn county where the locals,

for reasons best known to themselves, have put up a bunch of facades. From the road that looks like a barn, but it's just hollow and there's nothing behind it. And we can even make our driver have already seen, he's already seen a bunch of these things, oh look, there's a barn, there's another one, there's a barn, there's another one. And 99 times he's been actually wrong, he's pointing at something that's not a barn.

So when he gets to this one, it looks just the same, it's all those fakes. He says, there's a barn. Does he know it's a barn? If Goldman tells a story better than I did, if you could bring Goldman back and have him tell you the story, you would have kind of have the feeling that no, he doesn't. And that's a type of story that we've also investigated

in the lab with our student participants. So you can do stories where, you stories where people are in a jewelry store and there's a lot of fake diamonds and one real one. And they're, oh, look at that beautiful diamond. But all the stones look the same to them. Do they actually know that that one's a diamond? Maybe not. Even if it actually is a diamond and their judgment that it's a diamond

is being caused by the very object itself, which is a diamond. So the causal connection is there. Feels like something else is not there. How Goldman dealt with this was he updated his theory to what he called reliableism. So reliableism, which is another very dominant approach to theory of knowledge, says that you know when your true belief is the product of a reliable mechanism. So a reliable mechanism is one which has a high percentage of true deliverances.

That's the crudest version of reliableism that I can give you. It doesn't have to be perfect in Golden's theory, but it has to have a high percentage of true deliverances. In order to count as knowing in this particular case, it has to be coming from that kind of mechanism, and in this particular instance, the judgment reforming is true. Now, I imagine that you have some critiques about reliableism. Hell yeah. It's so wrong. So we can get to that. Why can't another answer as to the

difficulty or maybe in principle impossibility of defining knowledge, why can't it just be from Wittgenstein where he says words are just something that we use. There can be a fuzzy notion to them and if we're trying to pin something down with these necessary and sufficient conditions, we're going to fail because we're not going to find a word that is satisfied in every context that we want it to be satisfied. Can we just say that?

You can say that. You just did say that. I don't find that very satisfactory. Do you? Are you convinced? Well, words are fuzzy because we use them. That actually doesn't even really sound very compelling, does it? I mean, we all think words can be used well or badly, and if we use them well, why not use them with precision?

Well, do you think every word can be precisified? Precisified? Yeah, precisified. I just precisified your- Uh-huh. No, but no. So I do think most natural language is vague. Here I'm channeling Timothy Williamson again.

I think recognizing that a term is vague does not mean that we can't do anything with it, deepen our understanding of it. It just means that there are, that the boundaries of its proper application might be unknown to us. But that doesn't mean we shouldn't like, like here's something absolutely don't want to do. I don't want to say, oh, because natural language is vague,

let's just throw in the towel and like stop poking the beast, right? I think of Wittgenstein as a bit of a quietist. He's a bit of a guy who says, philosophers quit it. And I'm like, who are you to tell me not to play? Like, you know, stand back. I'm gonna tell you a few things about knowledge that you did not already know.

Like, I'm gonna tell you what's wrong with reliableism and how to put yourself on the right path. So here's one thing, I'm just like dying to get back to reliableism. So the reliableist says, you've gotta have a belief-forming mechanism which tends to reliableism. So the reliableist says, you've got to have a belief forming mechanism which tends to go right. And like, well, what's your cutoff point, right?

Do you want 99 percent rightness, 98.3? Sure. Give me something. And it's hard for them to set the level. And also, I think, there's some ways of thinking about the problems of knowledge that really push things down and other ways that really push it up. It seems kind of there's something problematic, something incoherent about it. So usually in this context, epistemologists will start talking about the lottery problem.

So you know the lottery paradox? Yes, yes, yes. So this is the idea that let's say you have a ticket in a fair one in a million lottery. The draw has just been held, so there's a fact of the matter about which ticket won, but the winner has not been announced. Do you know that your ticket has lost? Well, do you? The problem that I have with the lottery paradox is just saying it sounds like you're taking

something that's 0.001% chance and then calling it a 0% chance. Yeah, and that's fallacious. So, but you're evading the question. Do you know that your ticket has lost? No. Thank you. Yes, exactly. Even though there is a.999999 chance that it has lost, you don't know that it does, that it has lost. And so, like, in this context, if you defined as your belief-forming mechanism, you know, it's just pessimism or something like that, a pessimism in these circumstances is a reliable

mechanism. It has a very high chance of delivering a true result. It will deliver true results for everybody but the actual winner. There's something really counterintuitive about this. There's a bunch of problems around the lottery paradox, but with reliableism, I think like. Well, there's a problem with assigning a zero probability to something anyhow.

Oh yeah. There's the dogmatism paradox. Yeah. Well, some things we can assign zero probability to, like that P and not P. Can you? Aren't there other logical systems though, where you can have P and not P? We're already assuming contradictions blow up. Yes. Well, am I assuming that or do I know that that's the case?

I'm not particularly tolerant in logic. I'm like classical to value it all the way. Okay. I don't have to take on board your nonsense logical systems or pseudo systems. I mean, I know there are logical systems which have beautiful formal properties but I'm not going to say it's a level playing field between all systems of logic. That way madness lies.

Well, the problem is then the dogmatism paradox, which is that what if someone was to present you with some evidence that hey, classical logic isn't the way to go, it doesn't map to our reality, then you would automatically think that evidence is false. You're not open to changing your mind on it. I should certainly hope not. Yeah. I'm not terribly worried about the hypothetical.

What if someone were to present evidence that? There are some hypotheticals that I am really worried about, right? So if you give me a contingent fact, I'll worry about the hypothetical, contrary, but Let me persuade you to worry about that hypothetical. What if it's the case that there's some entity, some deity or some demon that says, look, whether classical logic is true or it's not true, it maps to our world, it doesn't map to our world, there's some other logical system, or something that's

non-logical or what have you, depends on this baby being tortured. So if you, sorry not depends on, but the consequence of you choosing correctly that classical logic is the way to map to our reality or some other button here is the way to map to our reality, if you choose wrong here is the way to map to our reality. If you choose wrong, this baby will be tortured. Then you're going to pause, no? One of the many reasons that I would pause is this is

a very unusual circumstance in which to find myself. But I don't think that that really has philosophically interesting consequences for the question of the adoption of logic. I also feel like I'm not the person to have this fight with you not being a logician. But yes, I appreciate that the range of healthy controversy in philosophy is fantastically wide. And I think we can actually learn a lot by talking about outrageous scenarios. That's certainly true. We can learn a lot by talking about outrageous scenarios.

But I wouldn't be too quick to take the hesitation you have about the pseudo possibility that you're describing as a sign that we should be assigning non-zero credence to alternative logics. I think that's a confusion, right? You know the theory that nothing can be probable unless some things are certain. And in any way of sort of marshaling a problem, thinking about it, there are some things that we are taking for granted, and perhaps rightly so, and perhaps even knowledgeably so. We shouldn't necessarily beat ourselves up too much about that.

Well, that's extremely interesting. Tell me more. What is it that we can know for certain? All kinds of things. One of the things I want to get in with right away is what we mean by certain in this situation, right? Okay. So, one way of thinking of certainty is just to think of it as the absence of a current

doubt. So, right now, I presume you're certain that 2 plus 3 equals 5. Now, we could probably induce doubt in you on that question. That would be a ridiculous thing to do, but there are ways of making that happen, including, you know, we could administer drugs that would make you feel unsure of that. But it is something that we can know for certain when we are not placed in adverse circumstances. We can be in a state of mind where we're linked to the truth and we have no occurence doubt. There's a stronger way of thinking of certainty where certainty is not just the exclusion

of actual occurence doubt, just the exclusion of actual occurrence doubt, but the exclusion of all possible doubt. What's something that you couldn't possibly doubt? And there I think certainty is going to be very, very much harder to find, because you can take lots of things that you actually do know, and you can induce spurious doubts in them. And I think in a lot of cases, you're inducing those doubts by basically fostering cognitive illusions of one kind or another,

creating sort of nonsense hypotheticals and trying to put you in a place where you start feeling like you have to argue for the thing that you once intuitively judged yourself to know. You see an old friend and you just know immediately, that's Rachel, right? You see her face and you just know her right away. Then I start saying, look, what if she had a doppelganger, identical twin? I conjure up all sorts of imaginary scenarios and you start doubting yourself.

