Decoding 'the Most Complex Object in the Universe'
IRA FLATOW, HOST:
This is SCIENCE FRIDAY. I'm Ira Flatow. Your brain has nearly 100 billion neurons, and one of my next guests compares that complexity to the Amazon rainforest. In fact, he says there about as many trees in the Amazon as there are neurons in your brain. Think about what the Amazon looks like for a second.
And the roots and the branches and the leaves and the vines, all of that can be compared to the tangled network formed between your brain cells because many of your neurons are in fact wired to tens of thousands of other neurons. That incredible complex network is packed into a soft, three pound organ inside your head, making it, as my next guest says, the most complicated object in the known universe.
Is it possible we'll figure out the secrets of what's inside our heads? I mean, if we are inside that black box, can we understand the whole black box? Like can you see the forest for the trees? Is consciousness beyond our understanding?
My next guests are both decoding the brain's inner workings by observing how we learn and how we speak, by eavesdropping on how individual neurons, the neurons, the cells, speak to one another. Christof Koch is chief scientific officer at the Allen Institute for Brain Science here in Seattle. He's also the author of "Consciousness: Confessions of a Romantic Reductionist." Sounds very romantic title. Welcome to SCIENCE FRIDAY, Dr. Koch.
CHRISTOF KOCH: Good morning, Ira.
FLATOW: Good morning. Patricia Kuhl is director of NSF Science of Learning Center. She's also a professor and co-director of the Institute for Learning and Brain Sciences at the University of Washington here in Seattle. Welcome to SCIENCE FRIDAY.
PATRICIA KUHL: Thank you.
FLATOW: Is it possible - let me begin with you, Dr. Koch. Do you think that consciousness might be possible to figure out? I mean, how can we get inside that black box? If we're in the black box, can we see what it's made out of?
KOCH: So for the past 3,000 years that we've been trying to understand the mind-body problem, mainly by philosophy, mainly by talking about it, we've not been very successful. Philosophers are very good at asking questions, but historically they're pretty bad at answering them in any decisive way.
KOCH: But fortunately, since 150 years, we have clinical science, we have neurology, and we can study what happens if you have a hole in your head and if you have - depending where the damage is, you might have a specific loss of consciousness. So we know, unlike the Greeks, we now know consciousness rises from the brain, and we know if you lack particular parts of the brain, you will lack specific content of consciousness.
And over the last 50 years, particularly in animals but also in humans, we've been able to look at actually the stuff out of which the brain is made, those many tangled neurons that Ira was just mentioning. There's an enormous number of them, and just like as in the rainforest, there's an enormous diversity of them, and that's the complexity we're facing just like in astrophysics over the last 100 years, every new - each new generation of astronomers and astrophysicists discovers that the universe is yet more bigger - yet bigger than we previously thought.
Now people talk about multiverses. Each time we look with better and better tools, with better with better microscopes and other tools, we see more and more complexity. So now we realize there are not just two types of nerve cells, but there are a probably 1,000 different types of nerve cells, just like there are 1,000 different species of trees in the rainforest.
And so the challenge is try to unravel it(ph) . And we have - the atoms of consciousness are really, are the nerve cells. And so really it's essential for us if we want to understand the mind-body problem, in health and in disease, in schizophrenia and other pathologies of consciousness, we really need to understand it at this neuronal level.
FLATOW: And are you actually able to go in to the neuronal level and understand that? Is it useful to study individual neurons? Could that tell us more about consciousness?
KOCH: It's essential. The only way we're going to find out about, totally uncover the underpinnings of consciousness, how - the brain is a physical system like any other physical system in the universe. But it exudes this magic thing, these feelings. I wake up each morning, I open my eyes, I have pains and pleasure. I remember who I am. How does that arise out of the physics of the brain?
And the only way we can study that is by actually looking at the stuff that computes this stuff that exudes mind. And the stuff that exudes mind is brain. It's not just some oozy, fluffy, tofu-like substance or overcooked cauliflower. It actually consists out of roughly 100 billion nerve cells, and they interact in a very complex way.
