Picower Institute for Learning and Memory Inaugural Symposium: “The Future of the Brain," pt. 4
[MUSIC PLAYING]
PRESENTER: I want to welcome you all back to our final afternoon session. And as the title suggests, Expand Your Mind-- Getting a Grasp on Consciousness, the session should be mind boggling to coin a phrase. I mean, we've been talking about consciousness. We've been dancing around a little bit all morning and a little bit of the afternoon. I once asked half a dozen psychiatrists, psychologists what consciousness was. They all defined it in their own way. They said there really is no definition at this point. So we'll try to answer that question and get some different viewpoints on it.
And there are going to be short presentations. And at the end of each presentation, we'll have a little bit of a sort of a different kind of panel discussion and open the questions right to the audience. Our first speaker studies consciousness by altering it. Dr. Alexander Shulgin is a chemist and an author. He's also known as Dr. X.
The New York Times calls Dr. Shulgin a one-man, psycho-pharmacological research sector.
[LAUGHTER]
Timothy Leary called him one of the century's most important scientists. By Shulgin's own account, he has created nearly 200 psychedelic compounds-- among them stimulants, depressants, aphrodisiacs, and pathogens, convulsants, drugs that alter hearing, drugs that slow one's sense of time-- I'm reading from the New York Times article drugs that speed it up, drugs that trigger violent outbursts, drugs that deaden emotion-- in short, a veritable lexicon of tactile and emotional experience.
Many of these drugs, perhaps all of them, he has tested out on himself and his wife, with a few friends included at times. And in addition to inventing the drug Ecstasy, Dr. Shulgin is a consultant and very much in demand for by an eclectic group of clients that include NASA, Bristol Laboratories, NIH, University of California for his view on how to experience consciousness and what consciousness is. Please welcome Dr. Alexander Shulgin.
[APPLAUSE]
Don't tell him I used his name, Alexander.
SHULGIN: Well, it's a great pleasure to be here and a great honor to be here. I'm not a great lover of Micro-- is it MicroPoint?
AUDIENCE: PowerPoint.
SHULGIN: Magic-- PowerPoint. PowerPoint. Had I known that you were going to have a PowerPoint availability here, I probably would not have used it anyway, because every time I've tried using it, about 30% of the time, at least, it fails. It falls apart somehow. It never quite gets to where I want to go.
So I-- rather than put a lot of chemical structures, which would be boring to about 80% of the audience and fascinating to the other 20%, I suspect-- I'll just use my hands and wave my hands as it is appropriate.
[LAUGHTER]
I mean, molecules obviously are rings and chains and nitrogens. And there's no problem about that at all. My interest in the area-- actually this is a nice opportunity. I spent a couple of years of my life up river some, oh, 60 or so years ago-- more than that-- at Harvard, where I had the unfortunate pleasure of having a national scholarship, which got me in there free. And I found that everyone else had parents who had enough money to get them in there and paid their way. And I could find very little rapport with the masses of freshmen that were around me.
So I found it much more pleasurable to go in the Navy and spend three years in the Atlantic in the anti-submarine patrol, which actually gave me a very nice beginning on chemistry, in that one of the books I had with me was a book by Paul Carr, a Swiss chemist written about 1938 or 1940. And it was a complete over-- a statement-- complete statement of the subtleties and the complexities of organic chemistry.
And when you're spending three years in the Atlantic waiting for submarines, you have a lot of time spare.
[LAUGHTER]
And I not only read the book but substantially understood it. And it was a very, very great pleasure to get out of the Navy and back into the university at Berkeley, where I took organic chemistry as my major. And the greatest compliment I had was from a fellow by the name of [INAUDIBLE], who was a lecturer there at-- or professor at chemistry.
And he said, by the way-- he met him in the hall during the-- I guess, the second year of organic chemistry-- we're having a midterm this coming Tuesday. And you can take it if you want to--
[LAUGHS]
--which I thought was quite a compliment, because I was-- the average on the first midterm was something like 62 points out of 100. And I had 100%. And he--
[LAUGHS]
--didn't know exactly why. I mean, I can answer the questions without any problem, because they're all in Paul Carr's book. And I had memorized the book. That's, I think, honest. Anyway, that-- after that, I got into my AB in chemistry at Berkeley, a PhD in biochemistry at Berkeley, got involved in a little laboratory-- there were five of us-- called Bio-Rad Laboratory. It's now a multi-million dollar operation. Had I stayed with them, I would have been a very, very rich millionaire with ulcers at this point.
And I'm very glad I just-- I split the scene when there were still only five of us present, did a little radioactive synthesis in their name, did some postdoctorate work at Berkeley, went to Dow Chemical Company-- the Dow Chemical Company out with a branch of it there in Pittsburgh and near the Bay Area in California.