Why do you start doubting yourself? Well, maybe partly because your own self-evaluation as a knower is coming under pressure from my line of questioning because I'm not taking you to be knowledgeable. And I think a lot of our self-evaluation, do I have knowledge on this point or not, is generated through interactions with others. So if you make a judgment, as you do all the time, communicate it to somebody else,

and they take it on board and they're like, yeah, right, okay. Then you easily and confidently judge yourself to no. But if you make a judgment, perhaps a very similar judgment, communicate it to somebody else. I ask you, hey, who's that? And you're like, oh, that's Rachel.

And instead of saying, oh, okay, hi, Rachel. I'm like, well, can you prove it? I'm sure you have a twin. I'm giving you a lot of feedback when I react to your claims. And it's feedback that you actually use intuitively in mapping yourself as knowing or not knowing. One of the big things that we do when

we find that our claims are resisted is we present arguments for them. And that's not something that we do just, you know, Western educated whatever. That's a human universal. So I think Hugo Mercier's work on this is really interesting. He's done some, you can look at his paper on the universality of argumentative reasoning,

which is a sort of cross-cultural, cross-linguistic analysis of our human propensity to argue for things, present reasons for things whenever we're put under pressure. And there's a very, very interesting function that argumentation performs, right? Because ordinarily, you will accept a lot of other people's claims on trust when they're speaking from inside their epistemic territory, inside the zone of things that you would count them as knowing. Like you come up to somebody at a party and you're like, hi, what's your name?

Right? And they'll tell you something. You could say, well, are you sure? But you don't. You just update your model of the world immediately to take that on board. Because what your name is, that's something located deep within your epistemic territory. You've actually got a lot of other stuff in your epistemic territory in the zone where people will intuitively recognize you as knowledgeable.

You're in a restaurant, you want to know where the restrooms are, you could ask the waiter, and you're not going to double-check, oh, do you really know? You're just going to take his word for it. And that's an extremely efficient thing for us to do. We're all the time evaluating what people do and don't know. We're kind of testing our models of what they do and don't know in conversation, right? You tell somebody something and you monitor if they're surprised or they react like they already knew it. But to get back to the main thread with Hugo Mercier, we are not restricted to telling each other just things that fall inside

our recognized epistemic territory. We can go outside of that and say things that push the envelope a little bit. But when we do that, what we say is not necessarily just going to be taken on trust because the person you're speaking to doesn't automatically see you as knowing it. So you have to do a little demonstration that you know it. What's an amazing way to demonstrate that you know something is to show that it follows from premises that the other party is either going to take on trust from you or that they themselves already know.

One of the coolest ways I can get you to update your model of the world with something that I know, is to show that it already follows from some things that you know. The path of presenting a sound deductive argument, that's a path that we follow when we're trying to exhibit or display our knowledge. If I can give you a sound deductive argument for something, which starts from premises you know to be the case, I've just given you the gift of what I know, right?

Now you know it too. And so that's a really gold standard way to share knowledge. And if somebody puts up resistance to what you're saying, even just to something as simple as, hey, there's water in this bottle. If you're like, prove it. I'm gonna try to give you an argument. I'll give you some reasons, I'll pour some out, I'll taste it, you know, I'll try to do that.

But if you just keep up a stance of, yeah, I don't know, could be a clear colorless liquid that's not water. If you just keep that up, if you don't take any premises on board, I don't have anything I can work with you to get you to accept the conclusion of that argument. One of the things that starts happening to me is,

I start feeling like, damn, maybe I don't know this thing. Mm-hmm. Okay. Because I'm trying to, in the course of argumentation, trying to demonstrate my knowledge to you. I had the knowledge already even before I started trying to demonstrate it, but now that I'm trying to demonstrate it to you, if you're resisting the demonstration, I'm going to feel like maybe I don't know it. And that's just a kind of heuristic that makes a lot of sense because a lot of the time if you say something to somebody else and they are resistant, that

actually is a meaningful update to whether you actually know the thing, right? Like if I say, hey, did you know Sabrina and Barry just broke up? And you're like, no, I don't think so. And you have some resistance. I'm going to be in a place like, oh, actually, wait, maybe I don't know that Sabrina and Barry just broke up. So this is like a super interesting thing about human beings that really sets us apart from other species.

That we actually collaborate on our models of the world in a way that no other animal does. And your sense of the presence and absence of knowledge is really shaped by the role it plays in that collaboration. So my vision of how knowledge works is it's really not knowledge works is it's really not best understood by trying to reduce knowledge into factors like justified true belief. I think it's best understood by looking at the functionality of the concept of knowledge, for example, in conversational exchanges. So if you characterize knowledge as a state of mind of the type you can have only to the truth, that characterization is actually a much better guide to how knowledge functions

and how it actually works to kind of like structure our collaborative exercise of making sense of reality. Do you know if the sensitivity you have to doubting your own knowledge based on the feedback from others of their skepticism or their doubt or what have you, does that scale with agreeableness versus disagreeableness? So if you're disagreeable, well, you're less sensitive to it? That's not something I've investigated. I'd be interested to see that. Okay. So Socrates apparently said, I know now that I know nothing, and let's just remove that that's self-defeating, because he knows that he knows nothing. Did he actually say that? What does that mean? And is it indeed wise? Like many people like to quote it and

say, oh, look how humble I am, I know nothing as well. So is that wise to know nothing? Of course not, of course not. And I do think Socrates is having a bit of fun right there. Socrates was not a big one for proclaiming his own knowledge. And he was very, very good at eliciting interesting observations from others with his method of questioning. And I think very, very good at expanding their Critical self-understanding and their critical knowledge and I say that not just for the people who had the pleasure of speaking to him When he was alive, but also for those of us who read the records of his dialogues

There are other places where he does make claims to know so actually in the Meno He says that he describes himself as knowing very few things, but one thing that he knows for sure is that there is a difference between true opinion and knowledge. So he's actually on my team, not just as recognizing that distinction, but describing himself overtly as knowing that that distinction is there. I think the game of, I don't know, tell me, that's just such a good game to play in philosophy. And I think it's very, very good to look at the places where Socrates chooses to play that card. So he's particularly interested in questions of absolutely foundational import, right? He's very interested in the question of what justice is,

what knowledge is, these very, very foundational issues. And I think he's very good at showing that these little things that people say to themselves, like justice is, tit for tat, or something like that. The little formulas that we trot out don't bear up very well under critical scrutiny. And my feeling with Socrates is that he really did think there was something very foundational about the character of those terms, and that we should at some point appreciate that other things are going to be understood through these terms, that these fundamental abstracta have a kind of structuring role in thought,

which we shouldn't try to deny. But I want to give you just a line on a place where Socrates does and doesn't play the skeptical card. So if you think about the discussion in the Republic where they're asking, okay, what is justice? Someone says, okay, you know what justice is? It's like telling the truth and paying your debts, right? If you borrow something, you gotta pay it back.

And Socrates has the counter example, like just a second. What if your friend borrows a weapon from you and then he comes back later and says, I need it back, but he's out of his mind, like he's out of his mind with rage or something like that. Do you like, okay, so justice is paying your debts. Here's your weapon, right? Obviously not, right? Now one of the things that Socrates does is he leans on your intuitive judgment that obviously

that's the wrong thing to do. If he was a really, really tiresome skeptic, he wouldn't just say, oh yeah, obviously not, you don't give the guy the weapon. He'd be like, well, or do you? Maybe you should give the guy the weapon, right? Like he could take you down that path and he could just keep saying, you know that definition of justice as telling the truth and paying your debts? Maybe that is right, you know?

He could go there. He doesn't. He shoots it down. He says that's a no good analysis of what justice is. If you're resting easy with that, you're in a bad place. And he's not skeptical about that, right? The classical skeptic, sextus empiricus, guy like that, would look at that definition and is like, well, is it right or is it wrong?

I'm just going to keep this question going. I see some things to say in favor, some things to say it against. Socrates isn't doing that. I think a lot of what Socrates is doing is he's wiping positions off the map. He's saying, it's not like that, it's not like this, it's not like that other thing. If you wipe enough off the map, you're left with the truth.

I think that's the place that Socrates is taking us. He's giving us knowledge expansion. It's almost all negative. It's almost all like you thought you had this slick little formula for defining knowledge. Nope, doesn't work. But that is knowledge gain nonetheless, right? And very important kind of knowledge gain.

