WEBVTT 00:00:03.677 --> 00:00:06.557 This is the Convergent Science Network podcast. 00:00:08.577 --> 00:00:13.057 Leading researchers in the domain of neuroscience, brain theory and technology 00:00:13.057 --> 00:00:16.757 are interviewed by Paul Vershoor and Tony Prescott. 00:00:19.417 --> 00:00:25.177 This is Paul Vershoor with our summer school, again, together with Tony Prescott, 00:00:26.077 --> 00:00:29.317 co-chair of summer school, and John Kass, one of our speakers. 00:00:30.117 --> 00:00:36.157 And John gave a talk about the evolution of the brain. 00:00:37.197 --> 00:00:42.597 And you showed this incredible overview and insight in how brains have evolved. 00:00:43.217 --> 00:00:51.237 So why is evolution, this evolutionary perspective on the brain so essential to you? 00:00:52.857 --> 00:01:00.017 Well, part of it I went in my talk, and that is that we had a curiosity of how 00:01:00.017 --> 00:01:02.077 we got here, what we are, who we are. 00:01:02.597 --> 00:01:09.537 And you can think of it in development, or you can think of in religious terms, 00:01:09.617 --> 00:01:15.657 but you can also think about an evolutionary past, from where we started and now we got here. 00:01:16.077 --> 00:01:22.397 And for a neuroscientist, I think the perspective of how our brains got to be 00:01:22.397 --> 00:01:26.737 the way they are is the most informative and interesting. 00:01:29.377 --> 00:01:35.497 Given that what we are is so much different than what other species are in terms 00:01:35.497 --> 00:01:38.097 of abilities, in terms of what they do, and so on. 00:01:38.097 --> 00:01:47.137 And one of the statements I remember is, if you're criticizing someone, are you a man or a mouse? 00:01:48.037 --> 00:01:52.837 The real difference, the main important difference between a man and a mouse 00:01:52.837 --> 00:01:56.057 is the brain, I think. The things that the brain can do. 00:01:56.537 --> 00:02:01.817 Of course, there are some slight morphological differences. They have a body 00:02:01.817 --> 00:02:03.917 that's appropriate for the brain as well. Yeah. 00:02:05.962 --> 00:02:12.062 But then you define, let's say, you identify evolution and let's say the phylogeny 00:02:12.062 --> 00:02:15.402 of brains as a phenomenon in and of itself. 00:02:15.722 --> 00:02:20.902 So this is one perspective. But do you see it also contribute to, 00:02:20.962 --> 00:02:25.382 let's say, extracting these fundamental principles along which any brain is 00:02:25.382 --> 00:02:27.742 organized? Yeah, so the same sort of thing. 00:02:28.302 --> 00:02:32.462 I take the evolution of the human brain because it'd be the most interesting 00:02:32.462 --> 00:02:36.882 for the most number of people. but you could pick anything. 00:02:37.082 --> 00:02:41.222 How would you get to an unusual brain such as in a duck-billed platypus? 00:02:41.522 --> 00:02:45.602 How would you ever get a brain like that? Why is it organized the way it is? 00:02:47.422 --> 00:02:50.722 And it applies if you're talking about rats. 00:02:50.902 --> 00:02:53.922 Why do they have such a wonderful vibrancy system and so on? 00:02:53.982 --> 00:03:01.082 You can start asking questions of how any mammal's brain got the way it is from 00:03:01.082 --> 00:03:03.262 the early beginning. any. 00:03:03.422 --> 00:03:08.042 But then if you let's take the platypus which is sort of very close to the common 00:03:08.042 --> 00:03:14.402 ancestors of our brain why is my brain not exactly like the brain of a platypus? 00:03:14.562 --> 00:03:17.062 What are the differences and why are these differences there? 00:03:18.322 --> 00:03:23.962 Well first there are some amazing similarities, 00:03:25.042 --> 00:03:30.402 and so all mammals will have this layered cortis which is quite different from 00:03:30.402 --> 00:03:33.502 any non-mammal And that's something we share in common. 00:03:34.502 --> 00:03:38.222 So one of the points I try to make is that given a six-layered cortex, 00:03:38.482 --> 00:03:45.662 it can be changed and modified in a number of ways to fulfill the needs of all 00:03:45.662 --> 00:03:50.062 kinds of different species and depending on what they need to do. 00:03:50.182 --> 00:03:56.122 A lot of species have a high reproductive rate and a short life. 00:03:56.122 --> 00:04:00.562 And if you're going to have a short life, you need a high reproductive rate, 00:04:00.602 --> 00:04:02.322 you need to reproduce early and so on. 00:04:02.362 --> 00:04:07.462 So you can't really invest under those circumstances in a complex brain. 00:04:07.582 --> 00:04:10.802 It takes too long to build, too much energy. 00:04:12.182 --> 00:04:19.862 So we happen to be one of those lucky species that has figured out how to live a long time. 00:04:19.982 --> 00:04:25.342 And so we can have a brain that takes 20 years to really fully develop. 00:04:25.342 --> 00:04:30.402 And that's almost unique, that luxury. 00:04:31.422 --> 00:04:37.682 But it leads to a very complicated brain that a whole range of other species never invest in. 00:04:37.762 --> 00:04:43.882 They just want to get reproducing within a few months, and they're unlikely 00:04:43.882 --> 00:04:45.622 to live more than a year or two. 00:04:46.824 --> 00:04:53.504 So you have those kind of brains. But a duck-billed platypus has taken a very 00:04:53.504 --> 00:04:59.164 unusual tact, and that is that it lives in water and feeds in water. 00:05:00.424 --> 00:05:03.904 Some other animals have done that, mammals have done that. 00:05:03.984 --> 00:05:12.144 But in this case, they were able to develop from mucous glands electroreception. 00:05:12.144 --> 00:05:17.704 And at high levels of current, I guess we all have electroreception, 00:05:18.104 --> 00:05:21.424 but to get the high level of sensitivity, 00:05:22.204 --> 00:05:26.464 it's not clear exactly how that's come about, but it's happened independently 00:05:26.464 --> 00:05:29.624 in several fish and so on. 00:05:31.624 --> 00:05:37.124 And once you have that, you can exploit this environment 00:05:37.384 --> 00:05:43.344 of turbid, muddy water that you can't see in, you can't really hear when you're 00:05:43.344 --> 00:05:47.524 adapted as a mammal to be above on land part of the time. 00:05:48.984 --> 00:05:52.864 So you can touch things, but then you're already up there. 00:05:55.044 --> 00:06:00.304 So electroreception is what makes them unique. It makes them different as getting electroreception. 00:06:00.384 --> 00:06:05.584 They refine that to a tremendous degree, come their most important sense. 00:06:07.124 --> 00:06:12.224 So the duck-billed platypus is interesting because it split from the main line 00:06:12.224 --> 00:06:16.764 of mammalian evolution a very long time ago and followed a very distinctive 00:06:16.764 --> 00:06:19.784 path, so it's almost now a unique creature. 00:06:20.844 --> 00:06:24.324 Now the common ancestor that the duck-billed platypus has with other mammals 00:06:24.324 --> 00:06:27.124 lived around 250 million years ago. 00:06:27.624 --> 00:06:33.744 And what is your view of what that animal was like, why it evolved to be different 00:06:33.744 --> 00:06:41.464 from a reptile, and how the evolution of that first animal changed the brain in such a dramatic way. 00:06:42.904 --> 00:06:49.464 Well, you're asking a tough question because of the lack of intermediates. 00:06:52.444 --> 00:06:58.604 But when it was proper to talk about the non-mammalian ancestors of mammals, 00:06:58.764 --> 00:07:03.444 they talked about them being mammal-like reptiles. 00:07:04.684 --> 00:07:05.564 And, uh. 00:07:06.809 --> 00:07:13.129 Their brains were bigger, but it wasn't clear whether they had a neocortex or just a dorsal cortex, 00:07:13.329 --> 00:07:19.709 and it's not clear how dorsal cortex of a single layer of pyramidal cells with 00:07:19.709 --> 00:07:26.329 a different kind of wiring diagram and processing of information and sending information back out, 00:07:26.389 --> 00:07:30.689 how that kind of cortex changed to the neocortex. 00:07:31.009 --> 00:07:38.109 And in my talk, I emphasize neocortex because because this structure is not new. 00:07:38.289 --> 00:07:42.369 A lot of people complain it shouldn't be called neocortex, and some people have 00:07:42.369 --> 00:07:46.729 called it isocortex, but it's new in the sense that it's a six-layered structure 00:07:46.729 --> 00:07:52.669 that evolved from something that was a one-layered structure with fiber layers on each side, 00:07:52.769 --> 00:07:55.669 so you could really call it three if you wanted to. 00:07:56.009 --> 00:08:01.689 But it's one layer of cells that gets the input, and the same cells provide 00:08:01.689 --> 00:08:07.549 the output, So it's very simple in kind of the wiring diagram it is. 00:08:07.849 --> 00:08:15.529 Changing from that into six layers gave the great flexibility and possibilities 00:08:15.529 --> 00:08:20.709 to early mammals to diverge and develop in so many different ways. 00:08:20.709 --> 00:08:27.869 A lot of the ways were just simply modifying the way sensory inputs were represented, 00:08:28.209 --> 00:08:34.929 enlarging what was important, putting more neurons into what seemed to be behaviorally important. 00:08:35.449 --> 00:08:41.529 So you get large vibrisci representations, for example, or large nose representations, 00:08:41.829 --> 00:08:47.089 or large mouth representations. All these are common variations in the somatosensory system. them. 00:08:47.709 --> 00:08:51.589 A lot devoted to vision, little devoted to vision, these kind of things. 00:08:51.649 --> 00:08:56.649 I'd say these are what we'd say is rather modest modifications, 00:08:57.209 --> 00:09:04.349 echolocating bats devoting a tremendous amount of space and cortex to the echo frequency. 