And we know from certain mental diseases, you know, I mentioned schizophrenia, there's also Alzheimer's and Parkinson's, it all involves various, various complicated mis-wiring. And so in order to help people ultimately, we need to understand the wiring and the mis-wiring, and that can primarily be done at the level of individual nerve cells.
FLATOW: Dr. Kuhl, let's talk about how much of your brain, how much of whom you are now, your consciousness, were you born with and how much did you pick over the years, I mean the nature-nurture question, right?
KUHL: Right, so we're interested in taking the approach of studying consciousness and all the magical things that humans can do by studying the baby, starting with baby brains and young children's brains and looking at the explosion of learning that takes place early.
And the tools of modern neural science now allow us to look and see activity. We can take snapshots of the brain and see individual areas, 433 of them, and the white and gray matter is their form. You can look at the fiber tracks that are connecting areas. You can look at activity using magnetoencephalography. That's happening online as baby hears a word, sees their mother or smells something that they recognize as milk.
FLATOW: Right. Your brain is made up of all these neurons, and yet we have all this DNA that we're born with. There's got to be, I imagine, somewhere along the line where the DNA gets triggered to mature or create something that helps your brain mature and develop.
KUHL: Exactly, exactly.
FLATOW: Is that where the nature part comes in on it?
KUHL: So think of the baby as growing up in an ecosystem, and part of the ecosystem is what the baby brings to it. There's a brain, and that brain's got windows of opportunity that are being opened by genes that are expressing. And also those windows of opportunity narrow. So you've got the baby trajectoring along that developmental path with a set of biological markers that are set to go off at certain times.
Then the ecosystem is the rest of us, the social environment. It's the people, the things, the schools, whether the child's growing up in poverty, that will provide or not provide what that critical window of opportunity is supposed to be receiving.
And so the baby's not on a turnkey system entirely. The baby's waiting for all of us to provide the necessary stuff.
FLATOW: Well, you mention a critical window of opportunity, and we all know how kids learn things like sponges, right? They suck them up much faster than we can. Can we find out how that works and extend that into later life?
KUHL: Well, there's a lot of things that we're doing to try to understand that. So when windows open for learning and close for learning, there are triggers. We're trying to understand what the triggers are. We're trying to understand whether you can move them so that they open later or close later. And some of the magic that kids are putting to work has to do with how they interact with the social world.
So in learning of language, which we know we're geniuses at between zero and seven but not so good every two years after the age of seven, you're falling off the curve with regard to your ability to learn a new language. After puberty, it really gets difficult.
And so what we're looking at is, well, what's the baby brain doing at, you know, age two or age three that you and I can't do anymore. And what the baby brain's doing is calculating, taking statistics as they listen to us talk and interact in the environment and also reacting to us socially.
The social brain has a good bit to do in controlling how babies react to the statistics of their world.
FLATOW: Could it also be what the baby brain is not doing, which is our taxes, our school, balancing our budget, those sort of - the noise?
KUHL: Right. Well, the other part of the baby brain is to actually be able to look. We can now look with diffusion tensor imaging at the superhighways in the brain. You can see how language areas and executive control areas are related to the reward system. You know, that squirt of dopamine that I hope you're all getting now as you listen to me talk, and you're imaging your little child and your little grandchild, that squirt of dopamine in the reward system has a lot to do with how it is that we learn at different ages.
So understanding what drives human learning at different ages, and it's different over your - over the lifespan, will help us, help all children because that critical first five years, when typical kids are cruising along, but kids with developmental disabilities are not, is very, very important and also for all of us who want to keep our brains alive all our lives. Understanding how it works is extremely important.
FLATOW: Dr. Koch, do we have a definition, does everybody agree on a definition of what consciousness is?
KOCH: No, nobody agrees.
KOCH: There's no universal agreement. But there's widespread - sort of it relates to the feeling, it feels like something. It feels like something to have pain. It feels like something to be interviewed by you. It feels like something to be a man. Or those are all different...
FLATOW: How about a dog? I mean, a dog must have - feels like something, right?
KOCH: There's no question that dogs, in fact all mammals, have conscious states, right? There's no question your dog can feel happy, he can feel sad. You know, people do, babies do, patients do. What we don't know how far it goes outside mammals. Is it possible that - some people are certain, I believe that most, if not all, multicellular creatures feel something that - even a bee.