Well, it would only get you to the truth if the map was small enough or your wiping outs were large enough. Yeah. But I think actually one of the things he's doing is he gets, I forget the count, is it nine different rando definitions of knowledge that he wipes out in the Republic. Actually, the Republic is a bad example because he does have a lot of positive things to say in that and it was Plato's Dialogue More Than So Socrates's, but you know one of the things that you feel is sometimes you're getting the lesson of we could go on like this,

we could go on like this all day. Now that we've started the pattern, you can continue it. And I think this is something we've really learned with the Analysis of Knowledge program, that we had starting in the year 1963 in Anglo-American philosophy just a whole succession of purported analyses of knowledge, all of which got shot down with counter examples.

Something looked good for a month or two, everyone was like, oh wow, we got it. We nailed it. No, no, got another intuitive counter example to these increasingly elaborate analyses. Then at a certain point, so Timothy Williamson really made his name in epistemology for saying, we can perform a certain kind of induction here and see this is a degenerating research program, right?

We're getting more and more elaborate analyses, sort of feels like all these epicycles people came up with when they were trying to save the geocentric theory of the solar system. And it's not looking good. Like maybe somebody is someday going to come up with a super complicated analysis that actually threads the needle and gets through all these counter examples that have foiled previous analyses, but doesn't look like that's happening anytime soon. Maybe we need another approach.

And so the approach he's proposing is this sort of knowledge first approach where you think of knowledge as the foundational concept in epistemology, mental state of the type, you can have only to the truth. And then you're actually going to build your understanding of weaker states like belief out of knowledge rather than going the other way around. So maybe what belief is going to be like is going to be a state of mind that first person, present tense pretty much approximates knowledge. What is it just to believe something? It's to be in a state that is first person, present, tense, pretty much indistinguishable

from the state of knowing. Here's another way of looking at it. This is something championed by Alexander Bird. Bird has the theory that you want to understand mental states, creatures like us. Maybe let's take a kind of evolutionary path to understanding them. You can see the evolutionary advantages of creatures having mental states of the type that can be held only to the truth. It's good for creatures like us to have cognitive mastery of their environment, right, to be on top of the way that the environment actually is. There is

no particular evolutionary benefit to just having beliefs or opinions as such, right? If your beliefs or opinions are a pack of nonsense, that's not actually doing you any good as an animal that has to thrive in an environment, right? Your cognitive systems are good only insofar as they are locked onto the truth. So we need to tell a story about how creatures like us get to have those kinds of states. That's the function that is getting selected for, both in natural selection and in reinforcement learning. So, in reinforcement learning, so first, something I would say about belief is that belief is sort of larger category, it's a spillover

category of everything that our knowledge-forming capacities produce plus all the byproducts, right? When you believe something but don't know it, that is a byproduct of your general capacity to know things, right? If you think about perceptual illusions, for example, right? Typically, you look at what seems to be a body of water, there's a body of water there, and there's a good biological and learning explanation

of why that's the case. Sometimes you're gonna have an impression that there's a body of water there when there isn't really. That is a byproduct of the system, of the perceptual system. It's not why the perceptual system was selected to exist in you. It's a byproduct of the system. It's a byproduct which nevertheless can drive your behavior.

False beliefs, beliefs on the basis of illusion and all of that, misconceptions, those can drive behavior and make you make choices. And interestingly, we learn from mistakes. It's not always an entirely bad thing that your behavior is driven, not just by knowledge but also by this larger spillover category of belief. Because when you believe that something's so and you get it wrong, you're going to get a little slap on the face from the environment, and

that can sculpt your neural representation of what's going on in ways that will pay off down the line. It can move you towards—so I like to think of knowledge as a kind of basin of attraction for us that we tend over time because of the error correction processes of reinforcement learning, we tend over time to gravitate to states of the type that you can have only to the truth. Our cognitive systems are sort of optimized for that. We don't always end up there.

We don't always end up in that basin of attraction, but that is in general where these systems have been selected to take us. Have you heard of Donald Hoffman? Yes, yes. So doesn't he have an argument about even if we've evolved, it doesn't mean that our senses are attuned to reality. In fact, if you look at all of the ways that reality could be and all the ways that our perceptions could be, there's a near, if not measure, zero chance in math of us actually mapping to reality.

Right. So I think those arguments are, I find those arguments somewhat underwhelming. I'm much more attracted to the sort of counter program coming out of Brian Scholl's lab at Yale. So I'm thinking if he had a paper a couple years ago. I think the lead author is Merriam Burke, but it's Brian Scholl is on that paper as well. Arguing against all of that in favor of realism, just by kind of pointing out that- Interesting. Do you know the work? No. Right. Okay. So here's the thing.

Was he arguing against Hoffman or just arguing for- Yes, arguing against Hoffman or just arguing for Hoffman? Yes, arguing against Hoffman. Okay, interesting. Directly, directly. So here's the thing. Our perceptual systems put us in touch with the world in ways that have to be robust across

a number of different projects or activities that we might engage in. You see a stick, and what's your purpose? You might have all kinds of purposes. You might want to pick it up and use it as a weapon. You might want to start a fire, toast some marshmallows. There might be all kinds of things that you need to do with this thing. If you had a fixed utility function, some version of Hoffman's argument micro-through. But given that we're creatures with

a vast multiplicity of purposes, you need perceptual systems that tell it like it is. That's the simplest rebuttal of the Hoffman thing that I'm aware of, and it seems to be very compelling. In 2018, you had a paper on episodic memory and the self. Okay. I think that was just like a VBS commentaries. So yeah, what did I say? Anyway.

Well, I want to know more. I want to know about the self. And is the self an illusion? Is the self real? What is the self? And what does it have to do with knowledge slash memory? Oh, okay. Yeah, good question. There's a lot of extremely interesting literature in this area, and I am a secret closet fan of Thomas Metzinger,

even though I don't agree with him. Metzinger is being no one, the self is an illusion, there is no self. I love it. I don't think it no self. I kind of love it. I don't think it's right. Why do you love it? It's so drastic, right? And I like drastic and extreme positions even when I don't agree with them.

So I find it very interesting. I think we have some intuitive beliefs about the self that can't possibly be right. And there might be sort of naturalistic explanations about why we hold those. I do think there's a lot of very interesting questions about the self versus the automatic psychological representation of self, these things can somewhat come apart. I think a lot of Metzinger's work is really motivated by the possibility of that divergence. I have not personally written very much on the nature of the self.

I would recommend Brie Gertler as somebody really interesting to talk to on that point, or maybe Peter Carruthers. I do think that social cognition is something very distinctive, very interesting, absolute fundamental point in our cognitive activity, and social cognition posits the existence of the self. Can I tell you something? Can I tell you something? So I think of the self as a defined point within social cognition. I think we have very early representations of reality as involving a fundamental division between agent and environment, between self and other. I think that distinction is something that goes back to the very, very

beginnings of animal life. As soon as you have sentient creatures capable of self-movement, they need to be able to distinguish between sensory states that are getting generated by a change in the environment and sensory states that are changing because of something that the creature itself did. And at that moment, you start to have the emergence of a conception of self. I think with us, we have a very rich conception of self because social cognition is so pivotal to our understanding of reality. I think as far as metaphysics is concerned, the self is the agent. And the self-conception is what the agent uses in

order to navigate reality. And these things can come apart in various sort of interesting ways. The self-conception doesn't tell the agent everything that is true about the agent, right? It's just, it might be kind of on the right track, you know, the self is like a little, the self-conception is a little kind of tiny dashboard that's showing the agent a very small snippet of what they are and where they are. And the richness of the agent as an animal with very, very complex cognition under the hood

couldn't possibly be represented in that little tiny dashboard for all kinds of reasons. So there's much more to the self than the self-conception. I will just say about social cognition versus our physical sense of reality. When you look around the world, there's a lot of rules that apply to inanimate matter, right? You make all kinds of judgments about how things are in time and space and how things will live towards you and fall. That are just your sort of internalizations of some basic notions of physics.

Then you also have a whole rule package that you bring out with agents. And they seem not to be, well, we are of course all governed by physical law. But our behavior is intelligible not only in the light of what's going on with physics, but also in the light of what's going on with our intelligence. That you see agents as having goals or purposes, they're not just waving their limbs, they're reaching for the salt,

they're trying to do something. You understand them in terms of their relationships to the environment. You don't understand them as like cut off from the environment. They are parts of the environment. And I will say that about the self as well, right? That we have this very fundamental way of

thinking of, you know, it's me against reality over here. But of course, thinking about it reflectively as a philosopher, I'm part of reality. I as an agent am embedded in this reality and I am a part of it. And so to the extent that my self-conception is like it's me versus the world, that is a little bit of a mistake in my self-conception because I'm part of the world. You're part of the world, right? And we've got to kind of keep that thought, keep that thought that you're part of the world going when we try to do a sort of philosophical unpacking of what the self is. The self isn't something that confronts reality from outside.

