“Time” and “the brain” are two of those things that are somewhat mysterious, but it would be hard for us to live without. So just imagine how much fun it is to bring them together. Dean Buonomano is one of the leading neuroscientists studying how our brains perceive time, which is part of the bigger issue of how we construct models of the physical world around us. We talk about how the brain tells time very differently than the clocks that we’re used to, using different neuronal mechanisms for different timescales. This brings us to a very interesting conversation about the nature of time itself — Dean is a presentist, who believes that only the current moment qualifies as “real,” but we don’t hold that against him.
Support Mindscape on Patreon.
Dean Buonomano received his Ph.D. from the Department of Neurobiology and Anatomy at the University of Texas Medical School, Houston. He is currently a Professor in the Department of Neurobiology at UCLA. His lab studies how the brain perceives time and constructs models of the external physical world. He is the author of Brain Bugs: How the Brain’s Flaws Shape our Lives and Your Brain is a Time Machine: The Neuroscience and Physics of Time.
[accordion clicktoclose=”true”][accordion-item tag=”p” state=closed title=”Click to Show Episode Transcript”]Click above to close.
0:00:00.1 Sean Carroll: Hello, everyone, welcome to the Mindscape Podcast. I’m your host, Sean Carroll. Long-time listeners and readers of my books and things like that will know that I have a very deep interest in time and what time is and the physics of it, but also how it floats into our everyday lives. We human beings use time all the time, as it were, and so that hooks up with another frequent interest of us here on the podcast, which is neuroscience, how the brain works and how the brain thinks about things.
0:00:28.6 SC: So there’s no better person to talk to about this stuff than today’s guest, Dean Buonomano, who’s a neuroscientist at UCLA, who specializes in how the brain tells time, how the brain thinks about time. He’s the author of two books, one is called Brain Bugs: How the Brain’s Flaws Shape Our Lives, and the other is Your Brain is a Time Machine: The Neuroscience and Physics of Time. I have to admit, in fact, that I’ve often gotten this wrong, when I go around giving talks about time, I said something that I now realize or I’ve recently figured out isn’t quite right, about how the brain tells time.
0:01:05.9 SC: We have artificial time-keeping and time-measuring devices, clocks and so forth, calendars, and basically these work by looking at oscillations, it’s something that happens over and over again, and we accumulate a certain number of oscillations to see how much time has passed. As Dean will explain, that is basically not how the brain works, the brain uses other mechanisms, it’s more like an hour glass than a ticking watch or something like that. But more importantly, it’s not just that the brain measures time, the brain uses time to model the world, because our world is four-dimensional. There’s three dimensions of space, one dimension of time, so what the brain does is to construct a model of reality because we human beings need to live in reality, we need to predict what’s going to happen next, we have a feeling that certain results will occur if we make certain actions and so forth, obviously time plays a huge role here.
0:02:01.3 SC: And I think that’s what is Dean’s deeper interest here, ’cause he’s interested in the laws of physics as well as in neuroscience, he’s trying to understand how the brain constructs a model of reality and what relationship that has to actual reality, right. Are we just constructing an illusion, are we finding, touching something real about the world, and in fact, we end up having more of a conversation between Dean and myself in this podcast episode that I usually have here on Mindscape because we’re interested in the same things, but coming at them from two different perspectives. So we don’t agree about everything, but we agree about a lot, and I think that we really got into some deep fun questions about what is reality, what is time, how do we measure it, how do we think about it. So I think this is a really great episode that I’m very happy with. Let’s go.
[music]
0:03:08.0 SC: Dean Buonomano, welcome to the Mindscape podcast.
0:03:11.2 Dean Buonomano: It’s a pleasure to be here, Sean, thank you.
0:03:12.1 SC: Every time that I deal with someone who is a neuroscientist or thinking about consciousness and things like that, I can’t help but think, wow, this is really, really complicated, I’m glad I don’t do it for a living. Are you the kind of neuroscientist to help our audience figure this out, who sort of studies neuron by neuron, or function by function, or thinks big picture about consciousness and what that is?
0:03:35.2 DB: Yeah, it’s interesting that you put it that way. And it’s… Neuroscientists feel the same way about theoretical physics, of course. It’s glad that I don’t do that. I’m certainly… Consider myself to be an neuroscientist who tries to understand the big picture question, so it’s a huge field, as you know, and some neuroscientists study molecular biologies and spend a long part of their life trying to focus on a particular molecule of protein that might be involved in a disease. I really focus on how the brain performs computations, how the brain makes, processes time, engages in decision-making and so forth, learns and so forth.
0:04:17.0 DB: But you start off with a question of, it’s very complicated, and indeed it’s so complicated that I don’t know if we have any right to think that we can understand, ’cause the brain is the most… Neuroscience is the most recursive field of all scientific fields because neuroscience is the only field in which the thing being studied is doing the study. So I don’t know if we’ll figure it out or not, but it’s certainly fun trying, and there’s a lot of job security, ’cause I don’t think we’ll actually make much progress within my lifetime towards the deeper questions.
0:04:50.1 SC: I suppose cosmologists could say that it’s the universe trying to figure out the universe, but it’s a little bit less direct than in your case.
0:04:56.6 DB: You know, there’s that anecdote of a neuroscientist and a physicist talking and the physicist says, the most important problem in science is what’s the origin of the universe. And the neuroscientist says, what organ led you that decision? So yeah, they’re interactive, and this is a fun aspect, I think, of… That’s fun to talk to you. I mean, there are links here between the neuroscience and the physics.
0:05:23.0 SC: Well, and also between the world of computation, you use the word computation as something that the brain does, and I always hesitate a little bit about that word, because clearly the brain does computations, but it lets us or it invites us to use actual physical computers as a metaphor for the brain, and in many ways, the brain is not designed the same way as my laptop is. Is that something you think about a lot?
0:05:51.8 DB: Absolutely. And again, we probably just want to define what we mean by the word computation. So of course, the brain doesn’t form computations in a serial manner like a von Neumann architecture like our serial computers do, but it certainly is a computational device in the sense that it process information, it has inputs, it generates outputs. Now, in terms of the comparison between digital computers and biological computers or the brain, at least, they’re adapted, they’re optimized for different computations. So I went into neuroscience because neuroscience, it was fascinating that the brain could bestow us with consciousness and the ability to communicate and to pat our tummies and rub our heads at the same time and other cool things like that, but as I grow grumpier the brain also sucks him anyway, right.
0:06:49.3 DB: The brain… If you ask me what 62 times 79, I’m like, let me go get a calculator, which is a bit embarrassing, because the brain is the most sophisticated computational advice we know of, and by almost any measure of computational complexity, multiplication is much easier than recognizing faces, so why is it so bad at some things versus others. So any piece of hardware is well suited for some computations and poorly suited for others.
0:07:22.4 SC: I do wonder whether or not we would be able to artificially enhance ourselves along those lines, like I think… Embarrassingly, I’m going to forget who it was that I was talking to about this, but I conjectured that there was sort of an uncertainty principle that made it difficult to be a conscious creature and do math perfectly at the same time, right. Not only were we not optimized for it, but there’s the reason why we’re bad at it, but on the other hand, if we just get a neural lace or an implant that we can plug into a phone or a calculator, could we imagine enhancing our abilities along those lines pretty straightforwardly?
0:08:00.6 DB: Yeah, I think that’s an increasingly topical and relevant question. First, in regards to the trade-off, what you’re getting to is a trade-off between, say, being conscious and doing mathematic computation. So I agree with that. And if you look at neurons as computational hardware, computational building blocks, they’re not well-suited for math, so no engineer or computer scientist would use neurons to build a calculator, because math benefits from digital precision, digital switch-like properties. So transistors are great switches, neurons are horrible switches, they’re great coincidence detectors. So people say, well, why are we so bad at math? And people say, well, because we didn’t evolve to do that. I would actually disagree with that, I think, of course, that’s one reason, but I think additionally, that the biological hardware evolution had at its disposal is not well-suited for numerical calculations, mental numerical calculations. So I think it’s a hardware problem.
0:09:03.1 DB: Now, in our ability to interface with a brain machine, interfaces, neurohybrid computing and so forth. Absolutely, I think we’ll take step in those directions. It’s interesting, right? And I think there’s a bit of fanfare or a bit of exaggeration in this field. We say, well, we can interface computational devices with the brain maybe directly in a neurolinks type fashion, directly implanting them into the brain, but the fact of the matter is, Sean, we already have pretty good interfaces, right, they’re called eyes and ears, and they’re not bad interfaces and we use them quite well, so this idea that we’re going to dramatically expand our cognitive abilities with these interfaces, I’m a bit skeptical.