A bee is a very complicated system. The brain of a bee is 10 times more heavily wired than we. Of course it has many fewer neurons, roughly a million neurons, but it may well feel like something to be a bee. And so the question is: How does this feeling, how does this subjective feeling come into the world? That's always been the mystery. It's always been the central mystery about consciousness.
FLATOW: Now, Eric Kandel spent his whole lifetime looking at cells of a sea slug and won the Nobel Prize eventually for it. Can you bring that to a higher level? Because he was probing an individual cell almost like you're talking about probing individual cells in the brain. Is that - can you get - can you learn as much as he might have from - or learn more about how consciousness is from those individual cells?
KOCH: Yes. So, ultimately, what we're doing - and President Obama is launching a new initiative to try to speed this up - is to record, to listen in. So you can put a piece of wire into a human brain during neurosurgery, or you can do it in animal brains, and you can listen to the way neurons chat to each other. They have this sort of hash...
(SOUNDBITE OF MAKING SOUNDS)
KOCH: It works very well if you're German.
KOCH: I mean, the brain itself doesn't sound like anything (unintelligible).
FLATOW: We actually have a sound of the brain. We're going to play - we have - give us a minute. We actually have - I think is probably some audio that you...
FLATOW: ...gave to us. Let's listen to that now.
(SOUNDBITE OF CRACKLING)
FLATOW: Tell us what we're listening to.
KOCH: So here, we're listening to sound track that's made by putting a piece of wire into, in this case, a human brain during neurosurgery and amplifying the electrical signal. And each of those little clicks is actually a little pulse that's being sent by one neuron to roughly - we don't know - anywhere between 5,000 and 10,000 other neurons. And so you have to imagine, as we talk right now, you know, there are literally billions of neurons that chatter away in this code either - and they're sending them out, and they're receiving it. And somehow, out of this cacophony arises, you know, this stable perception arises the fact that I can - I have a voice in my head, and it sounds like you. And so that's...
FLATOW: Sorry for that.
FLATOW: Yeah. And you think that you can actually learn something - was that - just - is it being too simplistic to learn the macro size of something from the individual pulses in the brain?
KOCH: Well, by and large, science has been...
FLATOW: Too reductionist here, I guess is what I'm asking.
KOCH: Well, but science has been very successful over the last 200 years at exactly taking this...
KOCH: ...ever more reductionist point of view. But the good question you ask is that maybe consciousness requires more than just a reductionist approach. And so there are some theories that try to address that, that, in fact, argue that the part of the brain that's not reducible, in some sense, that's what gives rise to consciousness. So we don't know. But it has been so successful, so that's, of course, what we should do. In the meantime, we're discovering all sorts of other interesting facts about the brain, but there's really no good alternative to doing it this way.
FLATOW: Dr. Kuhl, you - one area you do a lot of work on is in language. Tell us what you're trying to find out about language.
KUHL: Language is one of the quintessential human abilities, and we're trying to understand how it comes together in the human brain. As Christof says, we're - we know that the activity is happening at the neuronal level, but it's when millions of neurons are coordinated to do something like listen to a word and recognize it, to have an idea in your head requires this - that you will compute all the statistics and create an idea. We're trying to understand how that happens in the baby brain, and how being bathed in one language, as opposed to another, or being bathed in two at the same time - as in bilingual children - how does that change the whole brain? And, in fact, it does.
What we're seeing is that areas of the brain are different. The connectivity is different. The superhighways are constructed differently in bilingual and monolingual brains. And, in fact, bilingual brains are more creative. They're more - they're not smarter. It's not your general IQ. But it is your ability to be a flexible thinker. When solving a problem, if you're bilingual - whether you're a baby or whether you're an adult - you're going to be better at solving a novel problem.
So we have all these new imaging techniques, not looking at individual neurons, but when thousands of neurons fire in coordination and change current flow, we have babies and adults sitting in something that looks like a hairdryer from Mars, but it's completely safe and completely noninvasive and absolutely silent. So we can play things for themselves, show them pictures, show them movies and watch their brains as they are processing real words, and link that to their structures.