00:09:05.666 --> 00:09:12.446 Another example. So would it be true to say that the dorsal cortex of the stem 00:09:12.446 --> 00:09:19.086 reptiles from which these first mammals evolved was a relatively simple layered 00:09:19.086 --> 00:09:20.406 structure, just three layers, 00:09:20.946 --> 00:09:24.046 but was it also doing sensory processing? 00:09:24.626 --> 00:09:32.786 I would say it had more to do with something like memory or something like there's 00:09:32.786 --> 00:09:37.366 evidence that had to do with habituation, which would be a kind of learning, short-term learning. 00:09:38.266 --> 00:09:43.846 And this would be very valuable, but it wouldn't be very useful in detecting 00:09:43.846 --> 00:09:52.066 a bug from a ball or making discriminations of what kind of sensory information this was. 00:09:52.166 --> 00:09:58.226 That information would be done subcortically. So the auditory and the tactile 00:09:58.226 --> 00:10:03.226 and the visual brain of the stem reptile would be a midbrain, 00:10:03.226 --> 00:10:06.446 and then in the first mammals, 00:10:06.586 --> 00:10:08.546 this dorsal cortex would suddenly start 00:10:08.546 --> 00:10:13.386 to take over or be involved in the representation of these modalities. 00:10:13.386 --> 00:10:18.546 Yeah, midbrain and parts of the forebrain that would be now considered basal 00:10:18.546 --> 00:10:26.866 ganglion mammal, and certainly the amygdala was a large part of this subcortical processing, 00:10:27.226 --> 00:10:30.626 and still is for a lot of kind of behaviors. 00:10:31.166 --> 00:10:36.306 And the pressure to develop that new function for dorsal cortex, 00:10:36.646 --> 00:10:41.026 would that be to do with the lifestyle of those first mammals and how that had shifted? 00:10:41.026 --> 00:10:45.646 Well, it's presumed that the lifestyle would have been a little precarious for 00:10:45.646 --> 00:10:53.346 them because you had all these very vast reptiles, dinosaurs all around, many of them predators. 00:10:53.706 --> 00:10:58.226 And so it was assumed that they would be nocturnal. Their skeletal features 00:10:58.226 --> 00:11:02.706 suggest that they were nocturnal. They were small. 00:11:02.886 --> 00:11:07.566 They were living on insects or maybe some of the meeting some of the infants 00:11:07.566 --> 00:11:13.606 of smaller reptiles and things like that, eggs perhaps, but mainly on insects 00:11:13.606 --> 00:11:15.486 and scurrying around and hiding at night. 00:11:15.946 --> 00:11:19.766 And a couple things became important. 00:11:19.926 --> 00:11:25.806 One was the change in the auditory system in which the bones of the jaw became 00:11:25.806 --> 00:11:29.306 dissociated from the jaw. So you have the three. 00:11:31.179 --> 00:11:35.759 Bones of the inner ear, which really defines mammals as something unique, 00:11:35.979 --> 00:11:43.599 and that allowed a high-frequency transduction of sounds for the first time. 00:11:43.899 --> 00:11:49.859 So early mammals had a nice advantage over reptiles in that they could communicate 00:11:49.859 --> 00:11:52.539 mother to offspring, or mother. 00:11:53.579 --> 00:11:57.239 Offspring could make a separation call or something at a high enough frequency 00:11:57.239 --> 00:12:03.779 that predator couldn't find them, and they started to exploit this kind of life. 00:12:04.119 --> 00:12:09.519 And now we presume that auditory inputs to dorsal cortex, which would be now 00:12:09.519 --> 00:12:13.699 neocortex at some stage, would have emerged for the first time. 00:12:13.799 --> 00:12:19.879 And that kind of auditory processing at a higher level would emerge with the 00:12:19.879 --> 00:12:22.299 ancestors or with the first mammals. 00:12:23.059 --> 00:12:27.759 Do you see then also a link of olfaction here, because about nocturnal animals, 00:12:28.019 --> 00:12:32.579 so you want to sort of rely on sensory modalities that don't require light. 00:12:33.079 --> 00:12:34.439 Now you emphasize audition. 00:12:35.279 --> 00:12:39.299 Recently there was a paper out where people look at the fossil record, 00:12:39.499 --> 00:12:43.059 making the argument that actually this whole drive towards the cortex was very 00:12:43.059 --> 00:12:45.599 much grounded in olfaction. Do you support that? 00:12:45.879 --> 00:12:51.239 Well, clearly the first mammals emphasized olfaction a lot. 00:12:51.459 --> 00:12:53.999 And this is a really old idea, going back to Herrick. 00:12:54.219 --> 00:13:02.359 A hundred years ago, the evidence from looking at mammals in a comparative way, 00:13:02.439 --> 00:13:08.139 but also some of the fossil evidence, suggested that olfaction was very important. 00:13:08.279 --> 00:13:14.099 They used mainly comparison of living mammals, but it's been substantiated by 00:13:14.099 --> 00:13:19.819 everything that's been learned since then, so that if you had to look at the 00:13:19.819 --> 00:13:21.199 forebrain, it was dominated 00:13:21.459 --> 00:13:23.599 by olfactory processing. 00:13:23.939 --> 00:13:27.559 And this would be very handy for a nocturnal animal to find their food, 00:13:27.639 --> 00:13:34.779 find their mates, not use vision, not really use sound so much because that 00:13:34.779 --> 00:13:36.919 would lead predators to find you. 00:13:39.155 --> 00:13:44.075 So the olfaction is obviously crucial. But another interesting thing with the 00:13:44.075 --> 00:13:47.795 first mammals is the emergence of hair. 00:13:48.255 --> 00:13:53.595 And I think it's presumed that before hair could have a thermoregulation function, 00:13:54.475 --> 00:13:57.255 it would probably have had a function as a touch sensor. 00:13:57.715 --> 00:14:01.335 And of course, that was probably in the form of the whiskers, 00:14:01.335 --> 00:14:03.235 the fibrissi of the first mammals. 00:14:03.515 --> 00:14:06.255 And so then you also have somatosensory cortex. 00:14:07.855 --> 00:14:11.535 Yeah, it's very important. 00:14:11.995 --> 00:14:17.175 The sensory hairs probably were the most important feature of early hairs. 00:14:17.555 --> 00:14:24.635 They can touch something before they quite touch it. The longer the hair, the better. 00:14:25.135 --> 00:14:30.735 And this would be especially a great advantage in poor light or bad light, 00:14:30.775 --> 00:14:32.235 where vision wouldn't be so useful. 00:14:32.315 --> 00:14:37.375 You'd touch something and you could back away before it was too late or decide 00:14:37.375 --> 00:14:39.755 what kind of behavior you might initiate. 00:14:40.075 --> 00:14:46.775 And a good example of that would be in naked mole rats, which have been studied 00:14:46.775 --> 00:14:48.135 at Vanderbilt and other places. 00:14:48.435 --> 00:14:52.995 And they're not naked. They have these sensory hairs, but the other ones are 00:14:52.995 --> 00:14:55.215 gone because they don't need them for thermoregulation. 00:14:55.835 --> 00:14:59.855 And they're very driven by what hairs you touch. 00:15:00.015 --> 00:15:03.515 So they don't see you, You touch with a little stick or something, 00:15:03.735 --> 00:15:07.155 one part of the body, they go forward. 00:15:07.215 --> 00:15:12.215 Another part of the body, they go backwards because where they're touched determines 00:15:12.215 --> 00:15:14.035 what direction they're going to move in. 00:15:14.755 --> 00:15:18.235 So now we have sort of a sketch of this early mammal. 00:15:21.335 --> 00:15:25.575 But it also then had to develop a brain that would match these capabilities. 00:15:25.575 --> 00:15:31.115 And then the idea would be this would contribute from this step towards a six-layer cortex, right? 00:15:31.155 --> 00:15:34.175 It would just sort of highlight the isocortex in this case. 00:15:35.135 --> 00:15:41.415 But still then within that, so now there will be two questions at least. 00:15:41.415 --> 00:15:45.795 Or the one of this, okay, how do I get from a one-layer sheet of cells performing 00:15:45.795 --> 00:15:49.975 some very simple, possibly memory function, learning function, 00:15:50.095 --> 00:15:53.675 to a six-layer structure, and also so fairly rapidly? 00:15:53.955 --> 00:15:59.215 But then how do I make that match all these variations on morphology, 00:15:59.375 --> 00:16:02.335 on mixing of different sensor modalities? 00:16:04.604 --> 00:16:08.404 Efficiently and rapidly? Because now we're talking K-min explosion roughly, 00:16:08.584 --> 00:16:11.624 so we have all these different types of animals emerging. 00:16:12.724 --> 00:16:20.044 So how do you see the six-layer cortical sheet being rapidly adjusted in its 00:16:20.044 --> 00:16:22.204 details to then the specific requirements? 00:16:22.504 --> 00:16:25.444 So how do we get from one layer to six? And then once we have six, 00:16:25.624 --> 00:16:29.644 how do we tune it actually to the specifics of the specific organism in which 00:16:29.644 --> 00:16:30.784 you want to have that cortex working? 00:16:32.384 --> 00:16:39.644 Well, the speculation of how you get to more layers would be that the original 00:16:39.644 --> 00:16:45.564 organization really corresponded to layer 5 pyramidal cells. 00:16:46.044 --> 00:16:52.184 They get inputs on their apical dendrites and they send outputs subcortically, 00:16:52.304 --> 00:16:54.964 and this is what dorsal cortex does with a few inner neurons. 00:16:56.424 --> 00:17:03.004 How those other layers came about and any ball is really rather uncertain. 00:17:03.824 --> 00:17:08.704 People working on development probably will come up with and are coming up with 00:17:08.704 --> 00:17:13.804 some good suggestions about cell migration, cell division, changing the cell 00:17:13.804 --> 00:17:16.304 replication cycle, and so on. 00:17:16.404 --> 00:17:21.064 But it's really uncertain because we don't have examples of intermediate forms. 00:17:22.744 --> 00:17:28.724 Once you get to the six layers, you can modify the sensory representations representations 00:17:28.724 --> 00:17:31.484 that are there, as I already mentioned, in all kinds of ways. 00:17:32.284 --> 00:17:38.164 But the real powerful advantage of cortex is that you can start to replicate 00:17:38.164 --> 00:17:41.064 cortical areas and increase the numbers. 