0:09:58.0 DB: And let me just give you one example there. Thought, so knowledge is the infrastructure of thought, so in order to think about the problem, you have to have the data already built in, embedded into your neural circuits. If I ask you to think about a problem in physics, to create some novel idea, you need to have those primitive concepts already embedded in your circuits, that’s why you think about something differently than I think about something. So the idea that somebody who doesn’t know anything about quantum physics will just tap into their mental interface and load that up into their brain, they might gather information, but I don’t think they’ll be able to build upon that information, because that information is the substrate of creativity and thought.
0:10:49.5 SC: Could you… I want to follow up on that, but first I’ve just got to go back to this neurons not being good calculators a little bit. Can you expand on that a little bit because I think that, again, people do have this metaphor in mind about the laptop and the brain and the fact that the hardware is ill-suited to the task is actually not something I’d heard before, I’d heard the evolution explanation for why we’re not good at multiplication.
0:11:12.8 DB: So the hardware of the brain is based on neurons, so neurons are these amazing computational devices and they’re very extroverted compared to transistors. A transistor is a computation advice that switches from two states, a binary state, and it receives input from not that many other of its neighbors. Neurons are connected to thousands and thousands of its neighbors, and they’re very democratic, they sum up their [0:11:38.0] ____ inhibitory inputs and then produce an output. But once they produce an output, they immediately forget it. They’re not a switch-like property.
0:11:51.1 DB: So as you know, if you’re doing addition of binary numbers, switch-like properties are very important because once you switch, you’re adding two numbers, if you sort of want to carry the remainder, you need to fix the current result, remember what happened, and then bring in the other numbers. Neurons are horrible at that because they’re leaky, they’re not switches is… I think the clearest way to put it is transistors, the foundation of current computers are transistors which have switch-like properties. Neurons are integrated, and certain… I shouldn’t say integrators, I should say they’re coincidence detectors, they sort of sum up the votes and either produce an output or not, but once that output is generated, it’s gone.
0:12:38.0 SC: Okay, but nevertheless, we can add two digit numbers together, so there’s some collective property that is mimicking some virtual machine that is running in our head that is sort of a pretty shaky but still functional calculator.
0:12:53.3 DB: Yeah. Speaking for myself, it’s incredibly shaky, right.
0:13:00.5 SC: Well, me too, don’t worry.
0:13:03.1 DB: You know, if… There’s no way… No human being will ever divide 10, divide two 10 digit numbers mentally without using pen and paper with the same precision and accuracy as even the simple calculator on your watch. So yes, but remember, unlike face recognition, which is something that’s a relatively hard computation, that’s something that most computers have only excelled that recently, human babies do that, they don’t need to go to school to learn to recognize faces. You need to go to school to learn to do the math, or you’re clearly tapping in to symbolic reasoning representations that are not in the brain’s normal wheelhouse.
0:13:50.4 SC: I guess this is a completely different topic than I thought we’d eventually be talking about, but it will eventually lead into questions about how we represent the world and do physics and things like that, so it’s relevant, but do we know much about how the brain has decided to do arithmetic or simple math things?
0:14:10.4 DB: No, I don’t think so. The best we know about is counting, so even counting is tricky, right, so representing numbers symbolically. Can animals count? Animals can do something called subitization, which is judging, looking at an image and seeing which image has more elements, objects, seeds, food, whatever.
0:14:36.5 SC: So that’s a simple comparison task.
0:14:41.3 DB: Yeah, it’s sort of a… Subitization means sort guesstimating the numbers or the quantity or the magnitude of elements in an image, for example. So they can train monkeys to make a comparison task, as you just said, between does this image have more or less of those items, and then they can record from neurons in the brain, and they see that some neurons are tuned for magnitude. But that’s probably not how humans are doing math, we’re probably just tapping into language, symbolic reasoning, and these higher order cognitive abilities that are really more closely associated with human cognition.
0:15:25.9 SC: It’s just… One of the reasons why physicists will always be jealous of neuroscientists, ’cause it’s just so easy to ask a question that we don’t know the answer to, in neuroscience, there’s like so many things yet to be done, which is great.
0:15:38.1 DB: And as I said, yeah, it’s the brain trying to figure out itself, so it’s an uphill battle.
0:15:45.1 SC: And you mentioned the fact that an individual neuron takes in a lot of data and talks to a lot of its friends, and also there’s this well-known fact that the way that we make memories or something like that, it’s not… I say it’s a well-known fact, you’re probably going to tell me I’m wrong here, but it’s not that we change the internal chemistry of the neuron, but we change how the neurons are linked together and the strength of those linkages. Is that right?
0:16:10.8 DB: That’s exactly right. At least, the main mechanism seems to be what we call synaptic plasticity. So if we’re two neurons and we’re connected, learning, or one of the main mechanisms of learning, seems to be changing the weight of those synapses, the coefficients between those two elements. And here there’s very interesting physical analogies. And of course, this is the same way a lot of the current deep learning, machine learning approaches in AI function, of course, is by just adjusting the weights, what neuroscientists call the weights or the synaptic strength between two neurons, and machine learning or in other fields, you can just think of as the coefficients between two elements.
0:16:58.6 SC: Okay, yeah, and that was explicitly inspired by the human brain or by brains, right? Deep learning, neural nets?
0:17:04.1 DB: Yeah, I think a lot of the current… That’s right, a lot of the deep learning approaches were inspired by the brain’s architecture. There’s fundamental differences, of course, and I think we’re reaching a fascinating convergence between neuroscience and machine learning, where the question is, is can machine learning now inform neuroscience is opposed to just the other way around.
0:17:29.9 SC: Well, that’s… Maybe that’s related to what I was going to ask next, which is that not only… I think that what people hear, what I have heard for a long time, is that the strength of neural connections helps us form memories or something like that, but what about this question of algorithms? There’s some self-organization process that has to go on, is it the same thing, is it the strength of the connections and do we know much about how the brain grows its algorithms?
0:18:00.4 DB: Yeah, so that’s a great question. So first, the difference between algorithms and memory, in a way arises from, is influenced or is biased by our understanding of computers with von Neumann architecture, that’s sort of a key aspect of our digital computers, is that they have memory and they have these algorithms that are processing the information stored in those memories. So in the brain, it’s probably not a good idea to necessarily separate the algorithms from the memory, so this was what I was just hinting at is if we have these brain-machine interfaces, and maybe this is a better way to put it, is we have these brain-machine interfaces, the memory is the substrate for the computation, is the substrate for the algorithm.
0:18:54.6 DB: So those things aren’t as clearly distinct in the brain, so it’s absolutely valid to say that how you process information is in part dependent on changes in synaptic strength, and as well as the memory is also dependent on the change in synaptic strength. Now, what governs… What governs how synaptic strengths change? So maybe that’s what you refer to algorithms, and that’s what we refer to the learning rules. So the simplest learning rule we know of is called Hebbian plasticity or Hebbian learning, which just states that the strength between two neurons increases if they’re correlated with each other. So if neuron A is active at the same time as neuron B, then the strength between those two neurons increases. So that’s an algorithm and that’s sort of programmed genetically into the system, and that’s the self-organizing learning rules that govern the emergence of these complex properties, so any emergent system has underlying learning rules, relatively simple learning rules that generate surprisingly complex properties.
0:20:12.0 SC: Good. And one of the uses to which we put that capability is mapping or making a model of the world in which we live. I’m especially impressed by what we’ve learned over recent years. I know your expertise is in time, but let’s just talk about space a little bit, where we are in the world has really interesting reflections in our neural architecture.
0:20:39.2 DB: So yes, we have these very sophisticated spatial maps that allow us to interact with the world, there’s nothing more basic for animals to be able to navigate their world. This is one of the differences between plants and animals. So plants don’t need to navigate the world, they just stay fixed, but animals need to navigate the world, and that requires these sophisticated spatial maps. If you close your eyes, and most animals can navigate in the dark, if you close your eyes, you can navigate in your surroundings because you have an internal map that allows you to do some dead reckoning and to give directions on where you want to go, and this is something all animals are good at.