And it's there they really got initiated into what turned out to be a very important change in my life. I had my first experience with mescaline about 1960-- 1950-- 1960-- about 45 years ago-- 400 milligrams of the sulfate and had a good babysitter. And I had explored a great deal around various psychoactive drugs. This is supposed to be an erotic thing. That is supposed to be an amnesia thing. Each of these had their own little name. I had heard about mescaline, had never tried it.
And that one day, that eight or 10 hour experience, really changed my life for the next half century. I was totally fascinated with a drug that could get in there, allow you to see things you would not normally see, and yet, you knew to be valid. I have a reasonably limited knowledge of colors. Suddenly I saw colors that I had never really appreciated before. I could look at a flower and observe the beauty of a flower, could not open the flower, could not touch it. But I observed the beauty of a flower.
I had memories from childhood that I knew were valid. But I had not thought of them for years. It was a very, very delightful experience. But mainly what impressed me most thoroughly is that that experience was clearly not due-- the contents of that experience were not in that 400 milligrams of the drug. The drug-- what it did-- it catalyzed my mind. It got my mind back into that particular area. So I looked upon these materials as being catalytic, not productive. They do not do what occurs. They allow you to express what is in you that you had not had the ability to get to and express yourself without the help of a material.
So this really caught my fancy. And I said, if this little 400 milligrams of something could be an effective catalyst to reveal back to me what I had done, what I had seen, and such there is a great potential here for medical use-- and that caught me with my little knowledge of chemistry and my intense curiosity as what was going on upstairs in my head and that was revealed by this mescaline experience. I really went into a true new direction of chemistry.
And here is where I guess I kind of have two wave hands. Mescaline is a ring with three mythoxy groups out here. Don't worry what a methoxy group is. Someone near you will probably explain it later-- a carbon, carbon, and a nitrogen-- a very simple molecule.
And I said, you know, if this molecule can be this effective, what other kind of effects could be gotten by similar materials? So the first thing I did was stick a methyl group on down here-- so now I now have an amphetamine compound-- and took it very cautiously. We were talking a lot today already about experiments with mice and with rats and with various animals.
In my own case, the only animal I used was the human animal. I presume this is now a little awkward because of the various national and federal regulations that have come in. But I find that still the human animal is the only one that is really effective in evaluating and comparing these various psychedelic materials. And the work I do is still involved in that direction.
Here's a material that is identical with mescaline. I called it trimethoxyamphetamine-- TMA. And my golly, it was about twice as potent and totally different in its action. With the mescaline, I had this love and sensitivity to a flower that was on my coffee table, where I was living. And under the TMA experience, I got very curious about and tore it apart to see what was inside-- complete change of attitude towards something of precious beauty.
One was an absolutely [INAUDIBLE] sort of a reverence. And the other was one of dissective-- dissecting curiosity. And the activity was twice. So I went ahead and did that-- I put in ethyl, propyl, butyl, amyl-- put all kinds of different groups on that position. That's one of the beauties of things about having a little bit of fun with the art of chemistry is you can put things on and know where they're going and have ways of determining that their chemistry is going correctly. But the real charming thing and the really, to me, exciting thing was the fact that each thing that you came up with was a new material, had never been made before.
So you're looking at a white crystalline solid in a little beaker there. And you've never seen it before. No one in the world has seen this before. Far as you know, no one in the universe has seen this before. It's a new thing you've just made. And it's never seen you before.
So you, in essence, have no dialogue at all. How much do you start with? How much material do you use as a first experiment on a new chemical that's never been tried before by anyone? Well, obviously an amount that's small enough to not have any effect. But how small an amount is that?
And it's a very interesting additional nuance in this relationship that I developed over a period of time that you go with great caution, decide what is an amount that would have no effect, and take 1/1000 of that amount. It doesn't take much-- it takes time. But it doesn't take much more chemicals, because you use 1,000-- up to where you were, you'd use another milligram perhaps.
And so each of these materials had to be learned as an individual new meeting. And one of the outgrowths of that I discovered is that the beauty of your final results of finding out what the effects are-- you really can't be wrong, because you'll say, I found that this material caused a visual enhancement of that and a recall of memories of this and this and yonder. Anyone else who tries it who finds the same results will say, he is right. Anyone who tries it and doesn't get the results says, what did I do wrong? So in essence you come up with a winner--
[LAUGHTER]
--very nicely. Anyway, what I did-- put these on there. The methyl group was twice as active. I put a propyl with no activity at all-- the ethyl mescaline. And by that time, I had made materials up through the, oh, 9 or 10 carbon chains. So I didn't bother trying them. I went back, put stuff on the nitrogen-- your nitrogen atom-- no lost activity entirely, a couple of methyl groups out there-- you can go almost a gram and not get any effects.