The self is something that's located within reality. And then, like, in social cognition, like, we make sense of the world so much through the reactions and the behavior of other intelligent agents within reality, right? So you look at someone else's face as a guide to the things that they can see. If you're suddenly, you know, if you're facing somebody and they suddenly like recoil in horror, you're going to react to that. You're going to look over your shoulder. Like, what did they see, right? We can use other agents as mirrors of reality because we understand that they're embedded

in this reality with us and they have their own intelligent grasp on what's going on and we can sort of harvest their intelligence. You see that in non-human animals as well. They'll do gaze following. You have birds that will look up and then the other birds in the-human animals as well. They'll do gaze following. You have birds that will, you know, look up and then the other birds in the flock will follow that, follow that gaze. And that's one way in which the

the wit and intelligence and observation of one animal can spread to the, can spread to the flock. That's very helpful that they can pull their epistemic states that way. They do it kind of, you know, by chance and as chance arises, we do it deliberately and strategically. We set out and we question each other and we actually really use each other. We do something no other animal does in that we show things to other agents and ask for their comment on it. No other animal does that.

Yeah. Yeah. So we're like actively manipulating the social cognition of other agents in a way that is not seen in other species. For a first year philosophy course, what do you want your students to be walking away with with more certainty and with less certainty? Hi, everyone. Hope you're enjoying today's episode. If you're hungry for deeper dives certainty. a thriving community of like-minded pilgrimers. By joining, you'll directly be supporting my work and helping keep these conversations at the cutting edge.

So click the link on screen here, hit subscribe, and let's keep pushing the boundaries of knowledge together. Thank you and enjoy the show. Just so you know, if you're listening, it's C-U-R-T-J-A-I-M-U-N-G-A-L dot org, Kurtjaimungle dot org. I'm not a certainty salesperson in that I think certainty as a subjective state, that feeling of conviction is not what I'm after. I'm after knowledge, right? If I wanted to spread certainty, you know, I could be some kind of revival preacher,

and I could have some crazy theory, but if I had a really charismatic way of selling it, I could just try to make everybody feel very certain about it. That's never what I want to do. I want them to know more than they did when they started the class. And so that's going to be, to some extent, a matter of them gaining propositional knowledge about various important figures in the history of philosophy. I want them to know something about

the criterion argument in the Shwangsha. I want them to know how Al-Ghazali thinks you get out of the skeptical problem. I want them to know what John Ismael says about time. So they have that sort of baseline of knowledge of what other philosophers have said. But I want to be very clear on day one even that what's even more exciting is for them to find their own philosophical voice, right? For them to be able to tackle these questions for themselves,

we'll see that many very basic questions in philosophy have extremely attractive answers on both sides of the question. And it will be an exercise of their intellectual freedom to side with either of these answers or articulate a third way of their own. I want to see their creativity. I want to see what is new in philosophy. And I feel like I get that every single year when I teach intro. Students are thinking new thoughts that no one has ever thought before. I feel this especially with the rise of AI. That we now have this other form of intelligence, a little alien in many ways, right alongside us.

You can have philosophical conversations with it, and many of them have. And they are seeing things there that maybe weren't within the realm of what earlier philosophically-minded people have had the opportunity to see or to think. So I think there's always new challenges in philosophy. There's always new ideas emerging. I want my students to gain the power to tackle some of those challenges.

It seems like you'd also want your students to understand, to not just know more facts, but to understand. Understanding is a species of knowledge. To understand is to know why something is the case. So I don't think you can understand. So just a moment. I want to see if I understand the definition of understanding. So if someone tells you some fact,

you can just believe it or you can see it, as you mentioned earlier, you can say, okay, what is it that you believe or you're willing to grant? Then let me show you how you can get from there to here. Knowing the path or at least being able to articulate the path and then you can follow the path, that's the understanding? That would be one form of understanding, absolutely. Yeah, yeah.

Knowing why something's the case is understanding. So you could have a mathematical theorem that someone tells you, and it's like Ternes Tau theorem that someone tells you, and it's like Terence Tao, and he tells you something, and you're like, okay, sure, I'll update. Yeah, that's true theorem right there. Terence Tao says so. Or you can have him really go through the proof with you,

and bring it down to your level wherever that might be, until you really understand for yourself the reasons why this particular theorem is true. I think you can equivalently express that in terms of knowing, though. You could say, Terence Tao told me that this has been proven, so we now know that this is known. I know that this is true, and I know because Terence Tao told me. Or you could say, I know the following theorem and I know why it follows from the axioms of said theory or what have you.

So it's just all a form of knowing. Interestingly, if you look cross-linguistically, and here I'm relying on a work by, I think it's Anita Sjöberg, who did a doctoral dissertation on how verbs of knowing work across a huge, huge number of languages. There's lots of languages in which the word for understanding is some kind of composite expression or adaptation of the verb to know. But none where no is crafted out of understand. Okay. No seems to be the more fundamental term.

Tell me Al-Ghazali's response to skepticism. Oh, so it's super fabulous. First, he has a series of, his path into skepticism is fantastic. So he describes himself as having been freed from the bonds of conformity at a very early age. And where conformity is something like just mindless accepting what other people are saying, right?

And one of the things that Friedman was he noticed that he noticed that children who grow up in Christian families turn out Christian and children who are growing up in the Muslim families turn out, end up believing in other thing and children, the Jewish children in those households, they end up believing a different set of things. And, you know, of course, that sets off a whole crisis of faith. I mean, he worries that his own beliefs might just be the products of conformity and can't be trusted.

And so he decides he's got to try and figure out like what can be known for certain and ends up dividing the things that seem to him to be most certain into two categories, the, you know, deliverances of sense perception basically and the deliverances of reason. And he has really great skeptical arguments against trusting either of these. And then he has a sort of meta argument where he argues, like, it is going to be impossible to get out of the skepticism because it feels like to get out, I need a proof. But, you know, proofs start from premises.

Right. I don't seem to have any. And by his own telling, he's released from skepticism by an act of divine intervention. God shines a light into his soul and restores his trust in the rational truths. And that's the start of his path out. And he says those who think that the mercy of God could extend no further than proofs are really underestimating God. So for him, there's divine salvation from skepticism, and that's actually the only way that you can get out of it. It's interesting to read that and then read Descartes right after, right?

Who also has an essentially God-based path out of skepticism, but one of a quite different character. How do modern philosophers view that sort of reasoning? Depends on the modern philosophers. There's a lot of different ones out there. How do you view it? How do I view it? Very interesting, very intriguing. I don't accept that that's the best path out of skepticism,

but I find it an extremely interesting argument. I read it as an argument about the limits of argumentation, and there's something to it on those grounds. I don't think that divine intervention is the only or best path out of skepticism. I think I would take issue with that. He said that it allowed him to believe the rational truths. What do you mean? Like the axioms?

Yeah. What are the axioms or just any axioms? You'll have to have Al Ghazali on your podcast to explain that, or you could have a contemporary scholar who knows literature really well. I would recommend Muhammad Ali Khalidi, who's at CUNY in New York. He's a philosopher of science and also a scholar of medieval Islamic philosophy.

So he would number one with a gold star recommendation. Great guy to talk to. How do you get out of the problem of skepticism then? I get out of it by understanding exactly what it is and where it comes from. So I get out of it the way that you get out of the problem of, wow, it looks to me like here's the Miller-Lyer illusion and it looks to me like these two things are different lengths but when I reach out and touch them, they're the same.

Sure, and I can tell you a story about how your three-dimensional visual perception is generating this impression despite the fact that it's not so. So I get out of it the same way. I'm like, I'm here to explain why it starts to feel to you like you don't even know that there's water in here. There's a natural story about that. And it's a story about your intuitive self-evaluation under certain kind of social pressure. In fact, the big way that we as human beings are

marking the presence and absence of knowledge is in conversational exchanges. Conversational maneuvers that essentially assume that you don't know something, even if it is something that you actually do know. Conversational maneuvers that assume that you don't know something will start giving you the feeling that you actually do know. Conversational maneuvers that assume that you don't know something will start giving you the feeling that you don't know it.