0:21:24.1 DB: And something here that I think is worth saying is that the same is not true for time, so most animals do not navigate time, and how could they, but most animals, they seem to have a map, so migrating birds can navigate over very long distances. Time is very different. So you and I can say, well, next month, I’m going on a trip that I planned last year, so we have these complex temporal relationships that we can transmit, but most animals seem to live in the present, so they don’t have these temporal maps to the same degree that we do.
0:22:09.0 SC: I did have Karl Friston on the podcast, and I joked with him that I thought that it was cats, domestic house cats, that was the level of evolutionary progress at which the ability to think about the future must have developed, ’cause I have two cats and one of them clearly does it, and the other one clearly does not, so I don’t know if that’s scientifically proven, but it seems to be evident.
0:22:31.6 DB: So give me an example of your cat thinking about the future.
0:22:34.9 SC: Well, so one cat, Caliban, he just lives in the moment, he’s either happy or unhappy, depending on whether the present moment is suiting his purposes, but Ariel, his sister, she knows that if she does certain things, other things will follow and she will plan them out, so she plays fetch, she’ll bring you a little ball and you’re supposed to throw it and come back, and she knows that if she walks to the bathroom, we’ll turn on the shower and she’ll be able to take a shower and things like that, and so her happiness or lack of happiness is clearly dependent on what she anticipates is going to happen in the future in a way that her brother’s is not.
0:23:10.1 DB: How far into the future do you think her actions can go? So are we talking seconds or…
0:23:15.1 SC: Minutes, I think minutes, maybe, I would say minutes, but maybe it’s only seconds.
0:23:21.5 DB: So this is an interesting question, and this speaks to the problem of mental time travel, which for humans mental time travel refers to our ability to mentally… I know we have differing views of whether time travel is possible in physics, but I think we can probably agree that the closest we’ll ever come to time travel is mental time travel. We can sort of revisit… We can revisit the past in our minds and we can simulate possible future scenarios, we can practice and prepare for these podcasts, for example. Now, in that, that ability in many ways is what makes homo sapiens sapient, it’s what makes homo sapiens wise.
0:24:12.9 DB: You think of something like agriculture, which is one of the most transformative technological inventions of our species. That requires planting a seed today and reaping the fruits of that seed maybe a year into the future. That connecting cause and effect over a year is far beyond the cognitive capacity of all other animals, as far as we know. So all animals, the brain is a device that helps animals predict the future, and that includes Caliban, so all animals can anticipate what’s about to happen on the seconds frame, and probably more on the minutes frame, because if they couldn’t, they wouldn’t survive in a world governed by the laws of physics. They need to know where the prey is, they need to know where the water is, they need to know how to avoid being predated.
0:25:11.3 DB: But to think about things in a flexible manner, like saving for retirement or brushing your teeth, everything we do is future-oriented, it’s hard to find something that we do that’s not future-oriented, and I think of my dog, like brushing my teeth, I know I can’t convince him to brush his teeth, it’s hard to explain that this is something that you should do for your future well-being, and it’s not something that most animals can grasp. There is discussion, there have been experiments on mental time travel in some birds, and so there’s a debate that some animals can engage in mental time travel, but it’s still open to me.
0:25:53.8 SC: I love my cats, but neither one of them is going to be retiring, planning for their retirement, effectively, believe me.
0:25:58.1 DB: So you can’t count on them to support you in your old age.
0:26:03.4 SC: I would love to, I would love to. But okay, the juicy time stuff is creeping in on us ’cause we both want to talk about it, but I want to finish with space first. There are literally neurons in our brain that help us, you said, as you said, we close our eyes and we can still sort of have a picture of what’s going on in the world. I was just really surprised to learn that there’s different neurons that have different jobs telling us where we are, telling us how we’re oriented, placing grids on the world around us and telling us where we are relative to different nodes in that grid. This is just amazing to me.
0:26:39.0 DB: Indeed, and this was a topic that led to the Nobel Prize, I think, a couple of years ago, of place cells and grid cells, and people have discovered place cells in the hippocampus, and if you record from place cells in the brains of rats, as they’re traveling a room those specific neurons might start firing in a specific location in that room, or specific amount of distance when it’s running on a treadmill. And then there’s these other more higher order complexity of spatial cells, these are grid cells that you’re referring to, that fire in these beautiful periodic patterns, they’ll fire at one meter in and then two meters in and then three meters in, and so you have these grids, this tessellation, if you will. I’m sorry?
0:27:32.7 SC: And not in between. It’s as if there’s a giant chess board, and without actually the chess board being colored at all, there are neurons that fire if you’re on a black square but not a white square and I don’t know why that’s true. Is there some evolutionary explanation for that?
0:27:49.3 DB: I don’t think there’s a generally agreed-upon explanation for that, Sean, I think mathematically, you’ve seen these patterns and different aspects of mathematically, grids can be useful for encoding information in numerous circumstances. And one thing nice about grids is invariance. Right, so with changing one variable or two variables of a grid, you can change, expand or dilate that grid, expand or contract that grid, and then remap the system to a larger room or a smaller room. But that’s all sort of speculation or theories. I don’t think we have, at least that I know of, I don’t think we have a universally agreed-upon answer to that question.
0:28:35.3 SC: How much of our very, what a philosopher would call the manifest image of the world, our very basic idea that there is three-dimensional space around us full of objects, these objects take up space, etcetera, how much of that is in some sense innate? Is any of it, or do you have to learn all of it? The human brain is not completely a blank slate, but is there a little tiny folk physics built into it?
0:29:00.0 DB: I think so, so clearly, it’s the case that there’s areas specialized in processing spatial information. So humans that have lesions to certain parts of the brain will have typical deficits in spatial navigation or getting lost and so forth, so the brain is absolutely not a blank slate in this sort of deeper level. Now, how that’s applied, everybody will develop specific skills to navigate their local environment, whether you grew up in a forest or a city and so forth. But something interesting about space, you’re talking about the space of around us. There’s another very interesting aspect of this, which is your body, you feel… If I were to stab your hand, you feel pain projected out into 3D space at the point in space in which that piece of meat that is your hand happens to be.
0:30:05.7 DB: So in many ways, I think spatial awareness, the driving force from spatial awareness was body awareness, in that the body is the brain’s most important possession, so it needs to have this beautiful, very compelling map. It’s really amazing, right? Why do I feel my hand as my hand, and it’s an amazingly amazingly powerful illusion, and let’s face it, it’s just a piece of meat attached to my brain through some nerves, but I have a very intimate awareness of where it is, that it belongs to me and so forth, so it’s an amazing spatial illusion.
0:30:46.7 SC: This sounds kind of like a very specific hypothesis that one might even ask about testing, the idea that we developed spatial recognition capabilities generally out of first developing a question of where we are, our bodies are located in space. Is that something that people try to figure out, if that was the right way that it happened?
0:31:09.8 DB: I don’t know if evolutionarily speaking, I think the way I guess I should probably put it as those two things probably had to co-evolve because they’re intimately tied together, so almost all animals have a map of their own body in their brains. So as you know, we have the somatosensory map, the homunculus that maps out on my body, and it’s a very sophisticated map, and by the way, it’s not one map, it’s many maps, there’s many homunculi in our brains that map out different aspects of our body. Now in order for that to be functional, it’s also important to, if I’m going to interact with external world, I’m going to try to eat a fly or pick up an object, that my body map has to also be mapped onto this external map. So now that you’re sort of bringing those up in context, I would say, I think they have to really be tightly, inextricably linked and have co-evolved with each other.
0:32:09.4 SC: It makes sense. I know that people who study virtual reality suggest that you can re-map your homunculus a little bit, right, you can sort of train it to believe it has more arms or legs than it has. I’ve always wondered, not always, that’s an exaggeration, but for a long time I’ve wondered, could you have four-dimensional space in virtual reality, could you train the brain to think that it was in a higher dimensional space?
0:32:34.0 DB: I don’t know, but I’m suspicious of you asking that question as a eternalist, Sean. So I hope not. No, I’m kidding. I think… I doubt it, I sincerely doubt it. I think that’s something that is so fundamental to evolution that it would be hard and hard to really compelling achieve that awareness, four-dimensional awareness. Some professions, and you’re probably in one of those professions, and many other mathematically-rich professions have a lot of benefit in being able to visualize or imagine high-dimensional systems. Neuroscience is that way, by the way, and we all have our little hacks, but I don’t think most people can claim to have a good intuition of four-dimensional navigation, so I think that would be a stretch to have that ability, but I could be wrong.
