MIT Compass
MIT Compass
What are truth & knowledge?
Use Left/Right to seek, Home/End to jump to start or end. Hold shift to jump forward or backward.
For a transcript of this episode, click here (https://sites.mit.edu/compass/files/2025/06/Truth-and-Knowledge-Transcript-1.pdf).
This episode explores the question “What are truth and knowledge?” with MIT professors from philosophy, history, mechanical engineering, and brain and cognitive sciences.
Featuring: Alex Byrne, Professor of Philosophy (host); Jim DiCarlo, Professor of Brain and Cognitive Sciences; Anette (Peko) Hosoi, Professor of Mechanical Engineering; and Anne McCants, Professor of History.
This podcast was created as part of the MIT Compass Initiative, "21.01: Love, Death, and Taxes.” For more information about Compass, check out compass.mit.edu.
This podcast was recorded at the MIT AV Studios and produced by Adina Karp.
Welcome to MIT Compass: Thinking and Talking about Being Human, a show about exploring fundamental questions,
Professor Emily Richmond Pollock:What do we value and why?
Intro Speaker 2:What do we know and how do we know it?
Professor Alex Byrne:What do we owe to each other?
Intro Speaker 3:by MIT professors from across the Humanities, Arts, and Social Sciences, each episode takes on the moral and ethical questions of the human experience.
Professor Alex Byrne:This week, we'll be talking about truth and knowledge. I'm Alex Byrne from the Department of Linguistics and Philosophy. With me are three other MIT faculty members, first Professor Anne McCants.
Professor Anne McCants:I'm Anne McCants, and I'm an economic historian of the late Middle Ages and early modern Europe. And I also direct a program here at MIT called Concourse. It's for first year students, and our goal is to embed the science General Institute Requirements in a context of social and humanistic study.
Professor Alex Byrne:Thanks, Anne. Next is Professor Peko
Professor Peko Hosoi:Hi. I'm Peko Hosoi in Mechanical Hosoi. Engineering and Mathematics. I also have an appointment in IDSS, which is the Institute for Data, Systems, and Society. My research focuses on fluid mechanics. I've worked on soft robotics, and most recently, I've gotten interested in biomechanics and sports.
Professor Alex Byrne:Thanks, Peko and last, Professor Jim DiCarlo.
Professor Jim DiCarlo:Thanks, Alex. I am a professor in the Department of Brain and Cognitive Sciences. My lab works on vision in non-human primates and building models of visual processing that apply to both monkeys and humans. And I also am the director of the MIT Quest for Intelligence, which aims to discover an engineered level understanding of human intelligence with impacts on human health and in future artificial intelligence.
Professor Alex Byrne:Thanks Jim, and thanks again to all three of you for making the time to do this. So we're going to have a chat about truth and knowledge, and let's start with truth. So the MIT Value Statement mentions the pursuit of truth, or perhaps the quest for truth would have done equally well as something worth valuing. So I thought I'd begin with a deceptively simple and probably unfair question: what is truth? Perhaps there's no easy answer, but if you had to answer in a sentence or two, what would you say? Anne, do you want to have a stab?
Professor Anne McCants:Sure, Alex. I think that this question is so fundamental precisely because it cannot be answered once and for all and for all people from all perspectives. But if you come at it from, let's say, a theological perspective, I think it has, I mean, historically, it has been answered from that perspective as being the transcendental reality, if you will, that is out there, that is beyond us. So that's one way that many people have thought about what truth is, something outside of ourselves that really exists, even if our ability to access it is limited.
Professor Peko Hosoi:Yeah, I'll build on what what Anne said. So I like this, her framing of something that exists outside of our ourselves. I think, you know, putting my engineering hat on, I think table stakes is that it has to be something that is supported by empirical evidence. And you know, like Anne said, that's not something that is an absolute, because the way we interpret empirical evidence shifts as we learn more or as time progresses. But I like this idea of something that is in some way absolute, beyond our perceptions.
Professor Alex Byrne:Excellent, Jim?
Professor Jim DiCarlo:I agree completely with both with Anne and Peko about being outside of us. I would think of it as something that's in the limit that we will never actually achieve, but it's an aspiration, and so I think we'd shift the conversation again, maybe thinking also as engineer from truth towards utility, and how we seek truth to actually gain utility and but truth is the aspiration which we never actually achieve.