Then you have the ring system. Now here you got really exciting-- you have these three mythoxy groups sticking out in the ring. If you can imagine a hexagon being held by a two carbon chain, you have the hexagon out here. You have 1, 2, 3, 4-- you have 5 positions, 3 of them were occupied methoxies. So here's your quiz of the day. How many ways can you put 3 methoxy groups on this 6 membered ring that are different compounds? And the answer is 6. You can have 3, 4, 5, 2, 4, 5, 2, 3 4, 2, 3, 5, 2, 3, 6, 2, 4, 6. If I had a slide, this would be obvious.
Anyway, so I synthesized the other five compounds. And by golly, the 2-4-5 was 10 times as active. 2-4-6 was also very active and very interesting. The other three were absolute duds-- nothing at all. So here suddenly I now know that you can get much more potency and complexity, more stimulation, more eye dilation, but also psychedelic effects with, say, 2-4-5.
So now you have a new material-- TMA-- I called it TMA-2. You have three methoxy groups out here at those positions. Try each of them into an ethoxy-- gives you three more compounds. And only the 4 position was sensitive. So suddenly you have a position out there that gives you more potency. So I put other groups doing it out that way and began realizing that this here structure-- these are all called phenethylamines, by the way. This structure is amenable to amplification complexity increasing if you substitute here but not there.
So that's where I go. We're talking earlier some talk about neurotransmitters. It occurs to me that that position with the methoxy group-- the methoxy group metabolize off easily. What about putting a group out there that won't metabolize often? Instead of methoxy, put a methyl group out there. So I made the compound 2,5-Dimethoxy-4-methylamphetamine.
And I said, it's either going to be much more active, because it can't come off and be metabolized easily-- and hence, or I'll have a more active compound-- or it will not be active at all. But it will go into the neurotransmitter site the psychedelics go into. And if there's anything to the argument that these are neurologically activating sites and may be activated by people who are with mental illness, you may have a therapy for mental illness. You can't lose.
So I made the compound, tried it. And it turned out to be quite a bit more potent yet. This is material called DOM, which that, of course, led to a whole new direction. If you have DOM out there, methyl-- what about ethyl? Active compound-- propyl? Active compound-- and if it's methyls and the propyls and so forth-- active-- put a bromine out there-- active compound. Put an iodine out there-- active compound.
So the thought occurred to me, if you have an alkyl group that's DOB and DOI. DOM is the one that got off into San Francisco into the name of STP and if any of you are young enough to know San Francisco in the '60s. But there was a STP. STP, I should say, was very active at that time.
And it was-- turned out that I found out that it was, indeed, DOM under another name, STP. They said Serenity, Tranquility, and Placidity was the name for it. And no one knew what placidity was. So it became Serenity, Tranquility, and Peace, which was a little bit more understood. To the police authorities who did not like this idea of this going around-- they didn't know what it was-- they called it Too Stupid to Puke, which was--
[LAUGHTER]
--their counterpart to the-- they used to hate Ashbury Clinic. And it was-- at this time, I was up the hill in the medical school. And this was going out there. And I had no idea what STP was-- one of my compounds. I talked at a media-- at a conference back in the east coast, here in the east coast, about a week or two earlier. And I talked about the material and gave it structure. And I suspect it was just synthesized from this seminar I gave.
[LAUGHTER]
Anyway, the bromo-- the--
[LAUGHS]
--it's a funny world. The bromo compound, iodo compound-- it occurred to me maybe it is because this alkyl group was active. And you have what's called a lipophilicity or hydrophobicness, where something likes something that's fatty. And maybe if I put something on there that was water loving like a nitro group, it would not be active when it goes into the neurotransmitter receptor site. [INAUDIBLE] nitro group-- active. Well, maybe it likes both-- it was putting his tail into this receptor site. Going to the right that's lipophilic and to the left is hydrophilic. What about putting a group on that is not philic at all? And they made fluorine.
So I put on-- I think it was a trifluoroethoxy analog. So I felt this would probably not be active at all-- also active. So it's just getting it-- the tail of the four position-- that molecule into the receptor site produced activity. So if that's an obvious step for it to go and make-- take off the methyl group, get away from the amphetamine chain. So I took the methyl group off. And that gave 2C-B, then 2CI-- a host of other materials in the same ilk that was just a beautifully rich collection of compounds-- many of them not as potent as the amphetamines but shorter lived and much more benign and much more friendly than the corresponding amphetamines.