0:33:41.0 SC: A good PhD project for one of your students. I’m just saying, the student you least want to talk to, send them down this terrible rabbit hole of impossible questions. Okay, good, so that’s… I did want to get space on the board there, as it were, but time is what we’re really here to chat about. So we have a map of space, we have cells that tell us where we are, is time-keeping completely different from that in the brain, or is there a time cell in the brain?
0:34:10.1 DB: I think the answer is both. So first, it’s useful to have a bit of an analogy here with physical devices, because it pertains to scale, so the answer to your question depends on timescale. So here to give some context, man-made devices, if you need to weigh out a gram of salt or your weight in the morning or your car, you have a different physical device to measure mass across scales. Man-made clocks are quite incredible, right, because the same man-made clock, it’s really astounding, can measure microseconds, milliseconds, seconds, minutes, hours, days, and so forth. It’s quite an achievement.
0:34:56.4 DB: The brain can also measure times across a vast range of scales, so what I’m getting to, of course, is your question is dependent on scale. So we can measure microseconds on the time it takes sound to travel from my right ear to my left ear, and days, for our circadian clock, of course. The brain has absolutely fundamentally different mechanisms for timing on those scales, so it’s what we might think of as a multiple clock principle. So the circadian clock certainly doesn’t have a second hand, it just has an hour hand and tells days, but it has no idea of how many days have gone by, so it’s very specialized for that 24-hour cycle.
0:35:43.0 DB: The timing that underlies your ability to detect sound coming from your right ear on the right side of your head to the left side can pick up tens of microseconds, but it has no idea how many minutes have gone by, so the brain, in part because presumably because of the evolutionary constraints of biological systems, it had to develop or evolve different timing mechanisms to solve the problems on different scales. And the circadian clock functions as one, as an oscillator, as a molecular oscillator, it has what’s called a transcription-translation autoregulatory feedback loop, and that just means that it makes DNA and the DNA makes RNA and the RNA makes proteins, and those proteins in turn inhibit the synthesis of more DNA, so you have this recursive system that cycles upon itself. Now, on this shorter time…
0:36:48.1 SC: Once every 24 hours.
0:36:49.7 DB: Yes, approximately once every 24 hours.
0:36:51.9 SC: So that is the thing that is most analogous to the mechanical clocks that we build, where it’s really an oscillator with some period that is more or less reliable.
0:37:01.6 DB: So… Exactly, and that’s a good point to bring up too. So man-made clocks can be very, incredibly into their degree of sophistication, but they really rely on an almost embarrassingly simple concept, just counting the ticks of a time base of an oscillator, that time base might be a pendulum swing or a vibration of a quartz crystal or the caesium atom electromagnetic frequency, but they’re all just counting the ticks of those oscillators. But Sean, it’s important to know that the second part of that requires integration or counting, so it’s sort of timing above the period of the time base, the circadian clock is timing below the period of the time base, so it’s the phase, of course, so it’s telling the phase of the oscillation.
0:37:55.9 SC: So sorry, so to say in other words, the circadian clock doesn’t tell you how many days have passed, it tells you what time of day it is, where you are within that cycle.
0:38:04.6 DB: 100%. So your circadian clock located… Your main circadian clock located in your hypothalamus has no idea how old you are, it just says, time to get up, time to go to bed. Now, this notion of having an oscillator and a counter was really one of the main theories of how the brain tells time for really most of the 20th century, it’s called internal clock theory. And it’s a reasonable theory in the sense that brains… I mean, sorry, neurons can be reasonably good oscillators, neurons are good oscillators, they can go up and down at a fairly constant rate, and that’s a good thing, ’cause that underlies breathing, that underlies chewing, that underlies walking.
0:38:42.6 DB: But as we’ve already discussed, neurons are horrible counters, they’re horrible integrators, so it’s pretty clear today that this internal clock theory is not really the correct one in terms of the main way the brain tells time. Rather, any dynamical system, any system that changes in some reproducible fashion can be used to tell time, in principle, and the brain, of course, is the most complex dynamical system we know of, so it makes sense that the brain uses its own dynamics as a way to tell time, and that seems to be… That’s something that my lab has worked on a lot, and it’s becoming increasingly clear that in many cases your brain is telling time by changing patterns of neural activity.
0:39:40.7 SC: I think you’re going to have to give me more details about what that means.
0:39:44.7 DB: So one way to tell time, you can imagine you have a circuit of neurons, so you have neuron A, neuron B, and the most trivial sort of population clock or neural clock you can think of is just the chain of neural activity. Neuron A activates B, activates C, activates D. It’s a very simple, almost trivial concept, and you can imagine that depending on the time delay between that activation, sort of like dominoes falling down, you can time the dominoes and you know that when domino 100 fell down, 100 milliseconds have passed, whatever. So neurons are dynamical systems. A excites B, B excites C, C inhibits A. So you have all this complex high-dimensional dynamics which can be used to tell time. You imagine in physics, you have a ball rolling down a landscape, and you could use that as a clock if you so desired, as long as it was relatively reproducible with a certain degree of precision. So I think on the scale of seconds, the intrinsic dynamics of neural circuits seems to be one of the key ways the brain is telling time.
0:40:55.6 SC: Is it fair to say, reaching for an analogy, the brain on the scale of seconds tells time with mechanisms that are more like an hour glass than a wrist watch, because there is something that is changing uniformly and you see where you are in that process.
0:41:10.6 DB: Yes, I think that’s fair to say, because the wrist watch is clearly what you’re getting to, is a wrist watch is clearly relying on an oscillation and counting the ticks of those oscillations, whereas a hour glass is just a physical system that’s governed by the loss of gravity in this case, and it has a specific reproducible dynamics that can be used to tell time.
0:41:35.4 SC: Something that I remember from your book is that you can train these mechanisms, you can get better at them, and you can get better at one of them on one timescale without any change in your ability to tell time on other timescales.
0:41:49.6 DB: That’s… Yes, absolutely correct. It’s not even on other timescales, it’s even within the timescales, so if I trained you… If I gave you some software and trained you on discriminating a short interval, very short, like 100 or 200 milliseconds, you would get better at that over the course of days, but your ability to tell time across 500 milliseconds would not improve. So this goes again to the idea of why it’s different from a quartz wrist watch, right. If your watch is tuned well for, it gets better at 100 milliseconds, it’s going to get better at other intervals as well, right? So the idea here is that the brain tells time through, not through a centralized clock, but through many different circuits that are specialized for different functions.
0:42:38.6 SC: I think that the whole lesson of this podcast is the brain is bad at counting.
0:42:45.4 DB: It could very well be, Sean, yes.
0:42:45.9 SC: It would seem like the thing that happens in a wrist watch where you have a second hand, an hour and a minute hand, like you say, it’s the same oscillator that drives them all, but there’s some differential gears relating the actual motion of the hands. That doesn’t seem to be a mechanism that evolution ever thought up, right?
0:43:08.4 DB: It’s certainly something that the brain is not inherently well-suited at doing. As you’ve said, we can do it, we can get by, and animals can too, but it requires… Rather than it requiring one neuron or two neurons, it requires a whole circuit of neurons generally to hack it, to do the counting. And this is clear from our behavior, babies have to learn to count, you have to teach them, there’s a whole native… Native Indians in the Amazon have a 1, 2 and many system of counting, they don’t really know their age, they don’t really count past the number of 10, which is very surprising to us, but counting is something that we in Western-educated… In our Western-educated world take for granted, but it’s not as universal is we think it is.
0:44:03.3 SC: So if you can train yourself, are there certain professions that are very naturally good at keeping time? I was going to guess musicians, and then I thought of, what about comedians? Timing is everything, like are there things that you do that naturally hone your skills at time-keeping in different timescales?
0:44:21.5 DB: How about podcasters? Has your timing…
0:44:24.0 SC: Terrible. Terrible. As you know, we had to delay this for a week because I’m not very good at organizing my time.
0:44:31.5 DB: I guess podcasters have more flexibility in the duration…
0:44:35.8 SC: Can edit.
0:44:38.0 DB: Of the interview and they can edit, but in a lot of professions, so like in radio hosts where they have a five-minute slot so they can get quite good at having an intuitive feel of what that five minutes, but of course… So you’re absolutely correct, musicians have very good timing on the scale of hundreds of milliseconds and so forth. But timing, we’re all pretty good at it, and here’s an example. So you said, you mentioned comedians, and of course, timing, as we often say, is critical in comedy, but it’s really critical in speech, and here’s an example. I’ll say two phrases, and depending on the timing of my speech, the meaning is different. They gave her… Cat food, or they gave her cat… Food. So there’s two very different meanings there, and so you’re continuously telling time.
