Well this is just an incredible pleasure to be here, and I feel very honored to speak in the context of these incredibly interesting speakers, and I hope that I can do justice to this topic. I work in a realm of inquiry, which is not exactly art, but I am very interested in trying to understand how the brain perceives the world, and how we actually create our internal reality. And what I will do is show you a couple of examples of work that we do in a technology that we call functional magnetic resonance imaging which is a modified version of what’s used in a lot of the clinical scans that people do, and with this tool are able to go and look inside the brains, and to a certain extent inside the minds of people who are performing various tasks, many of which are very interesting. Here, I am just going to talk about three general areas.
1) How can we actually look at the organization of the human brain? And I really mean by the organization in terms of how cognition is organized within the brain.
2) I am going to concentrate a little bit on specific information we have on how our brain organizes the process of seeing, and then
3) I’d like to put forward a relatively simple thesis that what we see is what we know, and I am going to try and show you some rather physical examples that might actually try and instantiate that claim.
So, how can we look at the organization of the brain? In fact, I’ve been on this particular task for quite some time, and somebody quoted me in “Science,” and this is something that I have to live with. It says: “Cohen has said he has absolutely no doubt that functional M.R.I. will make a major impact as a clinical tool . . . but he added what we’re trying to do here is optimize the system so we can use it as a neuroscience research tool. In effect what we are trying to develop a mind reader.” And it’s not as flip as all that, so hopefully I’ll be able to show you a little bit of that.
Well, our kind of self-concept of how our minds work probably looks more like the Cartesian view of four centuries ago than the neurophysiological reality that we are starting to experience. This is a classic anatomical drawing from Vesalius, the great anatomist, who looked at this squishy organ inside of our head and figuring that basically what it did was cool the blood. There was in fact a very interesting centrally located structure called the pineal gland, I don’t know if you can see my little pointer over there, which was probably the anatomic, and therefore most important, center of the brain.
Descartes, working another hundred years later, inferred that the brain was in fact the organ which one way or another controlled the body. You can find this out by, for example cutting people’s heads off, at which point their bodies stopped working, this kind of thing, but the kinds of experiments that were available were fairly crude. What Descartes realized was that there was a strong connection, a physical connection, from the eyes to the brain, and he kind of mapped anatomically into the brain, and he saw that the eyes were connected pretty closely to the pineal gland. So this is a drawing in Descartes’ view about how your visual system might interact with your motor control system in the world. Here this gal is looking at an arrow, you can’t really follow these lines in here, but he understood enough optics at the time to understand that the arrow is upside down in her eyes. And the eyes were thought to secrete some kind of a chemical on to the pineal gland. And the pineal gland was an interesting structure because that was where God was able actually to interact with the corporeal self. And at that point there was some sort of a somewhat magical interaction which physically involved in the pineal gland secreting some sort of substance through these vessels, which we call nerves, which would actually act on muscles and make the muscles move.
A surprising amount of that information is actually true, but the important part about this, from this context is that the mind was entirely in the spirit world. It impacted on the brain, but it didn’t really live within the brain the way we commonly think of it. In fact, I think that like most of us, if I look inside my own head and try and imagine what it’s like to be in there, it’s like a little me inside of my body, and it somehow seems that if I took it apart, the little me would just get smaller and smaller, but just be there. But that’s ludicrous; ludicrous at least if you are in the western science tradition I have to remind myself after the last talk. But it is not consistent with the facts we have at hand from experiments.
We are right now in the early 2000’s very heavily influenced by the work of Franz Gall, who was the great father of phrenology. Phrenology is now in the waste can of ideas, but Gall figured that the brain was actually a much more complex organ, and that in fact there was some sort of regional organization to the brain, and that if you looked at some part of the brain being involved in some function or other, perhaps hypertrophy of that region – excess growth of that brain – would express itself as bumps on your head. So it was a simple matter of going to people’s heads, palpating them, finding out where they were large or small and you get various information about what kind of a person they were, and in fact this kind of phrenology was used as a criminal investigation technology because there were criminal body types. The “psycograph” was an instrument designed to bring technical accuracy to the field. These days we’ve seen icons of these little heads in hundreds of examples of what part of the brain does what, but my job in the UCLA Brain Mapping Center is to actually try and resolve what those little bits and pieces of the brain do to experiments.
So, how do we do this? Well, the great psychologist, William James, some hundred years ago said,
“We must suppose a very delicate adjustment whereby the circulation, (that is, the blood) follows the needs of the cerebral activity. Blood very likely may rush to each region of the cortex according as it is most active, but of this we know nothing.”