So this is-- oh, another thing-- somewhere along the line, it occurred to me, if oxygen does a good job, put a sulfur on there. And you get now the 2C-T family-- 2C-T-2 up to about 2C-T-25 or so, of which about half of them are active. So this is kind of the handwaving world of synthetic chemistry.
I could go on for another 10, 15 minutes and get into thryptomenes and go through the same complexities. But do you have this as the active position? That is not as active. This is less active. Alkyl groups on thryptomenes are much enhancing in nature and complexity of action-- alkyl groups with the exception of MDMA and a couple of others. Naphthylamines destroys the activity of the phenethylamines. And so there are differences between these two families of compounds. But those differences are not negative. They are just informative.
Anyway, that's kind of the picture for I've been going for a while. I don't want to take too long here. What is, I think, a question that has often come up is how is this all going to work out? What are the goods and the bads of this entire area of psychedelic chemistry? Basically, the negatives are the terms many people from law writers to people in the street feel that this is an area of neurotoxicity, an area-- these materials cause neurological damage, cause people to lose control, commit crimes, and eventually collapse after 20 years of brain decomposition, which is, to a large measure, nonsense. However, I can't say it completely excludes. I've been into it for 45 years. And I'm having my usual expected amount of brain deterioration. But I don't think it's that serious yet. So I hope to have another decade or two of reasonable responses.
And you have the increasing urge to put laws against these things, because the psychology, the propaganda that they are negative, that they do damage is very real and very much believed by many people. Now I've been often asked, why use the word psychedelic? Itself is a pejorative term. I mean, there are empathogens, entheogens, hallucinogens, psychotomimetics-- other terms that are used widely in medicine that carry other messages but do not carry the intrinsic negativeness of the term psychedelic.
Well, my main argument for keeping continue to use the term is that people may not approve of what you're working in or what you're saying. But at least, they know what you're talking about.
[LAUGHS]
You stop nine people on Market Street-- 10 people. Nine people, you say, I work with empathogens when they ask you what you do. They have no idea what you're talking about. Nine people out of 10, when you tell that you're working with psychedelics would not-- maybe not approve. But at least they'd know what you're working with. So the idea of using a term that is in popular usage I consider to be quite positive.
Well, my clock is still going. The clock is supposed to go flashing three minutes from stopping. And the clock has stopped. So I don't know where I am. Let me-- let me wrap things up.
[LAUGHTER]
What are the positives? I consider the positives to be my main incentive for doing the work I've done for the last half century and continuing to do it now is I believe in this collection of materials you're going to develop tools that are going to answer many of the questions that have been brought up today, namely how can you find out how the brain works? You can use a rat. How does the mind work? What does the-- what kind of a probe can you make to look at the function of the mind?
To me, it's going to be a psychedelic material that has very little action in experimental animals to look into actions in man that are not seen in experimental animals. Maybe the idea of using these materials as eventual research tools, I consider to be extremely, extremely valuable.
I think what I'll do-- a point came up during lunch today. It brought up an interesting story that I think pretty well puts this into perspective about the need of tools for exploring research-- research tools for exploring this area of understanding the function of the mind. It was Eric this morning who was talking about animals being invested with the properties of schizophrenia.
And this was some years ago, back in the good old days before there were many inhibitory actions on human studies-- FDA approval, disapproval. You get clearance from the DEA, clearance from everything like that before you do any human experiments. Your board of your university has to see the research and approve of it. A lot of this experimental work was done back in the halcyon days when there were no such things as research approval boards.
I mean, in Berkeley we had the run of the place. We could fire up the cyclotron front and make an isotope and use it and try it. Their argument at Donner Labs-- that was the Donner, then went up on the hill in Lawrence Lab-- was stay if you want and do whatever you want. The tools are here. Here's the cyclotron. Here's your PET scanner. Do whatever you want. But just remember, when you leave, turn off the lights and lock the door. And we could work through the night there doing experiments, all kinds of beautiful things.
I remember one time and this is-- let me use this as sort of a windup. We had-- this was some maybe three or four decades ago. It was quite popular opinion that methionine was involved in schizophrenia, because some experiments had been done in which people who were schizophrenic were given methionine in rich dyads. And their symptoms became worse.
And yet, those people who were not diagnosed as schizophrenics with the methionine rich dyad had no changes at all. So we've talked about this-- pros and cons. And it was a neat experiment. What I did-- I took a-- I remember it was [INAUDIBLE] or some compound in that area. And I tucked on a fluorine-18, which makes it a positron emitter, which means you can go into a PET scanner and put this into a person and put the head of the person-- while the person is attached to the head still I mean. No--
[LAUGHTER]
--that didn't sound good. You had the person lie down on a little cot with the head going into a positron camera. And we've had a section of the brain just above the ear lobes that tells you where that chemical went. Being a positron emitter, it didn't have to have any reaction in the body. It just went where it went. And what we did-- this was work done with Tony some, oh, God, years-- a few decades ago.