00:17:41.884 --> 00:17:48.744 One idea that we talked about a long time ago is just you duplicate areas that 00:17:48.744 --> 00:17:51.904 are already there and then modify them differently for different functions. 00:17:52.084 --> 00:17:56.284 I don't know if this is the way new areas emerge or where they emerge gradually 00:17:56.284 --> 00:17:58.124 by internal differentiation, 00:17:58.404 --> 00:18:01.264 but the idea of duplicating is 00:18:01.264 --> 00:18:04.164 something that's an old advantage in evolution 00:18:04.164 --> 00:18:06.944 where you duplicate body parts and as you 00:18:06.944 --> 00:18:11.864 can see in segmented creatures that have more segments and more body parts that 00:18:11.864 --> 00:18:16.784 were originally all alike and then you start specializing them as in a lobster 00:18:16.784 --> 00:18:21.724 with different of the appendages modified for different functions you can think 00:18:21.724 --> 00:18:25.144 of cortex with more areas doing something similar, 00:18:25.244 --> 00:18:29.724 is that you're now modifying different parts that might have been quite similar 00:18:29.724 --> 00:18:32.644 in anatomy and connections and, 00:18:33.424 --> 00:18:37.844 making them more and more specialized for different kinds of functions. 00:18:38.024 --> 00:18:43.284 If you don't have a number of areas, you can't specialize for different functions 00:18:43.284 --> 00:18:44.624 with one area very easily. 00:18:44.844 --> 00:18:50.224 You can subdivide an area into more layers and get some of that in a single structure. 00:18:51.164 --> 00:18:59.584 By having layer differences, but really the best way to add is to get more areas 00:18:59.584 --> 00:19:00.924 and then specialize them. 00:19:01.124 --> 00:19:05.884 And you specialize them in terms of what their inputs and outputs are, 00:19:05.964 --> 00:19:12.164 but you specialize them also in adjusting them to what they're best at doing 00:19:12.164 --> 00:19:13.324 or what you want them to do. 00:19:13.444 --> 00:19:19.684 So you can have dendritic arbors that are large or small, and they collect information 00:19:19.684 --> 00:19:22.744 from a few neurons or a lot of different other neurons. 00:19:23.224 --> 00:19:27.344 You can use different neurotransmitters. You can have different neuromodulators 00:19:27.344 --> 00:19:29.044 that will affect them in a different way. 00:19:30.864 --> 00:19:35.884 And you can end up, as we hear in all these talks, with neurons that are sensitive 00:19:35.884 --> 00:19:41.584 to very particular kinds of combinations of sensory information. 00:19:43.664 --> 00:19:50.204 In my mind, there's sort of a gap still in that answer Because while I'm changing 00:19:50.204 --> 00:19:53.504 my morphology, it's not like I can just change my morphology and I flop around 00:19:53.504 --> 00:19:57.264 for a few generations and at some point in time I have a brain that matches my morphology. 00:19:57.884 --> 00:20:00.864 These changes have to happen in a very coordinated fashion, right? 00:20:00.864 --> 00:20:07.644 So that means I would hope that you have an answer, just to satisfy my own ignorance, 00:20:07.904 --> 00:20:10.884 that would mean something like, well, there is a very specific genetic control 00:20:10.884 --> 00:20:15.744 that allows you to change body parts and in some sense change also matching 00:20:15.744 --> 00:20:19.624 neural structures in some coordinated fashion. Yes, absolutely. 00:20:20.024 --> 00:20:27.164 So one thing that you can imagine is that you change brain parts without changing 00:20:27.164 --> 00:20:31.544 peripheral morphology at all and get more out of the kind of information that's coming in. 00:20:31.784 --> 00:20:40.024 But there is a cascading sort of modification that occurs in any system. So. 00:20:41.073 --> 00:20:44.813 The whole system is changed by that, at least in development, 00:20:44.953 --> 00:20:46.773 but also in mature animals. 00:20:46.953 --> 00:20:53.413 And that's due to the plasticity or the flexibility of the system to always adjust to change. 00:20:53.673 --> 00:21:00.913 And so one of the major mechanisms of evolution is that you change the sensory inputs. 00:21:01.213 --> 00:21:08.533 Say, you change the cochlea so that you can now send signals at higher frequencies. 00:21:08.613 --> 00:21:13.793 The whole system, If you did that in any animal that didn't hear high frequencies, 00:21:13.913 --> 00:21:18.753 if you gave them the possibility of high frequencies, the whole system in development 00:21:18.753 --> 00:21:22.753 would adjust to that and have neural space devoted to this. 00:21:23.473 --> 00:21:28.113 If you doubled the number of receptors that come from the face, 00:21:28.293 --> 00:21:33.513 the whole system would adjust to that and you would have more processing throughout. 00:21:33.513 --> 00:21:36.293 Out and so your answer is essentially say 00:21:36.293 --> 00:21:39.333 look the first trick was maybe to develop a brain or 00:21:39.333 --> 00:21:45.173 cortex as well that is let's say hyper plastic or hyper adaptive and then after 00:21:45.173 --> 00:21:48.993 that you can start to play with your morphology to to sort of see how you can 00:21:48.993 --> 00:21:53.033 optimize yourself to a specific niche yeah yeah you have things that will happen 00:21:53.033 --> 00:21:57.153 automatically and and one of the. 00:21:59.493 --> 00:22:06.553 Papers from Van der Loos was how the periphery instructs the brain how to build itself. 00:22:07.813 --> 00:22:12.893 And I think that paper was not as influential as it could have been, 00:22:12.933 --> 00:22:18.473 in part because he was looking for genetic reasons as well at the same time, 00:22:18.493 --> 00:22:20.453 and he deleted his own message that way. 00:22:20.453 --> 00:22:25.953 The powerful thing was working with mice, if they're born with an extra whisker 00:22:25.953 --> 00:22:30.013 or a whisker short or a couple whiskers extra or a couple short, 00:22:30.213 --> 00:22:36.193 you still have a whole vibrancy system that incorporates that new information 00:22:36.193 --> 00:22:41.153 or that lack of information into the number of barrels or barrelettes or whatever 00:22:41.153 --> 00:22:43.773 in different parts of the somatosensory system. 00:22:43.953 --> 00:22:49.593 So the system developed to accommodate the change in the periphery. 00:22:50.922 --> 00:22:55.302 And even that change in the periphery is somewhat indirect because it depends 00:22:55.302 --> 00:23:00.982 a bit on how folds in the face occur in development, and that determines how 00:23:00.982 --> 00:23:02.382 many whiskers there's going to be. 00:23:03.122 --> 00:23:08.982 So we have this early mammal brain, which is plastic and allows quite a lot 00:23:08.982 --> 00:23:11.802 of diversity to happen, but it's still really quite a small brain. 00:23:12.062 --> 00:23:15.902 And then you mentioned in your talk something that was very interesting, 00:23:16.002 --> 00:23:21.582 that around 60 million years ago, the reptile disappeared, 00:23:22.002 --> 00:23:26.922 and that this was an opportunity for a new radiation of mammals to diversify 00:23:26.922 --> 00:23:31.802 into a whole new set of niches in which they could live. 00:23:32.042 --> 00:23:36.462 And that there were then some really quite dramatic changes in the brain as 00:23:36.462 --> 00:23:38.682 a result of that for some groups of mammals. 00:23:38.922 --> 00:23:43.242 And one of those, obviously, is the group of mammals that led towards primates 00:23:43.242 --> 00:23:45.282 and eventually to ourselves. 00:23:45.942 --> 00:23:50.962 So what do you see as the most important changes from those earlier mammals 00:23:50.962 --> 00:23:54.082 to, say, the early primates? 00:23:54.762 --> 00:24:00.022 Well, that did give an opportunity for animals to become diurnal. 00:24:00.042 --> 00:24:05.802 It became a lot of opportunities for different kinds of environments that they could occupy and so on. 00:24:06.962 --> 00:24:13.842 Their risk of being preyed upon changed from these very efficient reptilian 00:24:13.842 --> 00:24:19.382 predators to the mammals that had to evolve into becoming efficient predators 00:24:19.382 --> 00:24:23.942 and opening up that predator environment for them. 00:24:23.942 --> 00:24:32.222 But early primates developed apparently from animals that were emphasizing vision 00:24:32.222 --> 00:24:37.882 more and living in bushes or trees and trying to go into the fine branch environment 00:24:37.882 --> 00:24:40.602 and live on insects and find food there. 00:24:40.762 --> 00:24:45.762 And it gave them an advantage of escaping some predators by being away from them. 00:24:46.622 --> 00:24:54.022 But it put demands on visual processing in dim light that were very powerful 00:24:54.022 --> 00:24:58.882 and so they had to devote more to vision. 00:24:58.922 --> 00:25:04.362 So eyes were big, a lot of receptors, but the whole visual system started to 00:25:04.362 --> 00:25:10.362 expand and devote more neurons and more cortical areas for sequential processing. 00:25:10.362 --> 00:25:15.542 Processing, and already in the ancestors of present-day primates, 00:25:15.622 --> 00:25:17.522 the temporal lobe would have been expansive. 00:25:18.382 --> 00:25:24.062 The fossil record shows the temporal lobe and occipital region are both expansive. 00:25:24.182 --> 00:25:27.422 Both of them are thought to be completely devoted to vision, 00:25:27.582 --> 00:25:33.382 and so the cortex now is going over and covering the whole midbrain and much of the cerebellum. 