Interestingly what James claimed was that there should be some sort of observable coupling between blood flow in each region of the brain and what part of the brain is doing something. By saying he knew nothing, what he said is that there were no means of measuring this at the time. It turns out we can do that now, we can measure what the blood is doing. This very complicated and intimidating-looking slide is just meant to remind me to tell you how this works – I’m not going to dissect the slide. What happens is that if any given part of your brain becomes more active, it demands more blood, and on certain kinds of Magnetic Resonance Imaging acquisitions it’s possible to have the amount of blood flow in a region reveal itself, and it shows up as a very, very small signal change in certain kinds of imaging acquisitions.
This in fact is an example; this is one of the very first examples. This is an experiment we did back in 1990. So this is a brain, it happens to be my brain. This is what we call the occipital lobe that, as you may know, has long known as the region involved with vision in your brain. And this is going to be shown as a movie. It’s a series of pictures. Right now: lights are turned on; lights are turned off; lights are turned on; lights are turned off. You probably can’t see much change, but if you concentrate on this little region in here, you will see that the signal fluctuates a little bit. Well, this is hard to look at, so we came up with another way of looking at it. This is the same series of pictures, but we’re just subtracting off the first picture from everything, so we’re only seeing differences in time. And now what you can see is this effect very prominently. Lights are turning on. Lights are turning off. Lights are turning on. Lights are turning off. Okay, does everybody see that?
So this is in fact just picture taking in the brain showing what part of the brain is involved, and in this case, doing something very exciting, looking at blinking lights. We have chosen in this field to express results such as these as statistics – because we all know that statistics is truth. And so truth is measured here in correlation with numbers from 0.4 to 0.88, which means very true. 0.4 means we would usually believe it. And these are the regions of my brain involved in looking at these lights. Okay, now looking at lights isn’t really the most interesting thing that we can do, but it’s a proof of concept. So, I believe I have shown you, at least briefly, how we might actually go about looking at how the brain works
Let’s look now at one little content area: How does the brain organize seeing? This is a stimulus we used in an experiment. This doesn’t stand up in an art context. Here we have people looking at just gradients, “contrast gradients”, and we can also vary the contrast between them, or we also have the opportunity to show them gradients which do this little back and forth motion. So the difference obviously is just whether these things are moving or not, but the stimulus is essentially the same. This is yet another cross-section through brains, I’ll be showing you a few brains today, showing a couple of regions who have nominated themselves statistically as being involved in these tasks. But the graphs here are really the more interesting point.
Let me try and walk you through this a little bit. These graphs show us how bright the signal is in the Magnetic Resonance picture over time, in this case over about six minutes, and in the periods shown in black here, which are each 40 seconds long, the subject is sitting in darkness. In the periods shown with these vertical bars here, they are looking at the same contrast gradients I showed you, but we’re varying them by contrast. So here is very low contrast and here is very high contrast. I am looking at two different regions, which have been given the names “V1” and “MT” for historical reasons that nobody needs to know here. What you can see here is that this area of V1, shown in yellow, prefers high contrast gradients. Its signal changes a whole lot with high contrast gradients and not very much with low contrast gradients. Whereas this region here, we’re calling “MT,” doesn’t care about contrast, but if you turn the experiment around and say let’s looks at medium contrast moving gradients versus stationary gradients, the story becomes very different here. So on the right here, is area MT responding to moving gradients and not responding to stationary gradients.
What this class of experiments has told us is that when you look at something as trivial as these moving bars, your brain separates it into components. There are different parts of your brain that are going to look at the “barness” of this and the “movingness” of this, but you don’t see it that way, obviously. I mean, you can talk about motion separately, but the entire percept is fully bound. In fact, in the human brain, we know that a simple visual stimulus is fractured into a huge variety of separate components. Some of which I’ve labeled here, most of which we probably don’t even know yet. But there are different regions of the brain that care about how fast something is moving, what direction it’s moving, where it is in the world, what it’s texture is, what kind of an object it is. Is it a line? Is it a box? Is it your best friend? What color it is. In fact, position and depth and left and right are all separately processed in different regions of the brain. So we actually separate the world out into these different pieces.
So what happens when we look at art? This is actually a relatively famous painting called “Enigma,” and it probably doesn’t look as quite as good for the panel sitting on the right-hand side, but for you in the front of the room it should have a kind of dynamic quality to it. It sort of quivers and moves, and the little rings around it go back and forth. I think that you can probably trust me that this is stationary piece of art. In fact, it’s a painting. It’s not a computer trick that is causing it move; it’s a brain trick that is causing it to move. This effect has actually been studied using the kind of imaging tools that I have just shown you. And here’s the punch line; these are two sets of pictures of a person looking at this same piece of art. Oh, and I should tell you that if you let these little radial lines connect even occasionally through the blue circles, the motion illusion stops on this.