I made this material. In fact, I made 10 batches over a period of time. The half life of fluorine is a little less than two hours. So you can't make a lot of it and keep it for a while. And he had good friends up at Mendocino Laboratory-- Mendocino Hospital. And he came back with five names of five schizophrenic patients, who were up at the hospital. And we had their names and the backgrounds of them. And in the Lawrence Lab, I managed to find five normal controls. That was a bit more tricky. But--
[LAUGHTER]
--we did. And 10 batches of this and we did 10 experiments. We put the material into these 10 people about a week apart and in each case, put them into the PET scanner. I remember one of the schizophrenics Tony had a lot of problems with, because he did not like radioactivity. And they said radioactivity is bad.
So we had down at Donner a great big sort of background counters, a big room with a big iodine crystal of 30-some odd inches in diameter and walls that were 3 inches thick lead overhead and the side. And Tony very nicely told him, if you go in here and spend a half an hour-- he'd give you a magazine, turn, and leave the light on-- if you go in here and spend a half an hour in here, your body will be so depleted of radiation that when we take you up on the hill and put you in the [INAUDIBLE], it will bring you back to normal. You'll be OK. He believed it.
[LAUGHTER]
Anyway, to wrap up with the result of the experiment, it was a fascinating thing. We ran 10 studies. And we had 10 photographs of the fluorine-18 disposition in the brain. And the 10 photographs were absolutely different from one another. There was no consistency through this group at all. And so we put them on the wall of the medical radiation thing up on the hill.
And across the back of the wall, every time someone come in from Washington to give a seminar or come in from somewhere of any importance, we say, by the way, here are 10 photographs of the fluorine-18 labeled material we gave. Five of these are schizophrenic patients. And five of them are normals. Which do you think are normals? Which do you think are schizophrenics? And we got absolutely random answers-- no pattern could be found at all.
Then about two months, three months later, one of the schizophrenic patients who liked Tony very much came down to visit and see how everything was going on-- very nice visit. And they were talking for a while. And he saw these 10 photographs on the wall. And he said, are those the 10 pictures you took of us? Tony said, Yeah.
And he looked at one and said, well, I reckon that's me. And he pointed to number seven or one of them over there. And he's absolutely right. He identified his own photograph from the PET scan of the distribution of that fluorine-18 thing. And Tony very mildly, casually said, oh, you know, you're right. Absolutely right-- how do you know? Oh, he said. That-- you see that little sort of star shaped shiny thing in the bottom right corner-- that little star shaped thing? Yeah, he said. I see it all the time.
[LAUGHTER]
So we have a long way to go before we really can understand how the mind works. But this is a start. Thank you very much.
[APPLAUSE]
PRESENTER: That's great. Have a seat. And then we'll call you back up here.
SHULGIN: [INAUDIBLE]
PRESENTER: That was terrific. We'll have a chance to question Dr. Shulgin and the other speakers when we sit down there in a little bit. You weren't the only one experimenting with psychedelic drugs in the '60s. I remember a friend of mine's father worked for the CIA was telling me about a few CIA agents who walked out of a window on LSD. And this was 10 years before it was uncovered in Congressional hearings.
Our next guest, Dr. Christopher Koch-- Christof Koch-- excuse me, says his aim in life is finding out how consciousness arises out of the brain. My long term goal, he says, is to discover and characterize the Neuronal Correlates of Consciousness-- the NCC. I collaborated for 16 years in this exciting endeavor with the late Dr. Francis Crick at Salk, a postdoc refugee from MIT.
Dr. Koch joined Cal Tech's newly started Computation and Neural Systems Option in 1986 as Professor of Computation and Neural Systems running K Lab as well as being executive officer of CNS. Here to tell us more about his views of consciousness is Dr. Christof Koch.
[APPLAUSE]
KOCH: Thank you. All right, so before we start, I have to make a conceptual distinction because different people use consciousness in very different ways. So first of all, particularly what clinicians are interested in their see different states of consciousness. Right now, nobody's asleep, at least yet, so that's one state of conscious. You're awake.
Then, of course, there's REM. There's different types of sleep, REM, and Slow Wave sleep. There's coma. There's Persistent Vegetative Syndrome. There's minimal conscious state.
There's stupor. There are all these other states. And so you can study people, folks, and you can kind of made a contrast between the awake state and the REM state.