Professor Alex Byrne:Okay, excellent, right? So there's a there is a sort of common theme that that truth is an ideal or an aspiration, and perhaps we'll we'll never reach it. You know, you can pursue a rabbit but never catch it. And similarly, the idea seems to be, you can indeed, we should pursue truth, but that doesn't mean we'll ever find it. But what do you think of the sort of flat footed response that surely we can all agree that it's sunny today, but then assuming you agree with that, you wouldn't disagree with the addition that it's true that it's sunny today. I mean, if Anne said it's sunny today, and I said, you know, Anne, that's true. It's not that you Peko and you Jim, would be saying, "Oh, well, that's a terrible mistake. Because, you know, yeah, Anne's assertion was fine, but when Alex followed it up with that's true, that was really overreaching, because truth is this aspirational thing that we can only aim to but never actually reach."
Professor Anne McCants:Alex, I I think you're pushing our linguistic strings a little bit here. The the form in which you use true, right? The casual conversational "Anne made an observation. Alex said, 'Yeah, I think that's true.' " What you're really saying is "that's how I see the situation as well. I look up in the sky, I don't see clouds. The sun is out. It's a sunny day." And I think that that kind of true, and the truth you asked us about at the outset are, they are semantically related, but they are not the same thing.
Professor Alex Byrne:So there's kind of truth with it, with a little T and truth with a with a capital T. And
Professor Anne McCants:That's more or less how I operate.
Professor Peko Hosoi:Yeah. I mean, so if we think if we were 2000 years ago and somebody went outside and said, hey, the sun goes around the Earth. Somebody else would say, yeah, that seems true. Yeah, that's what it looks like to me, right? So there isn't a, there is sort of an acknowledged agreement between those two people, but it's, it's not related to what actually happens between the Sun and the Earth,
Professor Alex Byrne:Right. Or I was going to say "that's, that's true," but maybe I shouldn't. But of course, in your envisaged example, the first person was was incorrect, saying that the sun goes around the earth. And so because of that, when the second person said, that's true, that was, that was just another mistake,
Professor Peko Hosoi:Correct, yeah. Actually, I want to, I want to second what Jim said early about the utilitarian value of getting close to truth. There's the famous quote that all models are wrong, but some are useful, right? And so I think when we make those models we are we are trying to uncover something fundamental and something that gives us predictive power, and sort of the more accurate, or the more predictive those models are, so they enable us to become better decision makers. I think that's sort of the direction of and putting air quotes now "truth" in engineering.
Professor Jim DiCarlo:That's exactly right. That's the George Box quote, and as you say, you approximate it. The models lead to predictive power. The models try to embody our current statement of truth, but the models also accept that they're not yet true. If we take them seriously. And to push on your point about who's wrong about the earth being we actually don't know. Right now, the evidence suggest, if you take it at its limit, right right now, the evidence suggests that the planets are rotating around the sun, but if you really push the limit, we don't really can't prove that right? So again, that sounds odd, but this is really pushing against the notion of you never actually get to absolute truth, but, but it's more utilitarian to accept the second view for the moment, it makes better predictions. So we're going to go with that for now.
Professor Alex Byrne:Okay, well, shall we segue to to knowledge. We've been talking about truth and reaching the truth or finding something out, discovering that the butler committed the murder, seems to be coming to know that the butler did it. So just bracketing Jim's skepticism about whether we can actually know anything for the moment, let's begin with the MIT mission statement. Perhaps, yeah, should be read as aspirational rather than as a claim about what MIT actually does. The mission of MIT is to advance knowledge and educate students in science, technology and other areas of scholarship that will best serve the nation and the world in the 21st Century. So the mission is advancing knowledge and also preserving and disseminating knowledge. And so Jim, I think I mean consistent with your view about about knowledge being aspirational, you could still say that, well, we, we are a massive improvement over Aristotle, as far as getting close to the truth goes, because our theories are much closer to the truth than Aristotle's. Or just putting it more, just putting it entirely pragmatically, our theories are much more useful than Aristotle's. They enable us to do things that Aristotle's theories didn't
Professor Jim DiCarlo:I agree and I liked how you put the word--again, it's not more we they're more predictive, they're more useful.
Professor Alex Byrne:Yeah, yeah, right.