0:45:33.7 DB: Everything you do is occurring in time. And another example of that is right now, your brain is unconsciously attempting to predict what I’m about to… Say, so by pausing my speech, that created what’s called a temporal prediction error in your brain, and you said, uh, has Dean forgot his word, am I supposed to say something? And so forth. So your brain is unconsciously continuously telling time and attempting to predict what’s about to happen.
0:46:03.0 SC: This is actually of crucial importance for podcasting, because a lot of people listen to podcasts sped up… I don’t know if you know this, right, and the best way, but you want to speed it up, but not obviously increase the pitch of everything, so you clip and you get rid of all the silences and things like that, and people have started to complain that if you listen to the podcast sped up all of the nuance of the timing and the speech gets evaporated away.
0:46:26.9 DB: Yeah, it’s interesting, and I wonder if you’re training people’s brains to process speech at faster rates and so forth, so again…
0:46:35.2 SC: Again, full of…
[overlapping conversation]
0:46:37.3 DB: It is an interesting… Somebody should be studying that, I agree.
0:46:39.1 SC: And other things besides just training, I’m sure there are other things that affect the accuracy at which we can tell time. We briefly talked when David Eagleman was on the podcast about being scared, having your adrenaline going and how that affects your time-keeping, but presumably the brain is awash in signals and chemicals and influences, and these have an influence on how well we’re keeping track of time.
0:47:05.6 DB: And psychoactive drugs, and everybody reports that this influences our perception of time, and this is a deep mystery, but in many ways, I think the question might be misdirected. So clearly, we all know that the subjective flow of the passage of time can be dependent on context, whether you’re bored or excited, or having fun or engaged. Hopefully for the listeners of this podcast, they’re saying, wow, time really flew by.
0:47:37.3 SC: Riveted.
0:47:38.2 DB: And then, of course, in near-death situations, time often is reported to have slowed down. So you can imagine a couple of things underlying that, so one, you can imagine that the brain is being over-clocked, so just like you can increase the speed of your computer, you can over-clock it. Everything happens faster, so things transpire at a faster rate. The brain, there’s no reason to believe the brain can be over-clocked, neurons don’t work like that. Maybe you can accelerate things by 10%, but there’s no clock to be over-clocked, and there’s physical time constants that are constrained by the physics of neurons and so forth. So it’s very hard to imagine that you’re going to get… If you’re full of adrenaline that you’re going to get neurons processing 20, 30, 40, maybe 20, 30, 40% faster, but not 100-200% faster.
0:48:32.9 DB: You might also imagine that it’s an illusion, in the sense that time doesn’t change, it’s just your memory of what happened changes. So it’s sort of like a flashbulb effect, that it’s not when I was in a car accident time seemed to slow down, it’s that when I have very good recall of what happened and that produces that effect of slowing down. I actually recount the story of when I was in a car accident and my car was spinning around, that I’ve caught myself thinking, wow, time really does slow down when you’re about to die. I don’t think it’s memory either. So I think the correct question is, what is our normal sense of time, and if you give me some leeway and let me go back to an example with body awareness.
0:49:23.5 DB: So as you know, there’s a syndrome called phantom limbs, so people can feel… People who have suffered an amputation can still feel that limb as vividly as you and I feel our own limbs, which sounds very bizarre, but the real lesson in phantom limbs is not that the brain somehow miscalculates and generates an illusion of the phantom limb, the real illusion is our normal limbs, as we just discussed, it’s just that the brain continues or insists on continuing with an illusion of body, even though that part of the body is no longer there. So the correct question is not what causes phantom limbs, but what causes normal body awareness, and I think this is the same in the context of time.
0:50:11.7 DB: So the question is not why time is distorted, but what is our true normal, what we call our normal sense of the passage of time. And we know that in some sense this is an illusion, and this… Maybe we can step slightly into the physics soon, and what I say by illusion is we perceive time to be flowing as a continuous linear narrative of what’s happening in the external world. That’s what it feels like, but we know that’s not the case at the conscious level. So of course, unconsciously, your brain is continuously processing information that’s coming in, but things only get created or sent to consciousness in sort of fits and bursts. So your brain, your unconscious brain, delivers into consciousness interpretations of what’s happening in the external world in sort of a highly edited, cut and pasted way, much like you’ll probably do to this podcast.
0:51:15.0 DB: And here’s an example of that. So again, I’m going to say two sentences. She had a stake in the company; she had a steak on the grill. Okay, so hopefully you understood the homophone stake there to have two very different meanings, but you could only der know the meaning by the words that came after. So I think for most people, they necessarily interpret consciously what stake means and then re-edit their interpretation of stake, they wait, their unconscious brain waits for the end of the sentence to provide a conscious interpretation of what stake made. I don’t know what your interpretation there…
0:52:01.2 SC: I immediately interpreted as food, but that might be because it’s lunch-time and I’m hungry, so that’s…
0:52:08.2 DB: Fair enough. It is a bit of a Rorschach test there, I guess, in terms of what you’re thinking, but in many cases, if we change to say… If I say chickpea or chicken, right? It’s not like you understood chick and then modified what chick means by what came after, so clearly, much of our interpretation of what’s happening with the external world is processed unconsciously and then waits, and then the unconscious brain waits until certain key points in that narrative to generate conscious experience. So in that sense it’s a solution, but at the deeper sense, the question is, is our subjective feeling of the flow of time, and the subjective feeling that the past is no longer real and that the future is not yet real, is that an illusion or not, and the degree to which that creates tension with physics.
0:53:08.8 DB: And I think you have one perspective, maybe you could tell me about what, if you view… The degree to which you view our conscious perception of the past no longer being real and the future not yet being real as a creation of the human mind or an illusion.
0:53:25.0 SC: Yeah, so I almost never use the word illusion in these contexts because I think that… I would much rather talk about higher level emergent approximate descriptions of reality that are not necessarily capturing the fundamental stuff, but are still capturing something real. But having said that, yeah, I think that our… So let me talk for just a couple of minutes, which I try not to do on the podcast, but that this is something I do think about. I think as a physicist that it makes most sense to be an eternalist in some sense about time, to attribute equal amounts of reality to the past, present and future.
0:54:00.0 SC: And then of course, the big question, which I completely agree is a big question, is, how is it that we perceive it so differently, that we seem to perceive the present moment as real and attribute such very different ontological status to the past and future and to no one’s surprise who has listened to me talk before or read my books, etcetera, it’s because of the arrow of time and increasing entropy, and I think that, aside from a few rhetorical flourishes, you and I are not so far apart in this. Maybe putting emphases here or there, I had Jenann Ismael on the podcast, I don’t know if you know Jenann’s work, she’s a philosopher, but she keeps up a lot with both the physics and the neuroscience side of things, and she gives this very… What is to me a compelling story of where our impression of flowing through time or traveling through time comes from in terms of exactly what you’ve already described, namely that our brains are constantly both remembering where we just came from and comparing it to where we think we are and predicting where we think we’re going to be and sort of updating that prediction in real time.
0:55:07.6 SC: So it’s exactly because the arrow of time gives us slightly unbalanced access to the immediate past and future and our abilities to predict or retrodict those that we have this feeling of flowing, of being pulled or pushed or whatever you want to call it, through time, even though at the level of fundamental physics, all the moments are equally real. That’s the spiel I would give.
0:55:32.0 DB: And you’re certainly wise in being cautious about the word illusion, ’cause particularly for physicists and neuroscientists, we use that word in very different ways. But in many ways, most of our subjective perceptions are mental constructs. The degree to which those mental contracts correlate with reality is I think what we’re ultimately having a conversation about. So in neuroscience, say, color, color is a mental construct, and color doesn’t really exist in physics. What you have is wavelength of electromagnetic radiation, and our perception of color is sort of evolution’s hack to create a spectrometer. So it’s an illusion, but most of the time, it’s fairly well-correlated with physical reality, unless we’re looking at gold and blue color dresses on Twitter.