So this sort of defines consciousness as such-- the ability of your brain to be conscious at all. So right now, we are conscious, but what most people who study this now are interested in is different states of consciousness. So while you're conscious, you can be conscious right now of my voice. I talk a little bit like my governor. And--
[LAUGHTER]
--or you can be conscious of blue or of something else. So this is the current content of consciousness. And the elements are called qualia. This is what philosophers refer to. So there's a qualia for sound. There's a qua-- you have a qualia for red, for blue, for being angry, for being you, for being late, living in the late, industrial post construction society, whatever.
Now, empirically, studying this aspect, I can show you is much, much easier because this, you can control very well in humans and in some animals. You can control the level of arousal, and then you can manipulate the content, particularly in the individual case. It's much more difficult to do this, for example, for self-consciousness. We don't have any good illusions. We don't have any good ways to manipulate self-consciousness.
For example, I can sort of, without taking drugs, which is a little bit uncontrolled, I can really switch my feeling of being Christof on and off in a time scale off of half a second. But I can do that perfectly well for visual consciousness.
And so here, the trouble is if you do a comparison between the wake and the REM or between the REM and Slow Wave or between Coleman and awake as people do, there are all sorts of different things that change at the same time. It's not only conscience. It's also tensions. It's memory. It's also motor behavior, so it's more difficult to control, experimentally.
What is it we can see for sure at this point in 2005 about the Neuronal Correlates of Consciousness? Well, first of all, we know it's a property of certain complex adaptive networks, not all networks. So you may know down here in your gut you have roughly on the order of 100 million neurons, and It's called enteric nervous system.
They're real neurons. They have synapses, a neurotransmitter. And for the most part, it's totally silent. You don't have conscious access to it, and if you do, usually it's bad what happens, right? The other systems like the immune system, we don't have any conscious access to the immune system, so it's not just any-- it's not just any cellular property. It's a property of particular neuronal networks using massive feedback.
Everybody believes without feedback, if you create, let's say a mouse that has no feedback connection, this mouse wouldn't be conscious. It's, of course, shaped by evolution by natural selection. Many species, certainly from a biological point of view, many species share aspects of conscience with humans. That's not to say that the consciousness of the mouse, of a monkey, of a dog is the same as us. It's certainly not true
But you can infer based on the very similar behavior, based on similar structure of the nervous system, and based on argument of evolutionary continuity, it is very likely that, let's say a dog or mouse or a monkey shares aspect like vision or like smelling that also a mouse. When the mouse, smells something, it's also feels like something to the mouse to smell something. We don't know right now the minimal. We have no idea about brain size, how brain size correlates with consciousness.
We all have the strong feeling that things like C. Elegans, like Drosophila, like a bee. Of course, they're bugs. They're not conscious, but we really don't know. And if you observe a bee can do in conditional experiment in one short learning, a bee has roughly a million neurons. It can do amazing feat of complex information positing and complex visual pattern recognition.
And you're not so sure anymore, non-reflective behavior. So I'm not really totally sure anymore that it doesn't feel like something to be a bee. The trouble is that right now we don't really have good acids to test that. As Eric emphasized this morning, and, of course, as we know now, 400 years of psychophysics and clinical research and physiology, much, if not most, of behavior occurs without consciousness, occurs either totally bypassing consciousness, or consciousness comes after the fact, like giving a talk like this.
Most of us when we talk, we have no idea what we're going to say. You hear these words coming out of my mouth. The only thing I know is sort of in the next two minutes, I want to make this point. I want to make that point, but it's not that I'm consciously Christof is sitting inside here, formulating a sentence, putting the verb and the noun together, translating it from English-- from German into English, and then sending it out to my larynx.
It all happens unconsciously. And we know this for sensory motor behaviors, for visual behaviors. There's lots of evidence for that. So this really gives rise to a contrastive strategy because you can now contrast at the level of brain architecture with the difference between those behaviors, so let's say in humans or in monkeys that give rise to conscious sensation versus those behaviors that do not give rise to conscious-- to conscious sensation.
The different brain areas, how that some brain areas, let's say down the brain in the catacombs of cortex, let's say in the basal ganglia in the parts of the midbrain. Are they never associate with consciousness? It used to be folks only believed in the 19th century, most thought that 20th century, that anything in cortex, if you have vigorous neural activity in cortex, that's always associate with consciousness.