00:25:34.002 --> 00:25:39.602 You can get that from the fossil And, of course, if you're going around and 00:25:39.602 --> 00:25:47.522 catching things rapidly and find branches, insects that could hide or escape or die. 00:25:50.616 --> 00:25:57.796 Defend themselves in some sense, but you have to be agile. You have to be able to, 00:26:00.356 --> 00:26:05.316 hold on to a branch that may be moving up and down and reach and get something 00:26:05.316 --> 00:26:09.336 that may be on another branch and isn't moving or is moving in a different way. 00:26:09.656 --> 00:26:12.756 And you have to use visual guidance for these kind of movements. 00:26:13.156 --> 00:26:21.456 And grasping things with a forepaw would be something to be important and new, 00:26:21.556 --> 00:26:25.056 sort of, instead of grabbing things with your mouth, grabbing them with a forepaw. 00:26:25.216 --> 00:26:28.416 So, eye-hand coordination would start to become important. 00:26:28.996 --> 00:26:37.476 And so, all primates then, all living primates, have a huge investment at the 00:26:37.476 --> 00:26:39.936 cortical level in processing visual information, 00:26:40.116 --> 00:26:49.696 but also in processing sensory information relative relative to simple motor skills. 00:26:50.776 --> 00:26:54.676 So, yeah, and I think that's quite important, the idea of the hand, 00:26:54.736 --> 00:26:59.576 and also the hand being away from the mouth, so you can look where you want 00:26:59.576 --> 00:27:02.876 to eat something or pursue something and go with your hand for it. 00:27:02.976 --> 00:27:06.496 And in a way, also, that frees you from having the need for the whiskers quite 00:27:06.496 --> 00:27:12.296 so much because you now use your hand as this organ for touching further away in the world. 00:27:12.996 --> 00:27:18.116 But one of the things about primates, I think, that people notice is that they 00:27:18.116 --> 00:27:19.756 live in large social groups, 00:27:19.896 --> 00:27:25.816 and that the development of these societies of primates was possibly one of 00:27:25.816 --> 00:27:28.456 the important factors in their success. 00:27:28.736 --> 00:27:33.056 And is that reflected in the changes in the primate brain? Not all primates 00:27:33.056 --> 00:27:36.576 live in social groups, and a number of the prosimians don't. 00:27:36.576 --> 00:27:41.056 So it's uncertain, I think, if the first primates were social or not. 00:27:41.056 --> 00:27:44.896 There'd have to be some kind of evidence one way or other, and I'm not aware of any. 00:27:45.676 --> 00:27:51.156 They might have been solitary and nocturnal. The suggestion would be that they're nocturnal. 00:27:51.176 --> 00:27:56.996 But if you're going to go to the ground, or you're going to be out in the daylight 00:27:56.996 --> 00:28:02.256 and the ground, it's a tremendous advantage to be in a social group, 00:28:02.276 --> 00:28:05.696 because then you can be detected visually by a lot of different predators. 00:28:06.136 --> 00:28:10.756 You need a lot of sentinels. You have to have somebody warning you that there's 00:28:10.756 --> 00:28:15.096 danger and you're not too distracted by trying to find food or other things. 00:28:15.236 --> 00:28:18.556 So a social group then becomes very important. 00:28:19.838 --> 00:28:24.058 And a side benefit of a social group is you can defend a territory against other 00:28:24.058 --> 00:28:27.098 social groups or non-social animals and drive them out. 00:28:27.338 --> 00:28:33.058 And that would then impact on the sort of non-motor or less motor and less sensory 00:28:33.058 --> 00:28:35.598 areas of the brain and the way that they would change. 00:28:35.898 --> 00:28:42.058 This would impact on, I think, especially frontal cortex and enlarging frontal cortex. 00:28:42.218 --> 00:28:48.638 And you see in the monkeys and all the anthropoid primates that the frontal 00:28:48.638 --> 00:28:54.158 lobe is really well developed, but it's not very well developed in the prosimian 00:28:54.158 --> 00:28:57.098 primates by comparison to monkeys. 00:28:57.878 --> 00:29:03.278 By comparison to most other animals, the frontal lobes are quite well developed 00:29:03.278 --> 00:29:05.058 in even prosimian primates. 00:29:05.438 --> 00:29:11.778 So now we went with our cortex from the first vertebrates, and. 00:29:14.038 --> 00:29:20.018 Up now to the primates. But in some sense, that's not the only part of the brain 00:29:20.018 --> 00:29:22.238 that evolved or that changed, right? 00:29:22.378 --> 00:29:24.858 There's, let's say, also an underlying architecture. 00:29:25.338 --> 00:29:29.098 And also in your talk, you mentioned the McLean proposal of the triune brain, 00:29:29.398 --> 00:29:32.018 which might be considered a bit naive. 00:29:32.958 --> 00:29:35.898 But it was like one of these attempts to try to say like, okay, 00:29:35.918 --> 00:29:39.338 but these are the different parts of, let's say, an overall neural architecture, 00:29:39.898 --> 00:29:44.458 that we have to consider when we think about this evolutionary perspective. 00:29:44.938 --> 00:29:50.698 So what's your view on then these key ingredients of that architecture that 00:29:50.698 --> 00:29:52.598 actually leads to a successful brain? 00:29:54.082 --> 00:30:00.922 Well, every mammal or every animal has a successful brain in some sense. 00:30:01.082 --> 00:30:03.722 It has to be successful or they wouldn't be here. 00:30:04.402 --> 00:30:11.682 Well, I share features. That's the point, right? But you're right in pointing 00:30:11.682 --> 00:30:14.382 out that it's just not neocortex, 00:30:14.402 --> 00:30:19.562 but the whole brain is being modified in each line depending on what the requirements are. 00:30:19.562 --> 00:30:22.982 But they interact 00:30:22.982 --> 00:30:26.022 and I'll give a couple examples 00:30:26.022 --> 00:30:28.962 one example of course is optic tectum or 00:30:28.962 --> 00:30:32.722 superior colliculus as you're talking about mammals you don't 00:30:32.722 --> 00:30:40.622 get a cortical input to the optic tectum in non-mammals but you get a cortical 00:30:40.622 --> 00:30:47.222 input to the tectum but the regions that project to the cortex or the tectum 00:30:47.222 --> 00:30:51.822 superior colliculus in mammals is variable across mammals. 00:30:52.082 --> 00:30:57.522 So the inputs from frontal motor areas, for example, frontal eye fields and so on. 00:30:59.222 --> 00:31:03.742 That's unique to primates and maybe a few independently involved in a few other 00:31:03.742 --> 00:31:07.402 animals such as some carnivores and so on that are highly visual. 00:31:08.162 --> 00:31:13.822 So you're modifying a structure that has been important for vision through all 00:31:13.822 --> 00:31:19.702 vertebrates, basically. But you're modifying it by adding complexity to it and 00:31:19.702 --> 00:31:22.662 adding other ways of influencing its functions. 00:31:23.062 --> 00:31:29.322 But do you see this adding of complexity mainly occurring at this level of this isocortex? 00:31:29.742 --> 00:31:35.002 Or is it then also matched by the addition of complexity at these supporting 00:31:35.002 --> 00:31:37.322 midbrain and lower structures? 00:31:38.752 --> 00:31:44.152 So the midbrain will be modified, and architectonically it's modified. 00:31:44.192 --> 00:31:48.492 You can see it's changed greatly in different lines in terms of architectonic 00:31:48.492 --> 00:31:51.572 complexity, so its functions will be changed as well. 00:31:51.932 --> 00:31:58.092 The important feedback into the thalamus is very old, 00:31:58.212 --> 00:32:03.372 that there's a projection from the optic tectum into the thalamus, 00:32:03.412 --> 00:32:07.992 but how that's distributed to cortex has changed, its cortex has changed, 00:32:08.232 --> 00:32:16.132 its downstream projections become more important because they're influenced by other inputs, 00:32:16.432 --> 00:32:19.812 and so they're not really doing the same thing as they did before. 00:32:20.192 --> 00:32:23.692 And you could think of almost any structure. 00:32:23.972 --> 00:32:32.572 The amygdala would be responsive to direct inputs from the thalamic inputs to 00:32:32.572 --> 00:32:35.132 the amygdala that wouldn't depend on cortex at all, 00:32:35.132 --> 00:32:40.552 and they're highly important in a lot of mammals, but certainly non-mammals. 00:32:40.812 --> 00:32:44.392 Do you see, let's say that, because in some sense, 00:32:44.552 --> 00:32:53.532 the simple view on the phylogeny of the human brain would be like, 00:32:53.612 --> 00:32:56.352 okay, you have all these subcortical structures and they were sort of a constant 00:32:56.352 --> 00:33:01.052 and then this magnificent isocortex grew on top of that and that added all these 00:33:01.052 --> 00:33:02.132 amazing functionalities. 00:33:02.832 --> 00:33:06.812 An alternative view would be to say no actually it's a bit like we discussed 00:33:06.812 --> 00:33:12.032 earlier morphology and brain has to match and it's something you could say okay oh isocortex and. 00:33:13.352 --> 00:33:16.912 Subcortical structures have to match as well so it's always a co-development 00:33:16.912 --> 00:33:21.112 and change in these structures as opposed to these subcortical structures remaining 00:33:21.112 --> 00:33:23.412 constant and then your isocortex exploding, 00:33:23.952 --> 00:33:29.012 so where do you position yourself in that debate how should we look upon that well, 00:33:32.288 --> 00:33:37.388 So you're talking about systems that include cortical parts and other parts 00:33:37.388 --> 00:33:43.688 of the forebrain and parts of midbrain, hindbrain, spinal cord, 00:33:43.868 --> 00:33:46.528 and they all have to be coordinated and interrelated. 00:33:46.528 --> 00:33:53.468 But a lot of what happens depends on the changes in cortex to sort of drive 00:33:53.468 --> 00:33:56.108 these other changes, in my view. 00:33:58.628 --> 00:34:03.588 As motor centers in the cortex become important for controlling digit movements 00:34:03.588 --> 00:34:08.708 in hand and so on, you have to modify the circuitry in the cervical spinal cord 00:34:08.708 --> 00:34:12.108 that will deal with this control of muscles. 