So it’s very easy to make up a minor variant of this piece of art that gets rid of the dynamic quality. So, these are pictures of somebody looking at the original “Enigma” piece. And these are pictures of somebody looking at the original with the radial lines projected through and then subtracted from the top one. It’s a slightly complex piece of experimentation. But what happens is that this region here, which I called “V1” in my other slide, which is responsible for just looking at the basic contrast features, is shared in common when you look at “Enigma” moving and “Enigma” not moving. This region over here, which is shown with these little yellow arrows, only appears activated in these Positron Emission Tompgraphy scans when the thing appears to be moving. So what’s happened is this stationary object has started part of people’s brains working that is involved in the processing of physical motion. So your brain is actually formed this motion from whatever is the illusion in this piece of art.
My own research, which I’ll show you a little of here, has been not only in developing these instruments, but I’m trying to understand a little bit of what we call “visual mental imagery,” and it’s coming more and more to the thesis that what we see fundamentally is what we know. Let me try and tell you a little bit about that. We did a relatively simple experiment some years ago where we had our volunteer subjects look at figures like these and they answer a simple question. They ask, “Is this figure the same as this figure, only rotated?” Now we have a bunch of very visual people here. I suspect most people here are able to solve this problem one way or the other. And the trick is to take the object in one position and rotate it into another position and see if it’s a left hand rotated into some funny shape, or it’s a right hand that can never be rotated into a left hand. And we simply compared people doing that to doing a very trivial version where these are obviously very different figures. The experiment resulted in data that looked like this. And without walking you through all of the little bits and pieces of it, these regions right here and right here are our old friends “MT,” the part of the brain involved in processing motion. There is no motion there, once again, there is not even the illusion of motion here, but in order to perform this transformation, our brain asks the world to move, as it were, and we actually form that motion perception.
In a similar vein, an experiment we’ve just completed, we asked people to do a very simple task. Start in Arizona – for those of you who are United States citizens it’s a little easier – start in Arizona and simply name the names of the United States going clockwise, without the screen. It seems that everybody does this task by looking at the ceiling. I don’t know why the ceiling is the place to look for it, I don’t have an answer for that, but when we ask people how they do it, universally the response is, “I form a map of the United States and I read the name of the states. Or at least I name them by their shape, because I can do that.” Well, that’s a pretty neat trick, because of course that map doesn’t exist in any obvious way inside of our brains, so we ask how is that we can form those relevant and apparently stable percepts from the internal organ. These were the kind of results we get these days.
This is a picture of a brain, and I’m just going to let this movie go as we do this. These areas, which are showing up in hot colors, are just hot statistics. But these are the network of brain regions, that people evoke in trying to perform our little geographic imagery task. And, importantly, we have these areas right in here, called the lateral geniculate nuclei that are actually part of the early visual areas involved in just seeing things. So, unlike that Cartesian view of the world where the pineal gland receives the spirit world in order to get inside of your head, what our brain actually does is start becoming active in the visual parts of the brain when we are thinking about visual things.
And what I would claim that this class of experiments is telling us is, first of all, what now is a moderately stable and easy to accept finding, that in order to form internal percepts – to construct our imagined visual world – our minds must use our brain’s organs of perception. In order to think about a house, we have to actually use the house visual part of our brain in order to see that. I think that the more interesting problem is that we don’t have really a way of thinking about the things that we can’t see; because we don’t have an organ for it; because we are limited in our percepts by what our organ is able to see. This is a very profound and very troubling limitation on human imagination, which I think is something that we kind of need to explore.
So, I think I m a just going to finish up here. These are just blah, blah, blah. I’ll skip that and talk about where we are going with this. Right now in support of this mind reading instrument, we are actually building technologies; in fact these are mostly finished right now, that allow us to really look in a fairly dynamic way at what the brain and mind are doing. This is what a person looks like when they are lying inside of our imaging device. We present them stimuli through little video game goggles. We take pictures, M.R.I. pictures, in a series, as I showed you before. We then present them to a computer for analysis, at the same time we keep track of what subjects are doing. These days I’ve also worked on looking at brain waves or, “E.E.G.”, which we integrated into this whole process, and we do some statistical analysis that compares the pictures with all of these other events that are going on we end of with statistical maps that indicate regions of the brains involved in things. But, the way this is drawn it should be relatively obvious that this little stimulus presentation device is the very same computer, which is actually forming the pictures of the brain. So it’s actually very possible to sit inside this, look at your own brain while you’re doing whatever you’re doing, and you can ask questions about “What does it feel like when that part of my brain is active?” “What does it feel like not you know in my hands, but what’s my emotional state? Or what’s my anxiety state?” As Dave Barry might say, I am not making this up. In fact, the very concept of the “Autocerebroscope” was put in print by the greaet philosopher Feigl many years before we could build one. Now, however, we are beginning to be able to work with these 21st century science fiction/brain reading machines.