We have like a - with MRI and DTI, you have a static photograph of that individual child's brain, the connections in that brain, and how this activity fires in the various areas. So that's brand new and very, very exciting.
FLATOW: You're listening to SCIENCE FRIDAY, from NPR. I'm talking with Christopher Koch and Patricia Kuhl. We're talking about brain, listening and learning from the brain. Is there a tool - I'm going to give you the same question I gave to Christopher - to Christof. If I could give you a blank check and you could invent a machine to study the brain, what would it be, and what would you like to see?
KUHL: Well, we want to see the next generation of something like magnetoencephalography. I mean, it is picking up a current flow, and it's able to see with millimeter accuracy, but we want to get even smaller than that - a millisecond accuracy where the activity is in the brain, what's the cortical dynamics that's going on as kids learn. And how does the structure that was built yesterday operate today as the kid looks forward in their world? So we want better and more efficient machines.
We want things that will allow us to see with absolute clarity how one brain differs from another. And why is that important? Well, we just showed in a study published May 29th that if you take a simple brain measure on a child with autism at the age of two, you can predict better than any other measure that we have how that children's - that child's outcomes will look at age four and age six, not only in language. This is a language probe at the age of two, how the brain responds to known words, but how that child behaves throughout the language arena and cognition and adaptive behavior, and it gets stronger over time.
So these brain measures are not only going to reveal the wonder of the human brain for all of us over the age - over our ages, but for children who have disabilities, we can get in there early with biomarkers to say where in the brain - one, this child's at risk, but, two, where in the brain, what structures, what superhighways are missing? What might we do to create better interventions that don't miss that window of opportunity? We've got to get in there really early when the brain is plastic and change the course of that child's life and that child's family's life.
FLATOW: Mm-hmm. Let me get a quick question from the audience. We've got about a minute. Yes, sir.
JORDAN BROWN: OK. My name is Jordan Brown. And I was wondering - you've been talking about processing information from outside, like language and different things, but where does that go? What is the physical thing that constitutes thought? And could you find that by just tracking the electrical signals from the outside inputs?
FLATOW: That's such a good question.
KUHL: Good question.
KUHL: Would you like a job?
FLATOW: And I'm going - he asked - yes - signed him up, as we used to say in baseball. I'm going to hold that for the break, because we have a break coming up in about 30 seconds. And we'll have - well, that's a meaty question to tackle in 30 seconds.
FLATOW: And, yeah, we're thinking he wants a job. We'll come back and talk more with Christof Koch, chief scientific officer of the Allen Institute for Brain Science here in Seattle, and Patricia Kuhl, director of the NSF Science and Learning Center, also professor and co-director of the Institute for Learning and Brain Sciences at the University of Washington here in Seattle. We're going to take, come back after this break and step up to the mike with your questions. This is a good time for them. We'll be right back after this break. Stay with us.
(SOUNDBITE OF MUSIC)
FLATOW: I'm Ira Flatow. This is SCIENCE FRIDAY, from NPR.
(SOUNDBITE OF MUSIC)
FLATOW: This is SCIENCE FRIDAY. I'm Ira Flatow.
We're talking this hour about the brain, how a complicated tangle of neurons can give rise to consciousness. Talking with my guests, Christof Koch, chief scientific officer of the Allen Institute for Brain Science here in Seattle, Patricia Kuhl, director of the NSF Science of Learning Center, also professor and co-director of the Institute for Learning and Brain Sciences at the University of Washington here in Seattle.
And before the break, a very good question came in from a young member of our audience, wanting to know, OK, so what happens to all the stuff you - that flows into your brain, all the information and stuff? What happens to it? Christof, you want to tackle that? Yeah?
KOCH: Jordan, so I think it's a very good question. So the brain is, in some sense, similar to a computer. In both cases, you have information that flows in, in one case through transistors, in the other case through nerve cells in our eyes. For example, if I look at you. And then as you said, we have to track - just like in a computer, we can track so that the electric activity in the various transistors, we can do the same in the brain. We can track the electrical activity, either using the methods that Patricia Kuhl alluded to from the outside, using EG or EMG, or we can put microelectrodes in it or other tools, and we can see how the electric activity moves through the brain and activates language area or activates a part of the brain that's responsible for seeing, or another part of the brain that's responsible for hearing. And then ultimately, it activates part of the brain that's responsible for motion, and then we move our eyes, or we speak.