00:34:12.108 --> 00:34:21.448 Muscles, but deciding when to do something or how to do something and so on would depend on cortex. 00:34:21.668 --> 00:34:29.708 So it would be a series of gradual changes, I think, over many, 00:34:29.848 --> 00:34:35.728 many generations that would modify different parts but always together so that 00:34:35.728 --> 00:34:40.568 it would be pointless to have some function you couldn't use. 00:34:41.488 --> 00:34:47.228 But then if we now reduce a bit the level of granularity in that discussion, 00:34:48.308 --> 00:34:52.468 so now we look at, let's say, this brain with its different parts, 00:34:52.668 --> 00:34:57.128 we might have to sort of define what the key subdivisions are there. 00:34:59.048 --> 00:35:05.788 Co-evolving with the morphology, but now these larger chunks like isocortex 00:35:05.788 --> 00:35:10.208 and its underlying structure, like thalamus or basal ganglia or amygdala, 00:35:10.908 --> 00:35:12.708 again, consists of the subcircuits. 00:35:13.428 --> 00:35:17.768 And these subcircuits also will have some invariant features and also some variable features. 00:35:18.308 --> 00:35:22.768 So what's your view on that? What are the invariant features of such an isocortex 00:35:22.768 --> 00:35:28.108 and what are the variable features as such a brain is changing phylogenetically 00:35:28.108 --> 00:35:31.868 but also changing across the different modalities where it has to deal with? 00:35:33.855 --> 00:35:40.415 Well, tough question, but I've tried to emphasize some of the things that all mammals would have. 00:35:40.515 --> 00:35:47.155 It seems unlikely that we would involve any mammal without a somatosensory system 00:35:47.155 --> 00:35:50.135 that had somatosensory cortex involved in it. 00:35:50.215 --> 00:35:54.795 There are several divisions, but at least one division because we depend on 00:35:54.795 --> 00:35:57.915 this sort of sensory information so clearly. 00:35:57.915 --> 00:36:02.855 And once you relegate it to the cortical level, it seems like you're unlikely 00:36:02.855 --> 00:36:09.235 to change that and do those functions some other place. 00:36:10.795 --> 00:36:19.175 I can imagine the visual structures completely being lost in animals that no 00:36:19.175 --> 00:36:25.655 longer have functional form vision, as we see that those systems can be greatly reduced. 00:36:26.415 --> 00:36:31.815 Auditory could be lost. I can imagine it's not so important in some animals 00:36:31.815 --> 00:36:36.435 that live underground in tunnels, for example. You know what direction the sound 00:36:36.435 --> 00:36:40.275 is coming from because your body is blocking it or not and muffling. 00:36:41.235 --> 00:36:44.875 So sound localization as reduced sound is still important. 00:36:44.955 --> 00:36:50.595 You can imagine things being lost. lost in terms of higher order functions. 00:36:52.095 --> 00:36:57.935 You could go in any direction, I think. 00:36:58.055 --> 00:37:00.775 It could be modified in any way. 00:37:01.595 --> 00:37:08.215 It'd be easy for midbrain structures such as the superior colliculus to become 00:37:08.215 --> 00:37:14.795 completely auditory or completely somatosensory in function and not use vision at all anymore. 00:37:14.995 --> 00:37:21.295 It is already multimodal, and so you can just change functions. 00:37:21.535 --> 00:37:27.435 But the motor functions might be always very similar towards orienting towards 00:37:27.435 --> 00:37:29.195 a stimulus or something like that. 00:37:30.669 --> 00:37:35.309 Relay of sensory information as another route to cortex might be completely 00:37:35.309 --> 00:37:37.709 used in a quite different way. 00:37:37.909 --> 00:37:41.469 Right. So, a little procedure-wise. 00:37:41.789 --> 00:37:48.669 So, Tony wants to discuss with you a little bit developments in certain domains 00:37:48.669 --> 00:37:54.609 of science, like comparative neuroscience and so on, but then he has another appointment. 00:37:55.449 --> 00:37:59.049 So, then after that, I want to go back to some more specific questions and then 00:37:59.049 --> 00:38:03.409 later we just we added the interview that sort of these concluding questions 00:38:03.409 --> 00:38:08.789 go towards the end so that you're not sort of all right shocked by tony leaving 00:38:08.789 --> 00:38:13.609 and if you want to stop for lunch then tell him because otherwise he'll talk to you until it gets dark, 00:38:15.049 --> 00:38:22.549 exactly i will have to go and get my wife for lunch pretty soon okay um so well 00:38:22.549 --> 00:38:28.629 a couple of things if um the first thing i wanted to ask about multi-sensory areas in cortex. 00:38:29.489 --> 00:38:36.589 So it's clear that these early mammals have these new areas or sort of radically reduced. 00:38:38.537 --> 00:38:45.157 More sophisticated areas for analyses of the auditory and visual and spatiosensory inputs. 00:38:46.097 --> 00:38:51.397 Now, at what stage and whereabouts in their brains do they fuse these inputs? 00:38:51.817 --> 00:38:58.757 And how does the multisensory system evolve as mammals diversify? 00:39:00.117 --> 00:39:04.777 Opinions on this have changed a lot in recent times. When we were working early 00:39:04.777 --> 00:39:11.337 on in mapping visual areas 40 years ago, it was popular, 00:39:11.497 --> 00:39:17.097 and we believed it, to talk about visual areas or auditory areas or somatosensory areas. 00:39:17.317 --> 00:39:21.557 And we said, you know, most of the processing is done within a modality, 00:39:21.637 --> 00:39:24.757 and only at the very end, when things are very sophisticated, 00:39:25.037 --> 00:39:27.117 do you start to integrate modalities. 00:39:27.997 --> 00:39:33.697 And this partly was due to the methods of recording from anesthetized animals, 00:39:33.917 --> 00:39:38.417 where you really have dampened the responsiveness of neurons tremendously. 00:39:38.777 --> 00:39:42.497 And so you're getting this dominant input and say, yes, that's visual. 00:39:43.657 --> 00:39:50.297 And now when the anatomical methods and recording methods have been improved, 00:39:50.477 --> 00:39:54.917 you'll start to see people saying, it's multisensory everywhere. 00:39:55.277 --> 00:40:05.197 And Barry Stein had a book, is the editor of a book a few years ago on and multisensory systems. 00:40:05.617 --> 00:40:13.337 And you could say then it's amazing how fast multisensory processing evolved 00:40:13.337 --> 00:40:17.557 because it wasn't there very long ago, and now it's everywhere. 00:40:18.397 --> 00:40:22.797 But it depends on what you're looking at. 00:40:23.757 --> 00:40:31.977 Clearly, people are talking about in primates that even primary visual cortex 00:40:31.977 --> 00:40:37.917 gets auditory input, maybe somatosensory input, maybe not so directly, but... 00:40:39.194 --> 00:40:46.274 Some inputs rather directly. If you look at a small-brained animal like a rat 00:40:46.274 --> 00:40:48.294 or a mouse, you're seeing these kind of connections. 00:40:49.614 --> 00:40:53.114 Primary visual cortex will connect it everywhere almost. 00:40:55.054 --> 00:41:03.194 And if you look carefully at striped cortex projections like Henry Kennedy has been doing, 00:41:03.194 --> 00:41:11.114 And 99% of the connections can be accounted for with maybe five areas that you're talking about. 00:41:11.314 --> 00:41:17.074 But that 1% probably goes to 15 other areas and has some influence. 00:41:19.014 --> 00:41:25.874 So I think it's hard to find a neuron in the cortex that's not influenced by more than one modality. 00:41:25.894 --> 00:41:32.454 But whether it's driven by more than one modality, that would reduce the number 00:41:32.454 --> 00:41:34.894 of places. that you would find that. 00:41:36.134 --> 00:41:40.234 To give an example within a modality, 00:41:41.394 --> 00:41:45.694 you'll find in primary somatosensory cortex representing the hand in a monkey, 00:41:46.594 --> 00:41:50.834 neurons with small excitatory receptive fields on different parts of the digits, 00:41:50.854 --> 00:41:53.254 rather small, responding nicely. 00:41:54.494 --> 00:41:59.294 But if you get the neuron responding, and even in an anesthetized animal, 00:41:59.474 --> 00:42:04.434 and now have a second somatosensory input put somewhere else on the hand. 00:42:04.734 --> 00:42:06.154 It'll influence that firing. 00:42:07.174 --> 00:42:08.894 Then you go to the other hand on the other, 00:42:10.027 --> 00:42:16.887 and you will also influence that firing. And there's practically very few direct 00:42:16.887 --> 00:42:18.767 connections between the two hemispheres. 00:42:18.927 --> 00:42:25.187 For the hand areas, more connections would be from higher areas and then feedback connections. 00:42:25.427 --> 00:42:30.447 So it's not clear exactly how this is done, but it is clear that our traditional 00:42:30.447 --> 00:42:37.967 pictures of what neurons are responding to has been greatly constrained by our methods of looking. 00:42:37.967 --> 00:42:44.507 And when you look in other ways, you'll see the neurons are influenced by much 00:42:44.507 --> 00:42:49.147 more outside what's they call the classical receptive field in their own modality, 00:42:49.147 --> 00:42:50.847 but also from other modalities. 00:42:51.567 --> 00:42:56.187 Well, this is something very interesting. There's a caution here for students, 00:42:56.347 --> 00:42:59.047 that if you're reading the older literature, 00:42:59.427 --> 00:43:05.287 you may want to make sure that the opinions on these things haven't changed 00:43:05.287 --> 00:43:09.847 because there's been some really radical changes in methodology here that have 00:43:09.847 --> 00:43:13.967 made us revise our views on some of these fundamental questions. 