We know that certainly, not two, for example, the experiments of Nico [INAUDIBLE] and the monkey, you can have neurons that fire very, very vigorously to a visual stimulus. However, if the animal does not perceive it because it's perceptual suppressed, the animal isn't-- it doesn't respond to it. The animal doesn't seem to be aware of it. Yet you have lots of neuron focus and visual cortex that fire to it. So that means just because you have some cortical neuron onset fire vigorously, it does not equate with the fact that those neurons, their representational content is made accessible to consciousness.
So what Francis Crick and I have advocated since many, many years-- and what people are doing now-- is look for the minimum neuronal mechanisms that are jointly sufficient for anyone conscious percept. Minimal, because we know the entire brain is sufficient. But we-- for example, in order for you to see this blue, do you need your cerebellum? Well, probably not.
Do you need your retina? Probably not, because you can close your eyes, and you can imagine the blue. Do you need your primal visual cortex? Well, that's much more controversial already. But-- all right, so this brings me to my next point.
[LAUGHTER]
Yes. When I talk to folks in the philosophy department, people ask me, well, how can you study consciousness without properly defining it? Of course, historically, as we all know, in science, progress almost always happens at the time when we don't have good definitions. And defining something too early is usually-- if you define it too prematurely, it can be a bad thing.
So here, this is what I mean by consciously. Well, what you should do now, fixate this course just look at this course here. Look at it very intent. Don't close your eyes. Don't move your eyes.
And then, what do you see? It's called Motion Induced Blindness. Okay, do some of you see this? So look at the course, and what you should see is that one or both of the yellow squares sometimes disappear. All right? You see this?
Okay, so here people-- so this sort of is an operational definition of conscience. When you're conscious of the yellow, you see it. You have sensations. You can see the yellow. It may remind of the yellow sun flower or van Gogh or yellow sun or whatever.
And when you don't see it, it's just not there. We're not conscious of it, but physically, the stimulus is all-the-time present, okay? So A, this sort of belies a common, naive, realistic notion that there's a simple one-to-one mapping between what's outside in the world, what's inside in the primacy of my head.
Because here, the yellow's present all the time, yet sometimes I see it, sometimes I don't. It also shows you in this less than ideal conditions, that we can manipulate now very well the relationship between the physical stimulus and the percept. In the lab, we can do it with a precision much less than a degree of visual angle, and we can do it with the precision of a fraction of a second that we can preci-- we can put something on your retina, but you don't see it unless we want you to see it.
So this is similar to a magician. You go to a magician like David Copperfield, and then he has this beautiful bikini-clad assistant, and she acts as an intentional distractor. Let me tell you, it works very efficient. And while you're looking at the scene in front of you, because you're distracted by her, you will not see things, although they are in front of your eyes.
And you can see he makes a sleight of hand. He makes things disappear. But if you go to the same show two or three times, you can perfectly, well see. So we can now do the same thing under laboratory conditions.
Now, you should continue to look here at the course because now-- next slide-- you should see an after image, right? You see the ghostly, bluish after image corresponding to the two yellow squares? All right, so we can-- so that's an after image. It's one of the oldest techniques in psychology, was invented sometime-- was discovered sometimes in the 17th century.
We can now ask a simple question, do you need to see the yellow square in order to get the after image? Okay, this sounds like a really stupid question, but just to show how we can now manipulate visual consciousness. Do you need to see the yellow square in order to afterwards see an after image, okay? And the way you can test this, you can compare all the times when you saw the yellow continuously, let's say, for two seconds.
Compare that with the-- and you measure these things of the after effect-- compare that with the after image you get when you didn't see the yellow, right? So I get subject to press a button whenever the yellow disappears, and then immediately afterwards, you check the strength of the after image. And it turns out, the after image is totally independent of whether you consciously saw the image or not. It doesn't matter. You still get the same after image.
What this tells us is that-- this is a phenomena, Motion Induced Blindness, has to occur somewhere early on in the system because usually in the retina, we think that after images arise, in this case, in the retina, and that the Neural Correlates of Consciousness of seeing yellow happens higher in cortex somewhere. And so what matters for the after image is just how long the image was physically present on the retina, but it doesn't matter whether you saw it or not.
So it allows you to infer something about the-- and that or both. This is what Bela Julesz called psychoanatomy, because it allows you to infer something about the location of the phenomena in the hierarchy without doing FMI and without doing single cell electrophysiology. You can also do-- and people have-- lots of people have done this. You can also do this to other after effects you can explore. This was done, in fact, here at Harvard.
Well, you have always so-called, the tilt effect, orientation-dependent after effect. This also doesn't depend on visual awareness. This affect do, so we had a-- you can do a very nice study show. There's something called face-specific aftereffect that for example, if you look at a morph of John Kerry and George W, and it's just morphed where half the time, you see George W. Bush and half the time, you see John Kerry, then I show you briefly-- I show you, let's say, for two or three seconds, I show you Bush. Then, if I now show you back the original image that's morphed, image much more likely to see John Kerry than Bush because you've adapted to Bush in something specific having to do with identity of the face.