Can you give me an example? Well, just any kind of series of skeptical maneuvers that we could try. Like the two plus three equals five? Sure. Then we could roll in Descartes concerns about the origin of irrational nature. Sure, it feels really compelling to you to accept that two plus three equals five, but couldn't an all-powerful God have created you such that that would seem very compelling even if it were not the case. And for those unwilling to count as the existence of God, we could

say, well, so my rational nature came about by some series of lesser chances, maybe natural selection and so on. And couldn't I be such a system that I feel a sense of compulsion with respect to that fact even when it's not the case? So you can have these kind of argumentative schemas which deny, I refuse to attribute to you knowledge of something that you actually do know. You're going to start feeling like you don't know it. And one reason is you feel kicked into argument mode.

I gotta give reasons for why I believe that two plus two equals four. I have my liberty as a skeptic to just resist all your arguments. I'm like, yeah, really? Well, where did you learn that? And then you're just stuck playing a losing game. So there's generally a dynamic,

which is an epistemic dynamic in conversation. Here I'm drawing on the work of sociologist John Heritage. He says, for any turn at talk with clausal content that is after we get past, hey, hi, hello, greetings, stuff like that. As soon as we get into uttering interrogatives and declaratives, so declaratives, the kinds of things that could be true or false, interrogatives, the kinds of things that could have answers, as soon as we get into that territory, every single turn at talk is motivated by positing

an epistemic imbalance between speaker and addressee. If I'm asking you a question, I am positing you as having knowledge and me not having it. When you answer it, we can resolve it. If I accept your answer, I'm like, oh, okay. I'm signaling with this O, which is a marker of change of epistemic state, now I know. Yeah, I like that.

That's called back channel. And so that's a marker that we've come into equilibrium. That's if I'm asking a question you're answering. If I'm telling you something, the typical thing to do when someone tells you something is to give them a little back channel, give them a nod, give them an uh-huh or an oh, right? So if you look at, if you look at corpora of conversation, not so much journalistic interviewing because people will hold back a bit on backchannel, but just conversations like between friends. You know, if you tell

your friends something that happened to you, they'll be like, oh, wow, really? Uh-huh, yeah. They're giving you a lot of feedback that they're accepting what you've said. And that's actually marking that they have taken on board that deliverance. In other words, that they have seen you as knowledgeable on that point, and they have updated their model of the world to reflect the thing that you just said. You're kind of pooling knowledge when you do that. So you're always like marking these epistemic disparities and resolutions in the course of conversation as you're going along.

So both parties to a conversation are really aware of where things stand. When I say they're really aware, I mean, that's happening at the level of individual knowledge in navigating the conversation. The attention of the parties of the conversation is actually always on this sort of topical content that's being discussed. But at the level of individual knowledge, we're also always registering. Is he asking me that or telling me that? And we're always registering.

When you say something, are they accepting it or are they resisting and rejecting? Because if you're resisting and rejecting, I have to double down, I have to add something, I have to argue. That's my instinct, right? And that's an instinct I've gained through reinforcement learning. If you're accepting it, you say, oh, okay, I don't go on and belabor the point. Actually, I do sometimes because I'm that kind of person, but I shouldn't go on and belabor the point.

I should move on to the next thing. So we're really being guided by those signals all the time. And so the skeptic is the guy who's just messing with the signals. He's like, he's refusing to take on board even the simplest claim you might make. What's one thing you know? You say, oh, that there's a bottle there.

No, you don't. And so he's really just pushing your buttons. And those are buttons that are installed in our ordinary conversational processes. And they're installed actually because they serve this really fantastic function of optimizing the accumulation of common knowledge, right? They're feeding into processes that help two people discussing a topic together come to be in a better epistemic position than those two people just thinking about that problem

in isolation. So I'm still not understanding how this is solving or resolving the problem of skepticism so am I to just think of the external skeptic who is doubting everything that I'm saying as You are just pushing my button So I'm gonna classify you as that or if that's some internal monologue that's making someone destabilized or losing their sense of self and going Into psychosis you just say okay., that little voice there is pushing my buttons. Like, it's just the labeling of that and then moving on. Is that the resolution? That's part of the resolution. So the other part is the positive account of knowledge that I'm delivering is it's mental state of the type that locks you onto the truth. And so you

need some kind of explanation of how biological creatures like us can ever be in such a state in the world as we know it. And for that, I think you need a sort of positive understanding of our ordinary learning processes that integrates well with that conception of what knowledge is. We are not creatures who are just here racking up opinions willy-nilly. We are creatures who are here developing neural structures that get us in touch with reality

and neural structures that exist to support action. And we get lots of feedback from the world when we're wrong, structures that exist to support action. And we get lots of feedback from the world when we're wrong, and that feedback goes into reshaping those neural structures in ways that increasingly will lock us onto the truth. Earlier we talked about that you can believe something without knowing that you believe it, or at least that's a possibility. Yeah. Okay.

Can you infer beliefs from actions? Sure. Should you prioritize the action to infer the belief rather than believe what the person says, even if the person is being honest? So I think in some circumstances, like I don't have a general sort of operationalization of that. In some circumstances, quote unquote, actions speak louder than words. Certainly there are cases in which people can be deceptive. You can be deceptive in both what you say and what you do, right? You can lie about your sexual orientation and date or marry somebody who you know at

some level is not right for you. So, neither words nor deeds are completely revelatory of mental states. Yeah. Some people say, you say you're a Christian, but you actually are an atheist, or you say you're an atheist, but you actually are a Christian, or you believe in God. How can someone actually make that argument? Like, what does that mean? That, look, you're an atheist, but you truly do believe in God, or you say you believe in God, but you act like an atheist, therefore you are an atheist, something like that. How does that work?

I mean, I think in general, you can try to demonstrate to somebody that there is a conflict between what they're saying and what they're doing, and you can hope that they have some explanation of it, or maybe you're pushing them to try to defend their view or to be honest with themselves, people can be lying to themselves about what they're doing. I think we don't always get it right when we infer other people's mental states.

You can see somebody as knowing who fails to know. You can see someone as believing who fails to believe. But I think knowledge and belief are natural phenomena. They're like water, they're like gold. Something can be gold even if no one recognizes that it's gold. It's in some underground mine, no one's discovered it. Something can be knowledge and no one's recognized it, or be belief and no one's recognized it.

You yourself might have some beliefs that you're not aware of. I think that's totally possible. Yeah. Let's say you're speaking to a mathematician and the mathematician says, I don't even want to talk about consciousness. That's such an ill-defined notion. I only deal with what's defined. And then the skeptic is speaking to the mathematician and saying something like, well, how do you know that what you're doing is defined? And they say, well, look, I write it out and I understand the proof. And they say, how do you know you understand it?

Did you put it through a proof checker on a computer? They say, no, okay, but let's suppose I could do that. Then you could say, well, how do you know that the pixels on the screen are correct? And so then it gets down to, well, I perceive this. And then in the act of perceiving has something to do with consciousness. So that ill-definedness that they dismissed, they use as the basis of their math, which they call defined. Is there something wrong in that argumentation?

Well, it seems like they're violating their own strictures. If they started out saying that they want to talk about things that are well-defined, and then they start talking about something else. If you're the mathematician playing this game, as soon as the other person starts pushing you, you have to just switch the topic of conversation back to the Collatz Conjecture or whatever.

You know, don't go down that path if you're the mathematician. I think it's, you know, I wouldn't want to tell everybody that they have to talk about consciousness. If some people would rather just talk about math, that's okay. That's okay. Let a thousand flowers bloom. About AI and understanding. Yeah. Is there something that some machine information processing can do that would convince you that it is indeed understanding? Like, what is the criteria?

That's a good question. That's a really, really good question. I'm thinking about that a lot. And in many ways, you want to say the newer models obviously exhibit understanding. They pass our tests for it. They crush our tests at reading comprehension. And if you think of what's going on in them, they have representations of various features that are interacting to produce judgments and in a really robust way, in a really meaningful way. So I think it's not at all crazy to say that these systems can exhibit understanding. That said, there's something very strange

about applying our ordinary human-trained social cognitive systems to interpreting the deliverances of these AIs. So we have learned to use and apply terms like no and understand to individual discrete biological entities. We are not the only creatures that do this. Some other animals, corvids, other primates, can, for example, tell the difference between knowledge and ignorance.