00:43:14.307 --> 00:43:19.027 I guess another thing would be to say that this question of the multisensory 00:43:19.027 --> 00:43:23.007 cortex is really still open, that there's a lot more we need to know about this. 00:43:23.287 --> 00:43:29.927 And there are recognized areas in primates that are where neurons respond very 00:43:29.927 --> 00:43:37.127 readily, even in anesthetized animals to auditory and vision or vision and tactile and so on. 00:43:37.367 --> 00:43:43.947 And these would be called in earlier times or present times multisensory. 00:43:43.987 --> 00:43:50.107 Some of these areas have been called multisensory for a long time because of these responses. 00:43:51.487 --> 00:43:59.127 But 50 years ago, all the recordings were from a few primary sensory areas because 00:43:59.127 --> 00:44:02.187 anesthetics at that time just blocked everything else. 00:44:02.327 --> 00:44:05.327 So you had no chance of really talking about multisensory. 00:44:06.840 --> 00:44:10.780 So going to the field and how it's developed, 00:44:11.020 --> 00:44:17.240 so you've been in this field for a long time, and I think you were lucky to 00:44:17.240 --> 00:44:22.860 come into it at a time when people had this strong interest in comparative approaches 00:44:22.860 --> 00:44:25.460 and looking at different brains and different species. 00:44:25.460 --> 00:44:31.820 And now we have this fantastic new range of techniques to do this comparative 00:44:31.820 --> 00:44:37.080 study, but how do you feel the field of comparative neuroscience is looking 00:44:37.080 --> 00:44:40.100 in correspondence to the rest of the field? 00:44:40.380 --> 00:44:44.220 I mean, is this work happening, and where would you like the focus to be? 00:44:45.140 --> 00:44:48.300 Well, first, it's true I've been in the field for a long time. 00:44:49.280 --> 00:44:55.120 And this year I got a letter from Science Magazine, where we like to publish 00:44:55.120 --> 00:44:56.880 every once in a while. It's getting harder. 00:44:58.300 --> 00:45:03.040 And they said, you get a free subscription from now on. You don't have to pay 00:45:03.040 --> 00:45:05.720 for it anymore because you have subscribed for 50 years. 00:45:08.940 --> 00:45:12.720 So there are some advantages. I would say not that many, but there are some. 00:45:15.020 --> 00:45:21.040 And your question now is, how 00:45:21.040 --> 00:45:25.780 do I see the changes that have happened in comparative studies of brains? 00:45:26.780 --> 00:45:35.060 Has this field been influenced by technological advances as much as other parts of neuroscience? 00:45:35.540 --> 00:45:39.240 It's how you see the future for comparative neuroscience really more generally. 00:45:39.240 --> 00:45:44.000 So, there's obviously still many questions to be asked here, 00:45:44.140 --> 00:45:45.880 and how should we go about it? 00:45:45.900 --> 00:45:49.360 How should we balance our resources as a community between this comparative 00:45:49.360 --> 00:45:54.480 approach that you've been exploring and other approaches that are maybe focused 00:45:54.480 --> 00:45:56.380 more towards single-model animals? 00:45:57.240 --> 00:46:01.640 Well, I think the model animal approach has been a good one, 00:46:01.680 --> 00:46:03.200 and I wouldn't disparage it. 00:46:04.040 --> 00:46:09.220 Because if you work on a particular animal, you gain a lot of information that 00:46:09.220 --> 00:46:11.960 others have developed for you. 00:46:12.000 --> 00:46:15.940 If you work to another model or another animal, you have to do a lot of things 00:46:15.940 --> 00:46:18.680 over just to get to the point where you can answer the next question. 00:46:19.580 --> 00:46:26.980 So that becomes a problem. And sadly, in some ways, I started off working on 00:46:26.980 --> 00:46:33.540 cats as a model system, and they've been largely abandoned as a model system. 00:46:33.600 --> 00:46:37.140 And there's a tremendous amount of information known about brain organization 00:46:37.140 --> 00:46:44.360 in cats from years and years and years of study that is in some sense now not 00:46:44.360 --> 00:46:49.180 used to its full capacity because people have moved on to other models. 00:46:51.384 --> 00:46:57.064 That said, I think that we should always work on some primate models and some 00:46:57.064 --> 00:47:00.144 rodent models and maybe a few others, simple models. 00:47:00.704 --> 00:47:04.964 But simple models are limited in many ways. 00:47:05.184 --> 00:47:09.624 There was a debate at the Vision Science meeting, a friendly one just for the 00:47:09.624 --> 00:47:16.644 fun of it, and Tony Moshman was going to defend using simple models. 00:47:16.764 --> 00:47:20.964 He works on monkeys, so he was picked to defend using simple models. 00:47:20.964 --> 00:47:22.824 And he said, why stop at the mouse? 00:47:23.484 --> 00:47:27.744 C. elegans would be perfect because they're so cheap, the cost per individual 00:47:27.744 --> 00:47:29.324 is not even worth mentioning. 00:47:29.524 --> 00:47:33.064 They're so cheap, and they can do so many things. You can train them to move, 00:47:33.164 --> 00:47:36.524 turn to the left in a maze, and all kinds of things. 00:47:36.724 --> 00:47:39.244 And he went on for about 15 minutes about their virtues. 00:47:39.924 --> 00:47:45.204 But he said, there are two problems for vision research. They have no eyes, and they have no brain. 00:47:45.204 --> 00:47:49.924 So you have 00:47:49.924 --> 00:47:56.324 to have a model that has the capability or the potential of answering the kind 00:47:56.324 --> 00:48:00.384 of questions you're interested in and that means that we need a range of models 00:48:00.384 --> 00:48:05.764 because you can answer very general questions or very specific ones depending 00:48:05.764 --> 00:48:08.684 on what model you're using. 00:48:09.004 --> 00:48:16.324 But we also have a chance to study now a whole range of species and do it very 00:48:16.324 --> 00:48:19.964 productively because the methods are so powerful. You can learn so much. 00:48:20.804 --> 00:48:26.464 So you could take a new species, something that's never been looked at before, 00:48:26.624 --> 00:48:34.424 and you could know a lot about the brain of that species from one laboratory in three or four years. 00:48:34.604 --> 00:48:40.484 I think you could come up close to having a basic understanding of the brain 00:48:40.484 --> 00:48:44.864 organization in a relatively short time because the methods are so powerful. 00:48:46.397 --> 00:48:52.657 They were so poor when I was starting out. We were so limited in what you could 00:48:52.657 --> 00:48:54.197 do, and they were worse before then. 00:48:55.017 --> 00:49:03.537 And that people came up with such good ideas early on on such limited information is totally amazing. 00:49:03.877 --> 00:49:08.137 And given your vast experience and knowledge of different mammalian brains, 00:49:08.297 --> 00:49:12.057 are there particular animals that you think we should be exploring more? 00:49:12.917 --> 00:49:17.757 Well, from the point of view of evolution, brain evolution, which is of particular 00:49:17.757 --> 00:49:23.297 interest to me, it's important to learn more about the relatives of primates. 00:49:23.637 --> 00:49:29.557 Also, how much variability is in the primate order. We know a lot about a few monkeys. 00:49:29.857 --> 00:49:34.757 We don't know much about the diversity in monkeys. We know very little about apes. 00:49:35.557 --> 00:49:39.817 We're learning quite a bit about humans now from non-invasive methods. 00:49:39.817 --> 00:49:44.617 So there are big gaps that haven't been studied very well. 00:49:45.957 --> 00:49:53.677 Primate studies aren't very frequent. A review of publications over the last 00:49:53.677 --> 00:50:02.797 15 or so years by Paul Manger showed that almost all the studies are on mice or rats or humans. 00:50:02.997 --> 00:50:06.097 That takes almost all those neuroscience studies. 00:50:06.097 --> 00:50:11.237 Studies, then out of primate studies that aren't human, then you have macaque 00:50:11.237 --> 00:50:16.657 monkeys studied, and then the next would be prosimian galagos, 00:50:16.817 --> 00:50:20.057 and they're only studied in two laboratories that I know of. 00:50:20.177 --> 00:50:26.797 So a single person or a few people could make a big difference by filling in 00:50:26.797 --> 00:50:28.677 gaps in particular areas. 00:50:28.957 --> 00:50:32.977 It doesn't take many, but we have a lot of gaps. I think. 00:50:34.010 --> 00:50:38.150 To study animals like the monotremes would be very important because they're so unusual. 00:50:38.630 --> 00:50:45.170 We don't know anything about the brains, practically nothing about any of the larger marsupial. 00:50:45.690 --> 00:50:49.030 We know a lot about opossums because they're available in this country. 00:50:49.050 --> 00:50:50.290 In South America, they're small. 00:50:50.430 --> 00:50:54.070 You can get them in the laboratory, but we don't know anything about the red 00:50:54.070 --> 00:50:57.830 kangaroo, giant kangaroo brain, how it's similar, how it's different. 00:50:58.050 --> 00:51:01.070 We don't even know very much about carnivores at different sizes. 00:51:01.070 --> 00:51:07.810 So you could pick, almost in any order, animals that are just a mystery. 00:51:08.030 --> 00:51:13.910 And someone studying them would be bound to find something interesting and informative 00:51:13.910 --> 00:51:17.250 just by falling in and trying to do this. 00:51:17.410 --> 00:51:21.570 So one of the things that could be done in a lot of different countries, 00:51:21.710 --> 00:51:26.450 and there's been a tendency to do this in Brazil, take the native animals that 00:51:26.450 --> 00:51:29.330 are there and try to understand their brain organization. 