Professor Jim DiCarlo:That doesn't make them right. It also doesn't make the other ones wrong. Back to Anne's point, they were just less predictive and less useful, right? So I think getting away from these black and white notions of right and wrong is part of the challenge here, for for students and for scientists, right, like you discovered oxygen. No, we have a better model that includes this thing that we call oxygen at the moment. That sounds odd, but that's actually at the core what's going on, right? And it's just, we're taught, of course there's oxygen. Well, you know, for a while, of course there was phlogiston, right? That doesn't make these--you can't accept the permanence of all of that, and it's that makes life feel unstable a little bit. But if you just think of it, this is our current belief system, and here's how well it can do. And it will evolve in ways that we can't yet predict. We'll work within that for a while, until somehow something doesn't work right. Then somebody smart will see, oh, maybe if we think about it this way, we'll shift, and that will require people and communities, as you say, to make that happen. But it's sort of this notion that we don't MIT students shouldn't show up and say, everybody that professors know at MIT, here's all the truths, and they're going to teach them to us, and then we'll go forth and we'll know it. No, it's the process of discovering truth. And how do we approximate? How do we even know we're getting closer? What do we even choose to measure? Is this sort of social part of that as well,
Professor Alex Byrne:Right, so really you want to, I mean, maybe it's not completely clear what this notion of getting closer to the truth amounts to, but what, I think, what you fundamentally, fundamentally want to say, is that our theory is just much more predictive and much more useful. And that's what that's what scientific progress consists in, not in discovering the existence of oxygen, or realizing that there's no such thing as phlogiston,
Professor Jim DiCarlo:Building a model that includes one but not the other has then in context with other assumptions, can lead to more predictive power. That's how I would put it. I don't think those are just a linguistic difference, I think.
Professor Alex Byrne:Right. Can I just ask you, just to go back to the claim about about models? Sorry, what was the famous quote that models are always wrong?
Professor Peko Hosoi:All models are wrong, but some are useful.
Professor Alex Byrne:Okay, so someone might say, "Okay, what if by model you mean mathematical model?" So we could have a mathematical model of the solar system, let's say or mathematical model of a neuron. Or...
Professor Jim DiCarlo:It doesn't have to be mathematical.
Professor Alex Byrne:No, it doesn't have to be mathematical. But just for simplicity, just take a mathematical model as an example. So we have a mathematical model of the solar system, and it makes certain simplifying assumptions to make the computations tractable. Maybe it treats the planets as point masses or something or something like that. So it's it's highly predictive. It enables you to predict where Neptune is going to be next year. But you know, it's not, it's not perfectly accurate. If we run the model out, like billions of years or something, it's going to get something wrong. But for the purposes of flying by Neptune or getting to Mars or something, it's it's good enough. So in that sense, that model is is wrong, and we understand why it's wrong, it's just like almost every mathematical model, it's just a simplification of the real situation. But that doesn't mean that we don't know anything about the actual situation that we are imperfectly modeling. We still know that there are planets and that they go around the sun, and that there's this massive object that we call Neptune and so on. So there could be some kind of mistake in focusing so much on the on the imperfections of models, because perhaps that can sort of blind you to the fact that, well, actually, we really are just flatly right about some things, like, you know, the approximate size of Jupiter, or the cause of the Great Red Spot, or something like that.
Professor Peko Hosoi:Well, I don't think the I think the argument which Jim is making, which I resonate with, and Jim tell me if I'm wrong, it's not. It's not a criticism to say all models are wrong. It's an acknowledgement that part of modeling is to understand where we made those assumptions, and are they good enough for the predictions that we need to make, right? So, so, you know, saying that all models are wrong is perfectly fine. That's That's not saying science or engineering is bad. They're, they are. That is our framework that we need to make progress right now, and we need to understand the limits of those framework, of that framework.
Professor Jim DiCarlo:Right, I agree.
Professor Peko Hosoi:Related to that. I wanted to go back. Jim also briefly said that part of this is getting comfortable with uncertainty, right? And I think one of the things, especially when I think about MIT students, these are students who have come in, who have been terrific in physics, who've been terrific in math in high school, where there are absolute answers, right? You have a problem set. There's an answer that people are looking for. And then you get to college, and you know, you realize, oh, I've learned F equals ma. And then you learn, but wait a minute, actually, it depends on your reference frame. And you learn about relativity. And then you read, Oh, but wait a minute. Actually, you shouldn't be thinking about things as particles. I should be thinking about them as probability distributions, because it's because now I've learned quantum and so we keep as you go through, as you advance, you're actually starting to dismantle some of the things that you took to be absolute truth, even in science, and realizing that there are subtleties that have not that that have not been accounted for through in the early parts of that journey.
Professor Jim DiCarlo:Well, I completely agree with that and getting comfortable with uncertainty.