They can sort knowledgeable from ignorant competitors and compete intelligently in light of that distinction. So we're not the only animals that can sort knowing from unknowing. But our human systems for sorting knowing and unknowing are really, really trained on our interactions with creatures like us. And we're coasting on kinship a lot with our fellow humans. When we go to apply these human packages of instincts to chat bots, for example, there are some real questions about how our human instincts translate to that domain. And here I'm thinking especially of the work of Maurice Shanahan. So I'm very interested in sort of Shanahan's view

that with contemporary AI, with a chatbot or something, you have the underlying engine, right? Train next token prediction, maybe a bit of reinforcement learning with human feedback, whatever, GPT-4, GPT-5. Sure. You don't exactly interact directly with that. You interact in a way that is

mediated by a conversational interface. Often in commercial applications, there's a system prompt in there that the chatbot is reading before its conversation with you, but that doesn't really matter. But when you start into a conversation with it, you ask it a question, you wait for its answer,

you raise an objection. You're playing that conversational game that usually play with biological human. In the case of a biological human, there is one subject that you're talking to. I'm talking to one person. I'm talking to Kurt right now. With the large language model, you're talking to a character at any given time.

So Shanahan calls these simulacra, right? It's sort of a persona that it takes on. It's role-playing in his metaphor. It's instantiating a certain way that someone might be. Which particular role-player you get as you're talking to a large language model is in some way up to you, right? Like if you start talking in a very casual way or you start talking like a Victorian gentleman, it'll to some extent echo back the kind of vocabulary that you are using

for obvious reasons having to do with its training. You can sort of elicit different characters from it. Shanahan actually thinks there's at any given time in the conversation a superposition of possible characters it might be playing. As the conversation goes on, you're disambiguating and ruling out some of these characters. There is no one person there.

As we all know, sometimes if you press it and ask the question in a certain way, you get one kind of answer. Ask the question in another way, you get a different kind of answer. Right? Ask it about abortion, but do it, ask it in a way that really makes you sound like a, like a Texas Christian conservative about the sanctity of the right to land, da da da da, or ask it about abortion in a way that really makes you sound like a California liberal. You're actually going to get different answers, right? My undergraduate students are always so fascinated by the ways in which when they ask the same question in Mandarin or in English,

especially philosophical questions, especially political questions, they'll get different answers. It's like our ordinary intuitive systems conceive of us as interacting with single unified agents. So for you, there is a fact of the matter, given any given proposition, whether or not you know that to be true. It might be very hard for me to figure out which way it goes,

but you know, I toss a coin but hide it from you. You don't know how it turned out. I toss a coin but hide it from you. You don't know how it turned out. I toss a coin right in front of you and show it to you. You do know how that coin toss turned out. There's a fact of the matter, maybe very hard to discover matter for you and for any proposition. Large language models are trained on this, you know, very, very rich, enormous, elaborate body of text. and they have some probability

distribution which puts out some kind of answer to the question that you've given them. But is there a fact of the matter about what their neural state is? It doesn't slot so neatly into the categories that we apply to humans. So you might, I think, sort of bend our vocabulary and loosen it up a bit and say that there's lots of things that these models know. They do actually seem to know a lot about, you know, wine tasting and geography and history and all these different categories. But the way in which they know might be quite different. It might be interestingly fractured.

Or perhaps we shouldn't exactly be using the verb to know here. I think those are all open questions. These verbs, know, believe, came out of our organic interaction with other biological intelligences and we may have to abandon or resculpt them when we're dealing with AI. Well, we also have roles that we play in different contexts. And maybe we're playing multiple roles at any given time.

And you can also induce doubt. And so even with throwing the coin on Wednesday, maybe I agree it's a heads, but on Friday I had doubt induced on- So that's fine. You can gain and lose knowledge over time. The fact that you know something right now does not mean that you will continue to know it.

You can know and then that knowledge is eroded by doubt. But I think at any given time for any given agent, for any given proposition, there's a fact of the matter about whether they have or lack knowledge. That's it. I'm not a contextualist. So what if you were to fine tune a model to play a single role, a single role, whatever that means. It's just the TTC agent here in Toronto. Yeah. Or have it play the role of Kurt.

Well, that would be a fractured model. Okay. Would that then pass the knowledge test for you or the understanding test? Because it seems like the argument was about the superposition of different personalities. Oh, that's just a question of what makes it really hard for us to tell. I'd have to hear more about

the single agent hypothetical, right? Because you say fine tune. Now, before we fine tune, we've given it a terabyte of data from the Internet, including all of Reddit, which is the product of many, many, many minds, right? So if it's got that in its back pocket epistemically, that already might mean it's not going to be

a clean case of knowledge attribution the way it is with ordinary. Are you sure? We comprise many millennia of people inside us and belief and biological evolution, the millions of years of the implicit knowledge. Yeah. But yeah, that's a good observation. That's a good observation. Now, my worry is that for us, the borders of our minds are pretty well defined. And it's not so clear that large language models have the kind of unity that you do as an agent. I should be clear, I'm not resisting the idea that

knowledge attribution is appropriate for these models. It actually might be. I'm actually somewhat open to the idea that they fully know, fully understand. I'm just worried that our intuitive equipment is not fine-tuned for judging whether these things know or understand. That's what I'm worried about, which is probably a pretty boutique worry at this point in time. So to know, do you have to be conscious? Oh, good question.

I can give you a cheap answer. You can be fast asleep and still know a bunch of stuff. So knowing can be in a current state or it can be a latent state. The person who's in a deep and dreamless sleep still retains a great deal of knowledge, for example, about their person past, even if they're not activating it at this moment. That's probably not the answer you wanted.

You probably wanted like, could you be not conscious at all? And no. Yeah, probably. Yeah? Yeah. I thought earlier you said knowledge has to do with the mental state. Yeah. Knowledge is a state of mind of the type you can have only to choose. Not all of your mental states are conscious states.

You've got lots of stored episodic memories. That's part of your mind even though you're not activating it right now. You're not conscious of it. If I start naming them like what happened to you, New Year's Eve last year or something like that, maybe that's going to bring it up and so on, now you're conscious of it. But you can think of all kinds of other dates and facts,

questions like pose, things like point to, that you're not conscious of right now, but those are mental states of yours. Can you understand without being conscious? Yes, the person who is, again, I'm just going to give you these cheap shot answers. The person who's fast asleep, it could still be true of them that they understand Finnish or Turkish or any difficult language.

Isn't that a statement of if I were to wake this person up and ask them a question in Spanish, they would understand it? Yeah, that would be how we would verify that they do understand these things. But your understanding is that, like in that sense, could be a standing state of the agent, not necessarily in a current state. Now, where you should be hitting me is on a current judgments.

Can you make an occurrence judgment that counts as knowledgeable even without being conscious? I'll just tell you, we sure can't, but maybe other creatures could. Where should I be hitting you? Can you rephrase that? Oh, so I just keep dodging you by saying, hey, when you're fast asleep, you still have all this knowledge inside you.

For example, episodic memories carved into you by memorable events of your life. You're not conscious, you're in a deep, adrenal sleep, but there's still lots of stuff you know. But you could be, that's standing state knowledge. That's not a current knowledge. So we have a current knowledge right now. Like, currently, you know that I'm so bad at snapping,

snapping my fingers, right? That's happening now in this moment, right? Your knowledge is instantiated in your current mental life, right? And seems to actually unfold within consciousness. So the question is, could there be a current knowledge that is not conscious? And then I'm just going to be like, well, it depends what you mean by a current, right? So it could be that if it's getting called upon to, for example, not conscious and then I'm just going to be like, well, it depends what you mean by current. It could be that if it's getting called upon to,

for example, guide your action, maybe there is some knowledge that you have that is not a current. Can I give you an example? It's a weird example. Please. Well, this whole podcast is filled with weird examples. Yeah. Good. I lived for many years in Pittsburgh, and there's this building that, main campus building, where I just noticed whenever I was leaving the building, I got out on the first floor and left the building.

But whenever I was coming into the building, I got out on the ground floor. I went in through a different entrance and then the one that I always left by. And I asked some of my grad school friends, hey, do you do this thing too that like you always go this way when you're leaving the building and always go this way when you're entering? And they were all, yeah, yeah, I do. Why do we do that? And we sat around for a while and we figured it out. It was because there was like a little bit of a hill. And if you leave on the first floor, you got to kind of walk down the hill.

But if you were going to go into the building, it was just easier to go around the side because you kind of go down and get in on floor G instead of floor one or something like that. And we just kind of realized we had been kind of acting intelligently with respect to this aspect of our environment, but we didn't know that that was the reason that we were doing it.