0:56:33.6 DB: But for the most time, most of the cases, the evolutionary value, the adaptive value of perceiving color is that it detects something that’s a mental, it creates something that’s a mental construct, but it’s correlated with physical reality. So we have to remember that subjective experiences are evolutionary adaptive. So now the question is, is let’s not focus on the flow of time, let’s focus on the subjective feeling that the past is no longer real and that the future is… It’s related, but I think it’s an important distinction that will help this discussion.
0:57:11.9 DB: So we subjectively feel, very saliently, incredibly saliently, that the past is no longer real, our local past is no longer real, and the future is not yet real, and that’s fundamental to how the brain works, ’cause one of the brain’s jobs is to predict the future in order to allow for survival. So as I understand it, I think that you would have to classify that as… And I’m going to use the word illusion, although I know you don’t like that, but something misleading about the physical world that doesn’t reflect physical reality, because the past is in some sense, as you just put it, equally as real as the present and the future.
0:58:00.5 DB: And I know we have to be very careful with words here, because language is inherently presentist…
0:58:06.0 SC: Terrible, yeah.
0:58:08.8 DB: And it’s complicated. But we’re stuck, there you have it. So I struggle… Now, okay, so the way you posed it, I think you used the word interpretation of the laws of physics, but I struggle to know if you mean interpretation or if you really think it’s the laws of physics are conclusively telling us that the world or the universe is eternalist. So the way I see it, the laws of physics aren’t inherently presentist or eternalist, the laws of physics are somewhat agnostic to that, and we’re imposing interpretation on those laws, and that interpretation can be biased by the architecture of the brain. So what I think is worth considering is the degree to which our interpretation of the laws of physics, in this case the nature of time, is influenced by how our brains conceive of time.
0:59:10.8 SC: Well, I think that this is actually a good direction to go in because I actually want to go back to the illusion discussion a little bit, ’cause I think it’s relevant here, it’s not just a footnote, but first, let me parenthetically say that Steve Savage, who is a philosopher of science, who writes a lot about time, wrote… He emailed me an article because I was talking about presentism and eternalism in the usual way, and he said, don’t do that, there’s no difference between presentism and eternalism, they’re just different words to actually describe exactly the same ontology. And it was actually a very compelling believable article, but I emailed him back and I said, I bet all of your eternalist friends like this and none of your presentist friends agree with it.
0:59:53.4 SC: And he said, yeah, how did you know? Because basically, he was saying that presentists are just determinists in denial, or sorry, eternalists in denial. But…
1:00:02.3 DB: I would… I’m sorry, I would disagree with that too, with what he said, and I think as you know, the benchmark distinction between presentism and eternalism is time travel, closed time, like closed timelike curves. Is that the right term?
1:00:25.5 SC: That’s exactly right.
1:00:30.8 DB: So for presentists, that’s off the table right off the bat.
1:00:35.1 SC: How can you [1:00:37.2] ____ if it doesn’t exist, right?
1:00:40.4 DB: Right, right. And you’ve made the same point in one of your books. So I think there is a very relevant distinction between presentism and eternalism. I mean, relevant, not that it’s going to change our practical lives in any direct way, but I think at a deep question, it’s absolutely fundamental to know about the nature of reality. I mean, to me, it’s very important to know if the past is in some way as true as the present is the importance of the present somehow not… Is just subjectively relevant, not objectively relevant, so… Yeah, so sorry, I didn’t mean to interrupt you, but I think there is a very important difference.
1:01:18.9 SC: That’s very good. Let me follow up with one comment on that before I get into what I want to say, and the comment is that I think the point you’re making is that the difference between being a presentist and an eternalist might, if nothing else, be relevant for counterfactual claims, modal claims, as philosophers would say. Presentists are going to think that certain scenarios are just impossible to physically realize even if we haven’t done it yet, whereas an eternalist might imagine that maybe some day we would. So in that case, if that’s true, and I haven’t thought it through, that was a new idea you just threw at me, but yeah, then I would certainly grant that it makes a difference if you’re a presentist or an eternalist.
1:01:58.9 SC: Let me go back to the illusion thing, because I know you said you get why I don’t like to use the word illusion, but maybe it’s useful. No, I don’t think it is useful ever, and I think that the… I mean, an illusion is when you think something exists, but it literally doesn’t, it’s literally not there, someone made a coin disappear, but it really didn’t disappear, it just went into their pocket. That’s an illusion. You think that there is an oasis in the desert, and you go there and it’s not there, that’s an illusion. But color is not an illusion. I can build a machine that detects color perfectly well, color is just a useful emergent way of packaging microscopic details about the spectrum of wavelengths of light.
1:02:45.9 SC: When people say color is an illusion, I just don’t get that. I think that color is no more an illusion than tables and chairs are illusions, they’re not there in the fundamental laws of physics, but they’re there, and I think the same thing is true with the flow of time, it’s an absolutely useful way of thinking and conceptualizing about the world in our everyday macroscopic manifest image, even though it’s nowhere to be found in the fundamental laws.
1:03:12.7 DB: Fair enough. I think the… Perhaps the word mental construct would be more useful, and I don’t want to defend the use of the word illusion, but when… We need a word to describe some degree of decorrelation between our mental state and the physical state. And I don’t know what word you’d prefer, but that sometimes we can think of hallucinations in which you have no correlation with external world because you’re internally generating mental states that are totally non-existent in the external world. Now, when you tell about the color illusion, not the illusion of color, but a color illusion where you misperceive the color of an object, there’s a decorrelation.
1:04:00.2 SC: That’s right.
1:04:01.4 DB: But that decorrelation is allowed because color is a mental construct. Now, so I certainly agree with you regarding the caveats about the communication, but we should be aware… I mean, different fields use the words in different ways. Now, regarding how the architecture of our brain influences our interpretation of these laws, and I think I’d like to ask you a question, but first let me say that, as we already discussed, the brain is well-suited for certain types of computations, recognizing faces, but it’s quite horrible at mental numerical calculation. So the architecture of the brain is certainly well-designed for some things and not others.
1:04:48.4 DB: In the context of science, one of the best debugging devices we have, one of the best ways we have to overcome the brain’s limitations is called mathematics, because once some genius conjures up some set of equations that captures some truth about the physical world, any nitwit like myself can use those equations, plug those equations into a computer and predict something that’s going to happen in the world, and that act is independent of the human mind. Those equations quote speak for themself in a manner of speaking, but the minute we try to interpret those equations, then we have to acknowledge that they’re being filtered through our brains, flaws, gifts, and all.
1:05:45.4 DB: So the question I pose to you is presentism versus eternalism an interpretation of the laws of physics, or an aspect of the laws of physics. And I’ll put you on the spot here. If a… To give you a thought experiment, if some alien contacted you tonight and you were convinced this alien was an omniscient being compared to us, their technology was far superior to ours, and they told you, Sean, we actually live in a presentist universe, only the local present is real. Do you feel that the laws of physics would have to be rewritten or just reinterpreted?
1:06:34.8 SC: Yeah, no, I should have said that earlier. I think I did use the word interpretation and I’m pretty happy with that. I do think that for similar counterfactual reasons that we discussed, it is interesting and important to decide what we think is real, but in principle, I could imagine the current laws of physics… This is exactly what you’re trying to get me to admit, but I’m happy to admit it, are 100% compatible with a presentist ontology, where you think that the world is fundamentally three-dimensional existing at the current time and evolving through time in some way.
1:07:18.5 SC: The fact that there are laws of physics, especially if those laws are deterministic, and we don’t know for sure whether they are or not, but the fact that they are is what suggests to me that there’s not embedded in the laws of physics any special pointer indicating the present moment as picked out in any way, so I would really want to quiz the alien on what they meant by saying that they were presentists, but I would be open to having the conversation without necessarily overturning my current understanding of the laws of physics.
1:07:52.2 DB: So you just said that the laws of physics, and that’s of course true, whether it’s Newtonian physics, quantum mechanics or relativity, don’t have any special pointer that point to the present, but it’s also fair to say, and correct me if I’m wrong, they certainly don’t also say that there’s not a special pointer. So to me it seems like they’re agnostics as not that they say there’s not a special pointer, but they’re agnostic as to whether there’s a pointer or not.
1:08:20.5 SC: Well, as you know, the closest that there is to such an indication is from relativity, where you can’t uniquely slice the universe into the present moment, you have infinite freedom in deciding what you mean by the present once you go very far beyond your current environment.