Professor Alex Byrne:So, well, just just touching on that uncertainty point. I mean, how does that manifest in the in the classroom, if at all? Because, after all, Peko, as you were saying, I mean, this is a big difference between most of the classes that students at MIT take and and their philosophy classes. For example, sometimes I have actually given Psets in in philosophy classes, but the questions are all about, you know, the reading or what some philosopher thought, or whether some argument is valid. Those do seem to have clear answers, but they're not about, like, grand philosophical questions, where there's where there's a lot of of disagreement. But of course, in most of the sciences, you've got your Psets and there's, there's either a right or wrong answer. Grading them is, is totally straightforward. So what do you do, or what, or what could you do to make students aware that there's really a great deal of uncertainty in the scientific enterprise as a whole.
Professor Peko Hosoi:I think there are a bunch of tools that we the students need to develop before they can start to apply them to these sort of bigger stories, right? So like before you can build a house, you need to know how to use a hammer and a screwdriver and a wrench and all those kinds of things. And so there are tools that we are equipping them with, like, how do you solve a linear differential equation? Is this thing going to be stable or unstable, right? And those are things where there's a logical progression. And so once you frame things, put things in these frameworks, you can get a sense of how you think, how you think these systems are going to evolve. So, but you have to be good at the tools. And the tools are things where you can just where you can assign a problem set, practice this a bunch of times, and you will now have the tool. As you move through your curriculum in I think, in most departments, things become more and more open ended. So for example, right now, I'm teaching the capstone in mechanical engineering, and this is a product development class, and one of the things they're thinking about now is feasibility studies. So if I have a product that I I might want to design, and there's no right or wrong answer as to whether or not you should design this product, but you should check, okay, will this, can I run this on a battery, or do I need to build a nuclear power station? Right? And so you can start to apply those tools to understand to kind of scope the problems that you're thinking about. And I think, and again, there's no right answer to that, but there's a, there's sort of a skill in how you apply those tools that you've been developing early on.
Professor Alex Byrne:Right, so just, I suppose, in a way, keeping the focus on the classroom. So sometimes people talk about the scientific method, and laud the scientific method as a particularly reliable way of finding things out. And when I was an undergraduate, I studied physics as an undergraduate, and I don't remember having a class on on the scientific method. No one ever taught me what it was. So do you think that there is such a thing as the scientific method? And if so, what? What is it? And do do historians, for example, we have a historian here, employ the scientific method?
Professor Anne McCants:So, so I had reason actually, for the Compass course, to go back and, you know, sort of look it up officially, right? The, you know, the five steps, although it's six now, because they've added persuasion and communication to the end, which is, which is a good thing, right? But the first step was, you know, to, sort of, to collect all your observations. I don't remember the exact words, and it really presumes a zero step, if you will, that's not listed there. And that is that there are somehow unfiltered observations. And we've already, I think, agreed between the between us, that that doesn't exist in the world. And so that's a real problem for the scientific method, right, as a, as a, as a canned box, you know, if it, if it relies on, you know, a prior step that can't be done, then it's a shaky foundation. You know, that said it's tremendously useful, I think, when students are learning tools, and maybe younger than college students, right to, you know, just sort of have a little list that they can follow, right? So I don't want to bash the scientific method, even though I don't believe in it.
Professor Peko Hosoi:I mean, it's one framework, but it's not the only framework that you would apply to advance knowledge. I would say, I'm going to now circle back, Alex, to your because you originally came to the work in the MIT mission statement, which is to advance knowledge. I mean, it's implicit in that, that we're not that there's no end point here, like to Jim's point, that to advance knowledge, that means there's more things to uncover, right? There's we're always peeling back more layers which improve our understanding or improve our predictive power of the world.
Professor Alex Byrne:So there are different ways of knowing.
Professor Peko Hosoi:Yeah, yeah.
Professor Alex Byrne:Right, right, right. Excellent. Okay, so, so the mission statement does presuppose that all of us at MIT are really in the same business, at least at a certain high level of abstraction. We're all in the truth knowledge business. We're all pursuing, we should be pursuing the truth or advancing knowledge in our different disciplines
Professor Anne McCants:And to different ends.
Professor Alex Byrne:Yeah.
Professor Jim DiCarlo:Yeah.
Professor Alex Byrne:Not totally sure what the takeaway is.
Professor Peko Hosoi:Good luck. Alex, yeah.
Professor Alex Byrne:So it's hard to sum up this very wide ranging discussion. So instead, I'll end with a completely different, different question, just to humanize the three of you, if you're not humanized enough already. So. So the question is, what artwork, whether that's a book or a piece of visual art or a musical composition or a poem or whatever was your favorite or the thing that influenced you most in in college.