00:51:30.290 --> 00:51:36.230 So they're looking at say the very large rodents and so on how their brains are organized, 00:51:37.430 --> 00:51:44.910 so then to follow up on that so you also showed in your talk an earlier map 00:51:44.910 --> 00:51:49.950 of the cortex and most of it was actually white so by now. 00:51:51.630 --> 00:51:56.690 How much of this map of the cortex do you feel have you really filled in can 00:51:56.690 --> 00:52:01.090 we say look these are areas that But we really have understanding how they're 00:52:01.090 --> 00:52:05.250 organized, how structure maps to function, and how big are the gaps. 00:52:08.309 --> 00:52:15.689 There are major gaps, but depending on who you ask, people will feel, 00:52:15.809 --> 00:52:18.309 well, we've got the gaps pretty well filled in or not. 00:52:18.729 --> 00:52:22.749 So you can look at Brodmann's maps. There are no gaps. 00:52:23.929 --> 00:52:30.729 Right. But there are a lot of errors and different kinds of errors. 00:52:32.509 --> 00:52:35.509 We can make complete maps 00:52:35.509 --> 00:52:42.829 at any time and it depends on a tolerance of ambiguity or we can say well we're 00:52:42.829 --> 00:52:47.449 basing this on a very limited amount of evidence and so on and the kind of map 00:52:47.449 --> 00:52:53.409 that for example that Thelman and Van Essen did had a lot of uncertainty which 00:52:53.409 --> 00:52:54.909 they recognized and talked about, 00:52:55.609 --> 00:53:00.349 and I think it's important to talk about uncertainty so people don't get the 00:53:00.349 --> 00:53:05.929 impression that we really understand all the little divisions and subdivisions 00:53:05.929 --> 00:53:09.949 of the brain and where they are and how they're organized and how they interact. 00:53:10.689 --> 00:53:13.629 But if you would have to give this a number, would you say, look, 00:53:13.769 --> 00:53:19.889 20% is reasonably well explored, so uncertainty is low for these areas, 00:53:19.969 --> 00:53:21.809 let's say primary visual, primary motor. 00:53:22.289 --> 00:53:25.949 There we know what we're talking about. Association areas is, 00:53:26.009 --> 00:53:28.029 let's say, synonymous with saying we have no clue. 00:53:28.929 --> 00:53:37.089 So out of the proposed about 100 areas in a macaque brain, I would say we probably 00:53:37.089 --> 00:53:43.069 have a good understanding of about 20, so that'd be about 20% in that brain. 00:53:43.689 --> 00:53:50.029 But it depends on what you mean. I would say that you can find good evidence 00:53:50.029 --> 00:53:52.009 for a functionally distinct region, 00:53:52.329 --> 00:53:59.229 but to define it precisely in terms of boundaries becomes a real problem. 00:53:59.449 --> 00:54:03.109 So if we talked about VIP today in the talk today. 00:54:05.680 --> 00:54:09.040 People are very uncertain about what you're talking about. You know roughly 00:54:09.040 --> 00:54:12.840 where you are, and you say, if I'm in this region, it must be VIP. 00:54:14.820 --> 00:54:20.300 But it's hard to pin down exactly where there he is, what the boundaries are. 00:54:20.420 --> 00:54:25.540 Are the boundaries variable across different individuals and so on? 00:54:25.800 --> 00:54:30.880 So I'd say that's a subdivision of the brain that's reasonably well understood, 00:54:31.620 --> 00:54:36.160 but it's not understood to the extent that you can say, I know I'm in it or 00:54:36.160 --> 00:54:39.380 I don't, I'm not sure whether I'm in it, those kind of questions. 00:54:39.640 --> 00:54:44.720 You don't know what all the connections are because we don't know how to define the area precisely. 00:54:45.240 --> 00:54:51.640 Right. But then in some sense, also in your recent work, I think this also illustrated 00:54:51.640 --> 00:54:54.020 maybe this issue a little bit. 00:54:54.020 --> 00:54:58.800 Because if you look at motor areas, then in some sense, the standard knowledge 00:54:58.800 --> 00:55:03.940 would be that you have a fairly abstract representation of, let's say, 00:55:03.940 --> 00:55:06.480 the direction of movements, acceleration of movements. 00:55:06.560 --> 00:55:11.640 We're close to the kinematics of the whole of the body that we're then controlling. 00:55:12.340 --> 00:55:15.940 But in your recent work, you showed that if you stimulate in these motorized 00:55:15.940 --> 00:55:18.720 areas, you actually can get whole coordinated movement pattern. 00:55:19.980 --> 00:55:25.700 So how do you then link or compare that to our standard understanding of these 00:55:25.700 --> 00:55:26.920 areas we thought we knew? 00:55:27.340 --> 00:55:31.420 We said, okay, there's something like a population response that gives you some kinematic control. 00:55:31.880 --> 00:55:34.400 But now it seems that in your recent results, you're saying, 00:55:34.460 --> 00:55:41.220 look, actually, we're talking about highly coordinated control of stereotype behavioral patterns. 00:55:42.120 --> 00:55:43.720 So where are we going with that? 00:55:44.930 --> 00:55:49.130 Well, motor cortex, even primary motor cortex, is especially interesting because 00:55:49.130 --> 00:55:54.630 the early maps didn't quite give the details of the way that they're organized. 00:55:54.730 --> 00:55:59.250 So you'll have a hand area, you'll have a foot area, a face area, tongue area even. 00:56:00.350 --> 00:56:02.430 But within the hand area, you'll 00:56:02.430 --> 00:56:07.630 get multiple representation of the digits movements, the same digits. 00:56:07.870 --> 00:56:12.610 You'll get a digit movement that the very next site where you stimulate it next 00:56:12.610 --> 00:56:19.290 to it, you'll get a wrist movement, or you might even get a movement at the elbow, or you might get, 00:56:20.150 --> 00:56:24.470 and the same digit movement would be paired with other digit movements or just 00:56:24.470 --> 00:56:29.670 that single digit movement, but that would be repeated over a large region of cortex. 00:56:29.750 --> 00:56:34.930 You would find the same movement again, and this seems to be different than 00:56:34.930 --> 00:56:36.610 some of the sensory representations. 00:56:36.870 --> 00:56:38.550 Why are you repeating these things. 00:56:39.110 --> 00:56:45.830 And you can say, well, it's important because any particular movement might 00:56:45.830 --> 00:56:49.870 be conjoined with any other kind of movement, and you want to have them close and interconnected. 00:56:50.730 --> 00:56:55.830 I think now that the long-term stimulation with showing more purposeful, 00:56:56.390 --> 00:57:02.330 behaviors or movements can be elicited from primary motor cortex or premotor 00:57:02.330 --> 00:57:07.450 cortex context is starting to give another perspective on this issue. 00:57:08.530 --> 00:57:14.230 Part of a cortical area of M1, of the hand area, will be involved in some kind 00:57:14.230 --> 00:57:15.810 of behavior with the hand. 00:57:15.930 --> 00:57:20.190 Another part will be involved in another kind of behavior with the hand and 00:57:20.190 --> 00:57:24.010 joined with arm movements and so on, depending on what it is. 00:57:24.370 --> 00:57:29.330 And so this adds a complexity as if parts of the area are. 00:57:31.502 --> 00:57:35.942 Are working somewhat independently from other parts of the hand area. 00:57:36.082 --> 00:57:39.282 You're not involving the whole hand area in one task. 00:57:39.602 --> 00:57:44.382 You're involving part of it, and other parts will be involved in other tasks. 00:57:44.582 --> 00:57:49.842 That's what seems to be suggested by these kind of stimulation experiments. 00:57:50.502 --> 00:57:55.282 So it's more complicated than it seemed originally. 00:57:55.562 --> 00:57:58.582 So are you saying with that, that in these motor areas, 00:57:58.582 --> 00:58:06.962 You would have like a library of discrete behaviors that are then sort of biased 00:58:06.962 --> 00:58:11.382 with respect to the limbs that are being involved in these behaviors. 00:58:11.742 --> 00:58:13.302 Should I look at it like that? 00:58:13.922 --> 00:58:19.742 I think you can look at it that there's an organization, a gross organization 00:58:19.742 --> 00:58:22.482 of foot to tongue from medial to lateral. 00:58:23.662 --> 00:58:28.902 And then within that, there's a more complex organization. but that organization 00:58:28.902 --> 00:58:35.002 is similar across individuals of the same species and even across members of 00:58:35.002 --> 00:58:36.922 the same family, different species. 00:58:37.642 --> 00:58:42.262 So you have this preserved, and the reason it's preserved must be that it in 00:58:42.262 --> 00:58:47.702 some way is specified during development to come out in a particular way every time. 00:58:48.442 --> 00:58:55.522 On top of that, you must have a tremendous amount of being able to modify these 00:58:55.522 --> 00:59:01.842 circuits by experience and training so that you not only want to be able to 00:59:01.842 --> 00:59:06.262 do some tasks that every animal has to do, every person has to pick up something, 00:59:06.582 --> 00:59:14.442 but you want to be able to develop specific skills that for a human may be making 00:59:14.442 --> 00:59:17.802 pottery or something like you practice and you get good at it, 00:59:17.822 --> 00:59:20.482 but it could be anything that you practice and get good at. 00:59:21.542 --> 00:59:27.782 For many species Species I practice in modifying the system may not be that important. 00:59:27.922 --> 00:59:35.162 If your average lifespan is eight months, you don't want to spend a lot of time modifying it. 00:59:35.202 --> 00:59:39.762 But for a long-lived species, this modification can be very important. 00:59:40.022 --> 00:59:47.222 So our motor cortex part of the motor system is greatly expanded. banded. 00:59:47.362 --> 00:59:52.962 It's got all these premotor areas and cingulate motor areas in it. 00:59:53.002 --> 00:59:55.422 So there's a tremendous amount of cortex involved. 