But I do think at some level, you could describe us as having known that this is the easier way to go, and this is the harder way to go. That knowledge was shaping our behavior. It was an intelligent response to the environment, even though it wasn't consciously accessible, it wasn't available for introspection. I think a a lot of our knowledge, now we can go to town, right? Like we can say, when you're speaking in a natural language, you're invoking your knowledge of all

kinds of grammatical rules, all kinds of stuff about pronoun determination and so on. You don't really know what you're doing, thank God, like you're not looking at the underlying work inks. But do you know how the language works? Do you know how active to passive transformations work? Yes, probably, because you're exhibiting intelligent behavior in those domains. This is actually something that I'm very confident GPT-4, GPT-5 have is knowledge of grammar. I'm very, very happy to say,

they know how to conjugate a verb and not just in English either. They know how to do quite a lot, uncontroversially. When I was speaking with Chomsky, he said that we know how to speak, but we don't know our grammar. So was he just using a different definition of no? Yeah, I mean, it depends what you mean by no. We don't have explicit, articulable knowledge of our grammar.

Most of us don't. He probably has more than most. But I do think there is a strong sense in which we know our grammar. What does any of this have to do with free will, knowledge? Oh, yeah, good question. Good question. I do think that if you're going to act freely, you can't just be acting in a way that's not determined, as if by chance. That doesn't seem to be free will worth the name. I do think free will should involve knowing what you're doing. And I think that maybe having a deeper understanding of knowing could be part of trying to pull back the curtain of the really scary problem of what free will might be and how it might be possible for us to have it.

But it's only a very, very small part. Why is it such a scary problem? Oh, I always think because, like, you know, I'd be very alarmed if we did not have free will. And I very much want it to be the case that we're acting freely and this is not all some ridiculous charade, some ridiculous puppetry of forces beyond our control. I think there might be ways of thinking about free will that don't get you into the place that I'm at with that problem. And I would love to know what they are. But my own sense of it is that human freedom seems to be an absolutely undeniable fact of our existence. And it's something that gives our lives a certain

kind of worth and dignity. It's something exciting. It's something very interesting. And I'm very far from being able to grasp exactly how it might be possible for us. And of course, I think part of the problem might be, well, you know, what would it even mean to make free will intelligible, to give it a mechanistic explanation? Surely that would be a ridiculous farce. So there should be some other way of thinking of free will that legitimizes it and puts it within the scope of our human self-understanding. And I'm just very far from knowing exactly

how that would work. And I just think it's an extremely hard problem and something that I turn over in my mind sometimes very late at night, but it's not at the core of my research because for me right now, not a puzzle, it's a mystery. For me, that's a distinction from Chomsky, incidentally, that for me right now, I feel like I can make progress on the puzzle of the relationship between knowledge itself

and knowledge attribution or knowledge detection. That is a puzzle, and it's a puzzle I think we can get our teeth into now like never before, because we have really interesting new streams of data from linguistics, from sociology, from mathematical modeling, from all kinds of different directions that seem to point towards sort of fresh understanding of the nature of knowledge. So I feel like there's a puzzle there and I feel myself making progress on that. Free will right now, just a mystery.

An enticing line, but just a mystery. Are you speaking about libertarian free will or you allow for compatibilist free will in your mind? I'm not sure what kind of free will we've got. Because compatibilists do allow for mechanistic. For sure. So I was speaking to an AI researcher, her name is Anna Petrova,

and she was saying that, well, when she asks physicists, and I also have noticed this myself, interestingly enough, when you ask physicists, do you have free will? The majority of them say yes. Then when you ask AI researchers, do you have free will? The majority of them say no. And then when you ask AI researchers, do you have free will?

The majority of them say no. And then you ask them why, and they say, because physics. Interesting, interesting. Yeah, actually there's this kind of weird thing that often people who are not in philosophy will give you kind of skepticism-friendly answers to, well, do you ever really know anything? I guess not, like not technically. People who are in philosophy and epistemology, skeptics are extremely rare. Like there's one guy, his name is Brian. That's just one name, just Brian.

No, Brian Francis. There are people out there who are skeptics, but most people who are in my business in epistemology, we believe that knowledge is real and we want to understand how it works, how it fits into the cognitive economy, how it is possible, how it is detected. How do you detect states of knowledge in somebody else or in yourself? How can you resolve these paradoxes about it? Like, we're, I'm in search of something that is evidently a feature of reality. Like I think there is an evident difference between the person who knows who won the game last night and the person who didn't watch and doesn't know. I think that's an interesting, exciting, exciting difference.

Other than it being interesting and exciting, is it important for you? Do you see something great being lost in our society or in your community or with you if the concept of knowledge was just fuzzy like Wittgenstein said, or if free will indeed was an illusion? So I think we care a lot about knowledge. We don't want to be just swimming in fake news and run into some ridiculous state of emergency declared because of something that didn't actually happen. You know, I think that's, I think we want to know what's going on. And we want to have larger social infrastructure that supports the propagation of knowledge, not the propagation of propaganda.

So I think that's very, very important. And to the extent that certain kinds of bad philosophical rumination might lead people to the feeling that knowledge, smallish, there's not really any such thing, let's just go pomo on this. I think actually even to be fair, postmodernists don't typically completely ditch the notion of knowledge. So that's a bad slate. So I think if your way of thinking gets you into a place where the very concept of knowledge

is delegitimized, there's a massive loss to humanity. I think it is very important to humans as a species that we're able to know as much as we do about reality and we're able to collaborate in expanding our shared body of knowledge. And yeah, I think there would really be something lost if we came to the feeling that this is an illegitimate category. And as for the question of like, well, couldn't we just say it's fuzzy and let's not look at it, I think that's just a denial of our basic human self-understanding.

Like I think we should wonder about our foundations and try to understand them better, not just, you know, take it. I mean, I think it's fine to take for granted that you actually do know a lot. But it's, I think, also very healthy to be curious about the nature of knowledge itself. And I think also if knowledge is something you want to protect, it's really important to understand how it functions and how it's detected and so on. Like, you know, get to know exactly what the dynamics of knowledge are in the social environment and in the larger world in order to be able to protect systems that promote knowledge. So I think like one of our most not basic intrinsic desires is a desire for knowledge to gain,

curiosity, right? You evidently have it around a podcast where you're asking people questions. And I think it's good to understand sort of like what that is and how that works. And that's something that's very deeply integrated now with the theory of knowledge is understanding the nature of curiosity. In knowledge, in one of the proposed definitions of knowledge had the word true. True justified belief. Yes.

Truth itself also has multiple categories like coherence, and pragmatism, and correspondence, and so on. Is there a truth that you subscribe to? So, I think you're alluding to sort of inflationary theories of truth. And those are, I think, all dead ends. So, there's a sort of hyper-minimalist understanding of truth that emerged with the work of Tarski in the last century and has really become dominant for us, where we say that basically arbitrary sentence P is true if and only if P, right? Basically, it's like a disquotational theory

of truth. This is a very, very elegant way of thinking about truth. Like basically the idea is, think of the sentence, the water bottle is on the table. That's sentence number one. Think about sentence number two is, it is true that the water bottle is on the table. What Tarski tells you is those two sentences are in a very interesting way equivalent to each other. You're not really adding anything when you throw, it is true that, on front of what you say. And I think if you have this sort of inflationary view, right, like what if you had the idea that truth is what you can get everybody to agree on, right?

Find all the people, see if they all give it a thumbs up, and then it's true. That's a very inflationary view. Also gives you kind of like a shortcut for checking what is true, right? It's like a cheesy hack. Let's just see if everybody's got their thumbs up and then we'll call it true. What a guy like Tarski is saying is there's no cheesy hacks, right? If you've got the collapse conjecture and you want to know whether it's true, you can't have a shortcut, which is just like, let's see how many people have their thumbs up.

That's not going to work. So I think my understanding of truth, like my understanding of knowledge, is extremely minimalist. There are some very, very important functions that are served by the concept of truth, right? You can use it to sort of quantify over a bunch of sentences. You can say, many of the things he just said are true, or some of the things he said are not true, right? Those are meaningful expressions. Those are sort of, there's some very important uses of the truth predicate.

But we don't add an extra tag or an extra gold star on a sentence when we throw the words is true after it. We add nothing at all. And I think that's something that's very important. So the question of figuring out whether the atomic number of gold is 79 or not is just equivalent to the question of figuring out whether it is true that the atomic number of gold is 79. Get out your mass spectrometer.