Professor Peko Hosoi:This is going to be a somewhat long answer. So the first when I, when I, when I got this question, the first thing I realized is that college was a really long time ago. Pretty hard to figure out what I was doing in college. The second, once I started thinking about it, is that this is actually a very personal question, and so, so I will, if you answer honestly, you will learn a lot about about us. So I will answer honestly. I'm going to give you a few things. The first is when I was in college, that was when this amazing new technology came out called the musical CD. And so I had, I I got very excited about buying CDs. And it was also around the time that sort of like, like the alternative rock seed came out. So I think that, I think the CDs I listened to most were the was probably Jane's Addiction, if I'm gonna be honest, and Peter Gabriel, I have the whole Peter Gabriel collection. Then I was thinking, Okay, what about books? Because I also, I also did a lot of reading, and there's, there's two that I remember from that time. One is a book called "The River Why" by David Duncan. And this was, I'm from Oregon, so this was recommended to me by my friends in the Pacific Northwest. And again, I think it, it sort of fits with this podcast, because it is about, it's a coming of age and and how to make choices about leading a good life. And the other one that I remember from college was"A Canticle for Liebowitz" by Herman Miller, which I randomly walked into a used bookstore and just picked it off the shelf, and it blew my mind. So those are the ones I'll contribute.
Professor Anne McCants:I'll start by saying that I grew up in a evangelical Christian family milieu with very thoughtful parents who were in the process of asking more complicated questions than the families that they had come out of. And when I got to college, I started reading theology, sort of as a side project, and I actually was having trouble, sort of thinking of any one particular influence, but it opened up a whole new world for me, in part because I discovered that people's religious beliefs had these deep histories that were complicated, right and and I think to some extent, Both my parents are scientists, as are three of my four grandparents, and so, you know, I'm the "which one of these is not like the other" in my family. And I think in part, I became a historian because that was a way of thinking about theology without having to become a theologian. And I think my interest in the kinds of questions we've been talking about today came, you know, came out of that.
Professor Jim DiCarlo:Yeah, so that, if I may, the personal side of this for me is I was raised like my parents were both scientists, also, my dad was a physicist. And they raised me in a Catholic family, and I remember hearing sermons about how evolution was impossible, and I was revolting at the time about going to college. And my book that I remember in college was the was Richard Dawkins'"Blind Watchmaker," which sort of explains how complexity could arise without a creator, right? And it's very convincing, if you're into reading about evolution and how biologists think about that. And so that in reading that, in parallel with, you know, looking at my other book I remember in college, is just this huge textbook, tome of the, you know, principles of neuroscience, which is essentially should be called the collection of facts about the brain, which beautiful drawings of anatomy, and, like, look at all these complicated wiring systems, so juxtaposing that those tomes of, well, here's all these wires. We don't actually know what they're doing with these kind of, hey, some process has produced this that was, you know, by selection. And there's an opportunity here that at some point we'll figure out this, who we are as a machine, and and understand what all those wires are doing, and that that sort of inspired me to to move on to sort of do this that I do now.
Professor Alex Byrne:I also second what Peko said about Peter Gabriel, early Genesis. You can't beat it. Yeah. Of course, no one listening to this podcast will have any idea what we're talking about, none whatsoever.
Professor Peko Hosoi:None. Zero. Yeah, exactly, yeah.
Professor Anne McCants:I was busy listening to Bach and Palestrina in college,
Professor Alex Byrne:Okay, okay.
Professor Anne McCants:And I sadly haven't changed at all.
Professor Alex Byrne:No. Well, there is, you know, there is that sort of imprinting effect. There's Peter Gabriel. And I do remember reading Bertrand Russell's little introduction to philosophy called--which still can't be beat. It's flawed in so many ways, but, but it's a, certainly a gripping read, I think, to an undergraduate interested in the big questions. And that's his little book called "The Problems of of Philosophy" 1912, I think. Thank you so much. And Peko, Jim, I really enjoyed the conversation, and I really appreciate you taking the time.
Professor Peko Hosoi:Thanks, Alex.
Professor Anne McCants:Thank you.
Professor Jim DiCarlo:Thank you, Alex for having us.
Intro Speaker 3:Thanks for listening to MIT Compass: Thinking and Talking about Being Human.
Intro Speaker 1:We hope you'll check out other episodes at compass.mit.edu
Intro Speaker 2:This podcast was created as part of the MIT
Compass Initiative, "21.01:Love, Death and Taxes."
Intro Speaker 1:This podcast was recorded at the MIT AV studios
Intro Speaker 3:and produced by Adina Karp.