1:08:37.9 DB: Yeah, but that’s just the concept of simultaneity at different points in space, and that’s a very in my mind unhelpful concept. One thing that I think the lesson from special relativity is in way that we should probably get rid of the word simultaneity…
1:09:00.6 SC: But if you think only the present moment is real, you have to define what you mean by the present moment, unless you have a fully local version of what you mean by that, which is a slightly more radical ontological shift.
1:09:16.2 DB: But what you were referring to in terms of the relativity of two points in time being arbitrary depending on how you slice spacetime, that’s comparing the present moment at different points in space, so what I’m saying that we both know is dangerous, but from defining at my own present, if I’m not worried about anybody else’s perspective, then defining my present now, you know, I think is just whatever my current state is by defining that by a clock or by the state of configuration of my atoms, I know it’s a bit arbitrary, and I realize it’s a problem, but…
1:10:04.2 SC: Well, no, so I think that is along the lines of what I was thinking of when I mentioned this very local version of what you mean by presentism. I mean, if naively, if I say only the present moment is real, someone’s going to say, okay, what is really happening now on Alpha Centauri. And there’s no answer to that.
1:10:23.9 DB: But the lesson is that it’s a bad question. It’s not that that’s an argument towards eternalism.
1:10:29.1 SC: Do you, by the way, get into the distinction between Humeanism and anti-Humeanism about the laws of nature? Do you know this distinction? I talked about it on the podcast with Dan Hall, and roughly for the audience who for some crazy reason hasn’t listen to that episode yet, Humeans believe that the world is just a collection of things that happens and we sort of encapsulate them into what we call laws of nature, the patterns that we see in those things, whereas anti-Humeans and think that the laws of nature have some oopmh to them, they actually sort of generate one moment from another in a more creative becoming way, so it’s a being versus becoming kind of perspective. And I wonder if that maps on to your feelings about presentism and anti-Humeanism versus eternalism and Humeanism.
1:11:19.3 DB: Yeah, I’m certainly not that familiar with that dichotomy, but I do… My perspective is that, of course, everything that happens in the brain is a direct, directly governed or directly constrained by the laws of physics, of course. Now, the brain evolved to survive in a world governed by the laws of physics, which is at least… Of course, our intuitions about quantum mechanics are way off, they’re horrible, our intuitions about cosmology are way off, they’re horrible, but our intuitions about the mesoscopic Newtonian physics are actually pretty good, and if they weren’t, we’d be dead, because it’s very important to have those intuitions.
1:12:00.5 DB: So in that sense, I think the brain has to be governed in one way, or at least tuned or calibrated to grasp the laws of physics, so I think Hume talks about cause and effect, contiguity, and certain fundamental laws allow the brain… Or I don’t know if we should call them laws, but I think he refers to them as laws of contiguity or associativity.
1:12:25.6 SC: Metaphysical principles of some sort.
1:12:27.3 DB: Yeah, so I think the brain is built in to… Is built to capture those laws, and again, we were just talking about simultaneity, and I was saying that I think in many ways, the lesson from special relativity is that the concept of simultaneity across different points in space is very unhelpful, but the brain evolved pay a lot of attention to simultaneity, we’re sort of addicted to simultaneity, because…
1:12:53.9 SC: Synchronize our watches.
1:12:56.3 DB: The brain was trained, and I think this is built in to the brain, if I see you moving your lips and hear your voice, my brain concludes that you are generating that sound, but if you’re a ventriloquist and you have a ventriloquist’s dummy on your lap and I hear sound and I see the lips of the ventriloquist move, of the dummy move, then I conclude the sound is coming from the dummy, and it’s a very compelling illusion. So your brain is continuously using simultaneity between modalities, between events to make sense of the world. So I think it does confuse our interpretation of the physics in some ways.
1:13:36.9 SC: Now, I think that’s perfectly fair and let me run something by you, ’cause I have this sort of half-baked idea that I don’t think I’ve ever run by a real neuroscientist, about exactly this question, how well can the human mind in principle do understanding the laws of physics given its manifold lack of capacity. And so my half-baked idea is that at some point, I don’t know when, but roughly speaking, there was a phase transition in which human minds became essentially Turing machines, right? Yes, we have an ability to add big numbers, but we know what it means to add a number, and there’s no in principle maximum number that we can add or anything like that. We have a manifest image of the world with folk physics where we think things are solid and they exist in three-dimensional space, but by doing experiments, by coming up with hypotheses, we can invent relativity and quantum mechanics and all these things.
1:14:34.0 SC: So I do think that obviously, there’s a lot of bugs, as in the word in one of the titles of one of your books, in our human brains that prevent us from being perfectly reasoning machines, but I do think we’ve passed some threshold where there’s no reason to think that the ultimate laws of physics are beyond us.
1:14:54.3 DB: Yeah, that’s a deep question. I struggle, I don’t struggle as much as you’d probably do with that, but I struggle with the self-centered question of consciousness. Can the brain understand consciousness? So you’re asking is… Listen, at the end of the day, we’re unusually smart apes, right? And I don’t think we have any real right to believe that we have some right to think that we can understand the universe or it can understand ourselves. That doesn’t mean we can’t, but I think a bit of humility here is appropriate. So I think that for the laws of physics, I don’t… Here’s a way I would put it, Sean, and neuroscience is struggling with this issue. Neuroscientists agree that we want to understand the brain. What neuroscientists don’t agree on is what it means to understand the brain, so the question I pose to you is, you said, can we understand physics?
1:15:58.5 DB: I don’t think physicists would necessarily agree what understanding means, so does having the equations comprise understanding, and ultimately that’s probably the best understanding there is, or does it have understanding whether a photon is a particle or a wave or not, what does understanding mean? And I think that’s what we’re really struggling… Let me give you another example. We started off saying that we might be at the phase and then traction between neuroscience and machine learning, where neuroscientists should be learning from the machine learning field. So in machine learning, they create these incredibly sophisticated computational devices based on neural networks that, say, kick our ass in chess, so Alpha Zero just destroys humans in the game of chess.
1:16:52.2 DB: Do we understand how Alpha Zero achieves that? Okay, so it’s a tricky question, because we know every single quote synapse of that program, all the weights, all the activity in the algorithm it generated, but in a way, we don’t understand how it generated that move or why it played that move versus the other, or if it made a move that happened to be wrong, how we would correct it. So in one sense we don’t understand it, but in another sense, we know the equations that created it, and ultimately I think it was Feynman that said, you know, I can understand something if I can make it. And I think that’s a very pragmatic use of the word understanding, so my question to turn around to you, which was if the human brain is able to understand the laws of physics, what does that even mean, Sean? What does understanding the laws of physics mean?
1:17:46.0 SC: Well, I think that’s an excellent question and in part, sorry, it’s both an excellent question and a useless question…
1:17:56.0 DB: Fair enough.
1:17:56.1 SC: Because… And the useless aspect is that part of the answer is going to be carefully parsing different possible construals of the word understanding, ’cause we use that one word to mean different things. But I do think there’s also a good question lurking in there, because there are different aspects and knowing what they all are, it’s not like there’s some right aspect of understanding and some wrong one, but there are different ones, and we can ask when we achieve different levels of understanding. If we had perfect equations that governed the fundamental stuff of reality, so if we had some ontology, some stuff that we said, or at least some representation mathematically of what reality is, and some equations which we thought were 100% complete and accurate in describing how it behaved, that is a kind of understanding, which we are nowhere near yet, but would certainly count as a kind of flavor of understanding. And like you say, it would leave us completely unable to do a lot of things that we would want to do.
1:18:55.3 SC: That’s like having the understanding at the neuron level of AlphaGo or something like that. We would also like, as you say, to have some higher level understanding, and maybe the best way to contemplate that is in, once again, this counterfactual sense, so you know what all the weights in the neural network are in AlphaGo, but if you changed one, what would happen? What would be the difference, right? That’s incredibly hard to understand. You basically have to start from scratch and just do the whole thing again. So that sort of higher level, seeing the emergent patterns understanding would be lacking in that, and that’s a perfectly good kind of understanding, also I’d be in favor of that, and there’s probably other layers of understanding that are relevant.
1:19:42.6 SC: This is part of why I’m an eternalist, and I don’t think that it conflicts with my personal feeling that time flows and I can make choices that affect the future and affect the past, and it’s also a reason why I’m a compatibilist about free will. I don’t know, what is your free will stance? We should get on the table here for the audience.