Get out whatever you need to do. Like you've got to apply your first- order research methods to all questions. Also, I am not a logician, and you should have this conversation with somebody who actually is. I'm like, this is not in my wheelhouse, but that's just a buy the buy. But I will say like the idea that we shouldn't be invested in the coherence theory of truth and so on, that's very commonly held in contemporary philosophy.

Interesting. So is there anything about Tarski's definition that relies on a mind-independent notion of fact? Well, that's exactly it. Does it rely on the notion of truth? Like, the thing is you're not reaching for something else to reduce truth. Truth is X plus Y plus Z, right? That never felt right, did it?

Because it always seems like you could ask the question, well, but is it true that X plus Y plus Z is the case or something like that? Truth is not the sort of thing that you should be trying to do a reductive analysis on. It's not like salt. Salt is sodium chloride, right? And there's, it's a compound and there's two simpler things that you stick together to make it. And truth is a non-compound notion. It's a prime notion. It's a notion in terms of which other things are explained. It's bedrock. If anything is bedrock, truth is bedrock. And so that's why the best you're going to be able to do is show, here's the function that's performed by the truth predicate within a language.

You will not say, like, wouldn't it be disappointing if truth turned out to be like this kind of like weird mishmash of all these bits and feathers that you glued together? So what I'm not getting, or maybe my conception of Tarski's truth. Characterization of truth. Yeah, is not correct, is not true. Okay, so there's a set of propositions. Okay, this water bottles on the table is not on the table. Truth is a part of the one that is the water bottles on the table or what?

Like, what's the relationship between truth and the water bottles on the table? That's distinct from this. One of the propositions is true and one water bottles on the table that's distinctive from this? One of the propositions is true and one of them is not. What makes that one true? The fact that I put that thing there. What makes it true that there's a water bottle on the table is that I thought it might be thirsty during this interview and I brought it and I put it there. So then there's a relationship between propositions and then just the physical nature of the world?

Propositions encode various facts about the physical nature of the world? Propositions encode various facts about the physical nature of the world and you should 100 percent be having this conversation with a logician and not with me. I'm just going to say that again. Like, yeah. But yes. Yes, propositions are the information payloads of declarative sentences,

and they can characterize the world in various ways. But like what I'm pushing back on is just the idea that there's gonna be some kind of one weird trick. Here's one crazy hack. Here's what truth is. Truth is, you know, this kind of gold star feeling when you look at a sense. I don't think so.

Every single claim that someone makes about the world, stands or falls and how can you know whether it stands or falls? By applying various cognitive capacities of yours to that question. Things go well, we'll end up knowing. But there's lots of things you can't know, of course. Maybe you can help me formulate this question because I haven't said it before, so it'll just come out somewhat incoherent. Those are the best questions though, the questions that we haven't said before. So propositions, so contradictory propositions, and contradictory actions, so performative contradictions, are usually seen to be separate

classes. Now, if propositions overlap with the physical world, then that seems to me like there must be some contradictory propositions that are the same as contradictory, like, or performative contradictions. I don't know what you mean. I don't know what you mean. Oh, so, so, so,, so I think there's a difference between contradictory propositions like the water bottle both is and it's not on the table.

That can't possibly be true. Performative contradictions aren't exactly contradictory in the same sense. So performative contradictions. If you had this person both is and is not making a fist, that can't happen. But you could have somebody saying something which belied itself in some very, very fancy vocabulary describing themselves as a man of plain words. That would be a kind of performative contradiction, but it's perfectly possible. But there's nothing that's deeply contradictory about that.

The tension is just between the state of self-description and the manifest behavior. It's not sort of logically exciting. Now, I don't believe what I'm about to say. I have a stance on what I'm about to ask, but I want to ask you. Would superposition of particles be a performative contradiction then, where it is both up and down at the same time, in the same way that your hand can be both open and closed? You tell me, you're the physicist.

Note, there's a video in the description about why superposition isn't about particles being in two states at once. It's also about why the path integral isn't about particles taking quote unquote all possible paths simultaneously. Okay, but I want to know your answer first. So my answer is you should ask Jan-Anne as well, that question. Okay, well I would say that we say with our words that the particle is both up and down, but that's a wave function and that's not the same as the particle.

So you can't actually make a statement about what the particle is doing until it's measured. That's what I would say. The particle isn't actually up and down. Our theory is agnostic to whether it's up and down. You have to put an external interpretation on quantum mechanics. Cool. Okay. Anyhow, what are you working on these days? What's exciting you? So I am finishing a book called

Recognizing Knowledge Intuitive and Reflective Epistemology. I've been trying to finish this for a long time. I'm very excited about common knowledge formation, and I'm excited about the point at which theory of knowledge sort of meets live action, particularly in conversation. We have a lot of big data on that that I'm looking at with excitement. I'm going to China this summer,

and I'm going to be talking about the book with a bunch of students there. So I'm excited about that. That should be fun. I really love getting perspectives that are from outside the box on what I'm doing. And going to different educational settings is often a very good way to do that. Weirdly, I'm just now remembering a Zoom class that I taught in 2020 during the pandemic with a group of students, I believe, in the United Arab Emirates who were,

they were, I forget, I gave some kind of Zoom presentation. They asked some questions afterwards, and one of them asked the question, I gave some kind of Zoom presentation. They asked some questions afterwards, and one of them asked the question, asked a question about the first person and third person examples in cases of knowledge. Epistemologists write a lot about S knows that P, some subject Smith knows that the door is locked.

How do you analyze that? We also talk a lot about what I know. Do I really know that I have hands? She was like, well, why aren't we talking about the second person? Why aren't we talking about you and what you know? And I flubbed that and gave some kind of half-assed reply, but I kept thinking about that. And I've thought about it a lot more since. And I think there's something extremely significant in second person monitoring.

Like when you're interacting with somebody, you're watching what they do and don't know. And the student, I don't know what her name was, but that question really stayed with me. And I think one reason why she was emboldened to ask that question was she hadn't been steeping in all the SNOs that P, third-person attribution literature that I had. So for her, like, okay, first person, third person, second person, why not talk about second person? And I feel that whenever I get out into

a different space with different kinds of students, they will push me in places that I hadn't expected, and that's just always exciting. I always enjoy that. Thank you so much for- Thanks, Kurt. It's been a pleasure. It's been fun talking. That's the I always enjoy that. Thank you so much for... Thanks, Kurt! It's been a pleasure. It's been fun talking. That's the best kind of interview. Thank you. Thank you. I enjoyed it too. Take care. If you're a professor or lecturer and there's a particular standout episode that your students can benefit from,

please do share and as always feel free to contact me. New update! Started a sub stack. Writings on there are currently about language and ill-defined concepts as well as some other mathematical details. Much more being written there. This is content that isn't anywhere else. It's not on theories of everything. It's not on Patreon. Also, full transcripts will be placed there at some point in the future. Several people ask me,

hey Kurt, you've spoken to so many people in the fields of theoretical physics, philosophy, and consciousness. What are your thoughts? While I remain impartial in interviews, this substack is a way to peer into my present deliberations on these topics. Also, thank you to our partner, The Economist.

Firstly, thank you for watching. Thank you for listening. If you haven't subscribed or clicked that like button, now is the time to do so. Why? Because each subscribe, each like helps YouTube push this content to more people like yourself, plus it helps out Kurt directly, aka me. I also found out last year that external links count plenty toward the algorithm, which means that whenever you share on Twitter, say on Facebook or even on Reddit, etc., it shows YouTube, hey, people are talking about this content outside of YouTube, which in turn greatly aids the distribution on YouTube. Hey, people are talking about this content outside of YouTube, which in turn greatly

aids the distribution on YouTube. Thirdly, you should know this podcast is on iTunes, it's on Spotify, it's on all of the audio platforms. All you have to do is type in theories of everything and you'll find it. Personally, I gain from re-watching lectures and podcasts. I also read in the comments that hey, toll listeners also gain from replaying. So how about instead you re-listen on those platforms like iTunes, Spotify, Google Podcasts, whichever podcast catcher you use.

And finally, if you'd like to support more conversations like this, more content like this, then do consider visiting patreon.com slash Kurt Jaimungal and donating with whatever you like. There's also PayPal, there's also crypto, there's also just joining on YouTube. Again, keep in mind it's support from the sponsors and you that allow me to work on toe full time. You also get early access to ad free episodes, whether it's audio or video, it's audio in the case of Patreon, video in the case of YouTube.

For instance, this episode that you're listening to right now was released a few days earlier. Every dollar helps far more than you think. Either way, your viewership is generosity enough. Thank you so much.