00:59:55.682 --> 00:59:59.682 And exactly what this means isn't so clear, but. 01:00:00.640 --> 01:00:05.500 There's a big investment in motor control at the cortical level that interacts 01:00:05.500 --> 01:00:10.560 with posterior parietal cortex, frontal cortex, sensory inputs of various sorts. 01:00:10.940 --> 01:00:15.060 Right. But do you see these behavioral primitives now? So we're not talking motor primitives. 01:00:15.140 --> 01:00:20.100 We're talking behavioral primitives matching the ones that are represented and 01:00:20.100 --> 01:00:24.560 controlled at subcortical levels, like in areas like the central grave, for instance. 01:00:24.680 --> 01:00:28.360 Do you see some matching there between these primitives? So that means the motor 01:00:28.360 --> 01:00:32.940 cortex provides you an interface to the subcortical areas on which you can then 01:00:32.940 --> 01:00:35.940 start to play using, let's say, your planning and learning mechanisms you have 01:00:35.940 --> 01:00:37.820 at cortex. Or should I look at this differently? 01:00:38.640 --> 01:00:45.160 No, I think you're looking at it in a good way because it has been unpopular 01:00:45.160 --> 01:00:50.900 to talk about primitives or built-in sort of things in a human brain, for example, 01:00:50.980 --> 01:00:56.720 and say it's all learning or almost all learning. 01:00:57.960 --> 01:01:02.040 But then if you start looking at mammals in general, you see that there's so 01:01:02.040 --> 01:01:07.600 many things that would be too costly to learn, 01:01:07.720 --> 01:01:11.820 costly in that you wouldn't live long enough to learn them if you didn't know 01:01:11.820 --> 01:01:18.140 how to run away or escape or recognize a dangerous situation or to find your mother or whatever. 01:01:18.380 --> 01:01:21.680 There could be so many different things. There might be very rapid learning 01:01:21.680 --> 01:01:23.400 in the kinds of imprinting, 01:01:23.400 --> 01:01:33.740 first many things that would allow a pre-adapted system to rapidly congeal into 01:01:33.740 --> 01:01:38.020 eliciting a kind of behavior that's hard to rule that out. 01:01:39.600 --> 01:01:46.860 But then you have a whole system of levels of, say, a sensory motor, 01:01:47.910 --> 01:01:57.750 system that would have to be working in concert, pre-adapted for certain primitives, modifiable, 01:01:58.490 --> 01:02:05.210 hopefully in some ways, and on top of that then completely learned and modifiable 01:02:05.210 --> 01:02:07.030 sorts of motor behaviors that 01:02:07.030 --> 01:02:11.290 would be unique to the individual if they bothered to learn them. Right. 01:02:11.530 --> 01:02:18.470 So, John, to get to the finish line with our discussion, I have two questions. 01:02:19.030 --> 01:02:23.710 So, as you already indicated yourself, you got awarded this free subscription 01:02:23.710 --> 01:02:28.490 now to science because you have been paying enough money to this. 01:02:28.490 --> 01:02:30.590 I haven't gotten the first issue free yet, though. 01:02:31.210 --> 01:02:35.730 Exactly. We'll see. So, based on this broad experience you have in the study 01:02:35.730 --> 01:02:37.790 of brains and this comparative study of brains, 01:02:37.950 --> 01:02:46.910 what would you now see as the law of John Kass in our attempts to understand the brain? A law? Yeah. 01:02:47.970 --> 01:02:51.510 I'm not sure what a law is. We 01:02:51.510 --> 01:02:55.870 used to talk about laws and science quite a bit, but not so much anymore. 01:02:56.750 --> 01:03:00.510 So, this is your chance. Yeah, so I could make a law. 01:03:01.310 --> 01:03:05.130 John Ullman and I, when we started defining visual areas, thought, 01:03:05.270 --> 01:03:09.610 maybe they'll name a visual area after us sometime, but it never happened. 01:03:09.850 --> 01:03:13.170 Or an asteroid, maybe. So now is the chance for a law. Yeah, exactly. 01:03:13.170 --> 01:03:20.770 I don't think I can think of something that would fit the law. 01:03:20.770 --> 01:03:28.970 What I would like to take from other people that have made studies in development 01:03:28.970 --> 01:03:36.030 is that one level of the system will specify what the next level, 01:03:36.210 --> 01:03:38.070 a lot of its organization, 01:03:38.330 --> 01:03:45.130 so that you have this chain naturally that starts with sensory inputs and modifies 01:03:45.130 --> 01:03:48.490 the system to adjust to whatever the sensory inputs happen to be. 01:03:48.490 --> 01:03:53.030 This principle probably works anywhere in the nervous system. 01:03:53.090 --> 01:03:59.270 You make a change at any level in a complex network and in development, 01:03:59.310 --> 01:04:02.830 and the rest of the system will develop to accommodate that. 01:04:02.890 --> 01:04:06.730 And this makes brain evolution and change... 01:04:08.283 --> 01:04:13.403 Fantastically more easy, because you can imagine if it's, in some sense, 01:04:13.423 --> 01:04:14.883 based on genetic change, 01:04:15.983 --> 01:04:22.343 that you wouldn't want something that you had to change the genes at 10 different 01:04:22.343 --> 01:04:26.283 places, or the gene expression at 10 different places to get what you wanted. 01:04:26.383 --> 01:04:33.463 You would like to be able to make a change one place by some fortuitous accident of genetics, 01:04:33.663 --> 01:04:38.483 and that dies out because it didn't work well or it's propagated, 01:04:38.563 --> 01:04:42.563 but for it to work at all, it has to propagate itself through the whole system. 01:04:43.263 --> 01:04:47.923 So the John Kars law would be that when we look at the brain, 01:04:48.723 --> 01:04:53.963 its power is actually its ability to adjust to many different kinds of sensory 01:04:53.963 --> 01:04:57.643 inputs, also across different morphologies, right? 01:04:57.743 --> 01:05:04.883 So it's very much a very general design of many possible brains as opposed to a single one. 01:05:04.883 --> 01:05:11.023 And that the driving force is really how am I changing the inputs to this hyper, 01:05:11.723 --> 01:05:17.523 plastic control system I don't know how we can summarize it in three words. 01:05:19.723 --> 01:05:22.443 So then the second question would be, 01:05:23.383 --> 01:05:27.043 if we're going to get you back here five years from now and we do a similar 01:05:27.043 --> 01:05:31.603 interview I want to ask you okay did your prediction work out or not so what's 01:05:31.603 --> 01:05:36.783 the one prediction you would like to make today day that you're most enthusiastic about, 01:05:36.903 --> 01:05:38.803 that you have the strongest confidence in? 01:05:40.703 --> 01:05:45.763 I think that we're going to have, at the cellular level, a much better understanding 01:05:45.763 --> 01:05:47.303 of a whole range of brains. 01:05:47.503 --> 01:05:53.043 And one of the surprises that's coming out of looking at just counting neurons 01:05:53.043 --> 01:05:57.903 is a whole change from views that were made 20 years ago. 01:05:57.903 --> 01:06:05.083 And the view then was that all cortical areas are basically the same in the 01:06:05.083 --> 01:06:06.943 sense that their hardware is the same. 01:06:07.623 --> 01:06:11.163 Their neurons are the same. If you run a pin down through cortex, 01:06:11.303 --> 01:06:14.703 you'll count the same number of neurons no matter where you are and so on. 01:06:14.783 --> 01:06:19.943 The variability is turning out to be really tremendous, and it's possible to 01:06:19.943 --> 01:06:22.763 look at different aspects of that variability now. 01:06:23.243 --> 01:06:26.563 And one thing that's true that's going back to the early studies, 01:06:26.563 --> 01:06:32.603 They said that the one exception is primary visual cortex in primates, 01:06:32.623 --> 01:06:41.823 that the number of neurons is twice as many per unit volume of tissue than other places. It turns out. 01:06:43.635 --> 01:06:48.455 That statement is absolutely true for humans and for monkeys. 01:06:49.195 --> 01:06:52.375 It's less true for prosimian primates. 01:06:52.475 --> 01:06:55.855 It's less true for new-world monkeys and old-world monkeys and so on. 01:06:55.895 --> 01:07:01.815 So it's variable for primary visual cortex, but it's also variable in many other areas. 01:07:03.175 --> 01:07:09.555 And we don't know completely for humans yet, but macaque monkeys have variable 01:07:09.555 --> 01:07:13.995 numbers of neurons or neuron densities for different cortical areas. 01:07:14.155 --> 01:07:20.215 So they're specialized by changing the density of neurons, which implies changing 01:07:20.215 --> 01:07:26.035 the sizes of neurons, bigger or smaller, because that affects the density that you can pack them. 01:07:26.335 --> 01:07:33.495 And that wasn't appreciated, and it's just starting to become known as a big 01:07:33.495 --> 01:07:35.815 variable. But this variable is very... 01:07:36.755 --> 01:07:40.715 It is a major variable in macaque monkeys. It's going to turn out, 01:07:40.815 --> 01:07:44.135 I predict, to be a major variable in the human brain. 01:07:44.255 --> 01:07:47.655 And it's not going to be a major variable for most mammals. 01:07:48.115 --> 01:07:52.515 The neuron densities and neuron shapes and neuron functions are going to be 01:07:52.515 --> 01:07:56.655 more standard for most mammals, including prosimian primates. 01:07:56.935 --> 01:08:00.095 Okay, great. Well, John Goss, thank you very much for this conversation. 01:08:00.615 --> 01:08:02.935 Okay. That was great, John. Thank you. 01:08:04.755 --> 01:08:10.575 The CSN podcast was produced by the Convergent Science Network of Biometrics 01:08:10.575 --> 01:08:17.315 and Biohybrid Systems, a project funded by the European Sevens Research Framework Programme. 01:08:18.555 --> 01:08:23.855 For more interviews, recorded lectures or upcoming conferences in the field 01:08:23.855 --> 01:08:30.095 of biometrics and biohybrid systems, go to csnnetwork.eu. 01:08:30.415 --> 01:08:33.615 And thank you for listening. Thank you. 01:08:30.960 --> 01:08:38.000 Music.