1:20:04.2 DB: So I think the first lesson to be taken from free will is really how much we can debate something without properly defining what we mean, right, that’s the amazing thing about free will. And in the context of free will, I’m reminded of the Zen koan, if a tree falls in a forest and no one’s there to hear it, does it make a sound? Right, and it sounds deep, but if you say, the moment you define sound, it becomes silly, if I define sound as the vibration of air molecules, well, of course it doesn’t make a sound. If I define sound as those air molecules hitting my tympanic membrane and creating a conscious perception, of course it doesn’t make a sound.
1:20:47.5 DB: So in free will, my view is that if I use my free will to define free will as the decisions my brain makes, conscious and unconscious processes, then of course I have free will, I am my brain, my brain is me. So the decisions my brain makes at the conscious and conscious level are, is my free will. Now, of course, if I define free will or if we define free will as what I believe you’ve referred to as libertarian free will, and which is somehow above the laws of physics, then I think it’s our scientific obligation to say, no, there is no free will, because we’re scientists and things are… Everything should obey the laws of physics, so we have to say, no, there’s no free will.
1:21:39.0 DB: So I think when we’re talking about the concept, it’s a bit of what some people have called cryptodualism, the question of itself is a bias the brain has, it’s like the brain evolved to make a mind that didn’t believe it was caused by the brain, so the mere question of free will, I think, is suspect, because it somehow implies that people are cryptodualists.
1:22:09.5 SC: Right, no, I think that we’re more or less on the same train here, so that’s good, but this is what… Again, I’ll finish the thought about why this makes me completely happy with being an eternalist, even though my immediate personal perception of time is so different. I think that there is a fundamental layer to reality, I don’t know what it is, I have my ideas, other people have their ideas. But there’s also numerous higher level emergent approximate descriptions of that, and what’s crucial is that those emergent higher level descriptions capture something real, and if for no other reason than purely practical ones, right, I can know when I see a table, if it’s in someone’s house that is not trying to trick me or something like that, it’s not like the Amazing Randi’s house, so what I’m walking into a pit of vipers, but he’s trying to trick me. I know that if I put my coffee cup on the table, it will not fall through, it’ll be supported by that.
1:23:05.2 SC: Now, that does not necessarily rely on the fact that I know that the table is really made of atoms and those atoms really have fermions in them, that obey the Pauli exclusion principle, and the whole thing is described by a quantum field theory, obeying the laws of physics, like all of that stuff, who care? Whether I knew that or not, it is not an illusion for me to say, oh, there’s a solid object here and I can put the coffee cup on it, because that captures something predictable and real, right. And I think that likewise, my perception that I exist in the present moment, I make choices about what I’m going to do in the future, all of that is very useful and practical and real. None of it is fundamental, but that’s okay with me, it’s not an illusion either.
1:23:49.0 DB: And as you said, I think we’re in large agreement about that, I just feel that ultimately the question of the reality of nature, and if the past, present and future are equally real, is something that we should strive to understand about the nature of the universe. And I think that, as you said, your interpretation leans towards or is eternalist, but there’s no empirical data one way or the other at this stage. Indeed, it’s not clear what any empirical data would look like. So I think if we try to be more objective… Listen, we accept a lot of strange counterintuitive ideas in physics, we accept entanglement, that clocks slow down at high gravitation potentials and so forth, we accept those because there’s a ridiculous amount of experimental data to support those.
1:24:55.9 DB: That’s obviously not the case with eternalism. Yeah, it’s a strange idea intuitively for most people, that somehow the present still persists or the future is quote out there in some way, so in the absence of any empirical data for that, it seems to me we should take as a data point in the lack of any other data points, trust our perception. So I would just like to put that one data point in there, and I’m willing to throw it away if we get better data points, Sean, but it just seems that that should be the tie-breaker for now, but I know, for what it’s worth.
1:25:35.9 SC: I think that’s perfectly fair, and I think as a final thought that will encourage listeners in the audience to try to go out there and construct some closed timelike curves and therefore prove you wrong. I think that would be a wonderful little project for the Mindscape audience.
1:25:52.8 DB: That would definitely help settle this debate, and so keep me posted if any of your listeners do that, Sean.
1:25:58.5 SC: We will not keep it secret, I absolutely promise you. So Dean Buonomano, thank you very much for being on the Mindscape podcast.
1:26:04.8 DB: Thank you so much for having a stimulating conversation, Sean.
[music] [/accordion-item][/accordion]
Wow – this is my favorite podcast to date! If the future is already there, I am willing to bet is was always my favorite. Wow. Rob
Pingback: Sean Carroll's Mindscape Podcast: Dean Buonomano on Time, Reality, and the Brain | 3 Quarks Daily
I agree, this episode was fantastic. Thank you Sean and Dean for sharing this conversation.
Thank you for this amazing episode, and all the others too, in fact. Very inspiring. I have listened to all of the episodes (well, there may be one or two missed ones) and I often formulate comments and questions in my mind but never actually ask them.
This time, though, there is one question that I can ask without writing an essay about it (looking back after writing it, though, I feel I might have been too optimistic). Dean Buonomano says towards the end, within the conversation about free will, that it is as if the brain evolved to make a mind, that didn’t believe that it was caused by the brain, and this expresses the bias of the question (the brain asking if it has free will already assuming it has free will to decide whether it has free will). This part to me expresses a blind spot, a kind of recurring doxa if you will, of the current philosophical and scientific conversation about the brain and the mind. The doxa is not about free will but that the brain creates the mind and it is not useful to take into account that we can only access the idea of the brain through the mind. So to me it seems that Buonomano’s phrase should be turned around like so: It is as if the mind evolved to imagine a brain, and then imagine that it was created by the brain. Moreover, it forgets about its own existence soon after, and thinks that only the brain is real!
The argument for this is simple and basically already explained above, but let me try to elaborate: we can only access the idea and the perception of the brain through our first person phenomenal consciousness, here dubbed “the mind”. I guess it could well be so that what we imagine (perceive, construct) as the brain does create (though let us not forget about the body and the environment etc) the thing that imagines it, but the fact remains that the hierarchy is not that we know definitively that first there is brain, then mind. What we know definitively is that there is a mind that thinks that there is a thing called brain. Makes sense?
Let me just finish this by saying that I don’t try to split hairs or be smart about this. This is a genuinely unsolved problem, and if it is solved, I would be more than happy if someone would explain it to me in the comments. Panpsychists like Philip Goff — btw loved that episode although I felt like in that episode the usually wise Sean was unwilling to accept precisely this idea that there is also consciousness, or mind, or first person phenomenal experience, which needs to be taken into account and not just assumed that it will be explained away at some point — try to tackle this question, framed as the question of consciousness as fundamental to the explanation of reality, but it has been an uphill for them too to even get the question taken seriously.
So after a long comment a question: how should we deal with the problem of the mind here? How can we say that the brain creates the mind, when it is also true that the mind creates (imagines, perceives, constructs) the brain? I don’t think this is much different to saying that the brain creates the phenomenal world that we experience, except we substitute the brain with the mind, and add the brain into the phenomenal world that is created by the mind. A sleight of hand? Let me know.
I hope I haven’t made a fool of myself here. In any case, thank you for this whole podcast series! It is a joy to listen to.
Física e neurociencia!
Os 10-15 minutos antes do final, foi um presente incrível. O “choque” de ideias de ambos, argumentos de “defesa” foram “Bum”.
Ótimo, quando argumentos inteligentes – de ambas as partes- divergem, e, discutidos!
Obrigada Sean Carroll
Obrigada Dean Buonomano
Presentists, eternalists, we’re the same until we get a bit closer to find out what constitutes reality 😊
Now I have two absolutely favorite Mindscape podcasts:
This one and number 44.
From Mindscape 140 about 9 minutes in.
0:09:03.1 DB: … Sean, we already have pretty good interfaces, right, they’re called eyes and ears, and they’re not bad interfaces and we use them quite well…
Thinking of Damasio speaking of the GUT in Mindscape 44 could we say eyes, ears, and GUT, are not bad interfaces?
I ask this because since Damasio I have been learning to use my gut as an interface and this has been very important to me.
Not sure Dean has a yard- otherwise he would recognize that those darned squirrels ‘plan’ far ahead in time by planting acorns all over the place. Of course their memory of where they planted the acorns may not be so great…
A certain percentage of men are color blind. Women can see color generally better than men and some women much better. My wife will show me two fabrics and claim that they are different and I can’t see a difference. In space, color perception shifts. Maybe illusion is not a good word but color perception is certainly subjective.