FLATOW: You talk - you've brought in some references to computer, transistors and things like that. In your book, you compare the brain with its hundred billion neurons to the Internet, which has several billion computers there. Could consciousness arise from the Internet?
FLATOW: We know what - we know what's out there now. But could something meaningful like consciousness arise from the Internet?
KOCH: In principle, yes. In practice, we have no idea, and I don't think it's a useful research topic right now. But in principle...
FLATOW: Why not?
KOCH: Well, I mean, let's first look at the analogy. The brain is a very complicated network that consists of individual elements called nerve cells. OK? Now, of course, it doesn't make any sense to say one nerve cell conscious. But a large collection of neurons - not all of them. My spinal cord isn't conscious. My second brain in the gut isn't consciousness. But some part of the neurons generate consciousness.
So now I look at another network, very, very complex, the Internet. It has 10 billion nodes. Each node consists of, you know, between two and 10 billion transistors. So maybe it's also possible that collective, as a whole, it also feels like something to be the Internet. And if you would turn it off, it wouldn't feel anything anymore.
But right now, we don't have a good way to actually test this. Right now, that's purely science fiction. So you're not going to get a grant funded to study whether the Internet is conscious or not.
FLATOW: But you think it might be, you said, someday.
KOCH: Yeah, in...
FLATOW: In principle.
KOCH: ...in principle (unintelligible) there's no reason. Of course, we see it in, you know, "Matrix" or other movies all the time, that, you know, machines become sentient. In principle, nothing in the universe - there's no physical law that says that that is not possible.
FLATOW: All right. Let me get - time for maybe one more question up here. Yes.
DANTE MANTEL: Yes. Dante Mantel(ph), of Northwest University in Kirkland. You kind of answered most of my question. In science fiction, you just have a big enough computer, and it might become conscious. So - but expanding on that, do you use computer simulations, and is it helpful for studying the brain?
KOCH: Yeah, it's essential. So we - for example, right now, at the Allen Institute, we have this very large project where we're recording from brains, human and animal brains, but we also are simulating it. Because even if you could record from - even if you could visualize the activity of every single nerve cell in something I call a brain TV - imagine a television with something like 200,000 pixel by 200,000 pixel. And every pixel is the activity of one single nerve cell, and sometimes it turns white when it fires one of these pulses, and sometimes it usually goes black.
And now I'm looking at this picture of 100 billion pixel, right? I probably wouldn't understand any of it. So I need a theory. I need a model. I need a theory to say, OK, this is language and this is vision. And it's vision for the following reason, and it's language for the other reason. And this happens, you know, if you look at the brain TV of a schizophrenic patient because it's hardly different, and that's why the person hears a voice. And so it's essential to theories, and then it's actually essential to the test in using computers.
FLATOW: Mm-hmm. Is it possible to make a computer where the brain picture that big, with all the neurons firing?
KOCH: Not right now, but there's a group, you may have heard. There's a very large project. It was just funded to the tune of 1 billion euro in the European Union called the Human Brain Project, where they're exactly trying to do that. Over the next 10 years, they're to build a special-purpose supercomputer, because you need - you know, it's terabytes of data. So they're trying to build a special-purpose computer to simulate the human brain.
FLATOW: Wow. It's quite interesting and quite...
KOCH: It's pretty cool stuff.
FLATOW: It's pretty cool stuff. And we want to thank Doctor Kuhl for coming with - to be with us today. Patrick Kuhl - Patricia Kuhl is director of the NSF Science of Learning Center. She's also a professor and co-director of the Institute for Learning and Brain Sciences at the University of Washington in Seattle. And also Christof Koch is chief scientific officer at the Allen Institute for Brain Sciences here in Seattle. Thank you both for taking time to be with us today.
KOCH: Thank you very much.
KUHL: Thank you. Transcript provided by NPR, Copyright NPR.