Well, this, for example, does depend on consciousness if you don't see it because we suppress it. So we put these trace images on your retina, but we suppress them so you don't see them. You don't get the artifact anymore. So this is just a way just to show you one glimpse of what psychologists can now do to control-- dissociated physical stimuli from what you can actually perceive.
Okay, so in 10 years ago, Francis and I wrote us a nature paper where we based on the new anatomy of the macaque monkey for reasons I don't want to go into. We made two predictions. We said that a conscious perception, the particular visual perceptual requires activity in prefrontal cortex. At the time, people were profoundly skeptical of this because it was believed that essentially vision is just done in the back part of the brain.
And then, more relevant for today, is that the neural cortex are conscious are not in primitive visual cortex. In other words, that the visual input projects is from the eyes when the intermediate relay station to the back of your head here, the visual cortex. And that's the first stage called the primary visual cortex.
And we made a sense that our ideas, the support for ideas would be much greater if it was shown that invisible gratings produce activity in V1 neurons, in humans. Because then you can argue, well, you didn't see them consciously yet they still give rise to finding activity in neurons in primary visual cortex. So m that seems to imply that the neurons in V1 do not give rise to conscious visual-- to visual consciousness.
So there's a paper just came out in Nature Neuroscience. I'm just showing you to give you the idea that you can make progress in this very old-- consciousness is one of the oldest part of mankind. Aristotle has written already about it, 2,00 years ago. And we're not forever condemned to just philosophical speculation, but we can actually make progress.
So these people did the following as simple imaging study, and this is in a magnet. So they showed you this grate. Either you saw this grating. You fix it here, and you get this grating or into it this way. Or you fix it here, and you get a grating onto it that way.
And you get a-- in different parts of visual brain, you get a very strong response to either this grating or that grating. Okay, not very surprising. And in fact, what you can do, you can use a decoding algorithm. So this is the of one a subject. So you take his visual brain and flatten it out. I mean, you do it in a computer.
[LAUGHTER]
This is human, after all. And then, you can do what's called decoding. It's a standard technique in information processing. You essentially ask a little mathematical algorithm, which pattern give rise to this orientation? And which pattern gives rise to this orientation?
And then, I'm going to show this machine, this little algorithm I've cleaned up a new image, which the algorithm hasn't seen before and ask it to, based on this what you learned in the previous image is, can you tell me whether the subject saw this grating or this grating? And in this case, you can do it 82% correct, so far above chance. Chance would be one 1 out of 2.
So you can do it 82% correct in primary visual cortex. This is not at all surprising. This is what everybody expected. But now, you can do a really cute trick. You can hide this game using a masking technique.
You show the same stimulus very briefly, just like before, but now you hide it. You put stimuli before and afterwards. This is called visual masking. Essentially, you show an image, and immediately afterwards, you show another image.
So physically, the image is present exactly the same amount of time. It's the same signal energy you put into the signal as before, except subjectively and behaviorally, you didn't see the image anymore.
And we know that because we asked people-- we can ask people, well, in this experiment, each time, tell me whether you think you saw this grating or this grating? And you ask people to do it, and they're at 50.3%, so they're totally chance. So they have no idea what the eye was presented was this grating or this grating.
Yet if you do a decoding algorithm, you can still see that, in this case, so if they're invisible, you can still get information out of the hemodynamic response and primary visual cortex if it's invisible. So what that seems to suggest-- well, what that says, is that you have information that's accessible, that's represented in primary visual cortex, yet it's not accessible to the conscious you.
It's not accessible to the-- if you ask the subject, did you see it? Were you conscious of it? They tell you, no. So they see this sort of evidence in support of the conjecture that some parts of the cortex, in this case, primary visual cortex, they may be necessary for certain forms of vision, but they do not give rise to conscious sensation.
Okay, almost everything we heard today, the electo-- well, all the electrophysiology we heard today was based on monkey data or data in rodents. You can also do these experiments very rarely in humans. And then, you can gain data that again directly speaks to consciousness because you can ask humans what they see.
So this was done with-- this is done it by a neurosurgeon, Itzhak Fried. And the work was done by several people particularly, Rodrigo Quian Quiroga and then two people, who are here at MIT now, Leila Reddy and Gabriel Kreiman. And the setting is, these are patients that have epileptic seizures, and they can't-- they're resistant to pharmacologic treatment. And so you try a new surgical approach, which is very successful.
So what you do, you locate the seizure origin using MRI or behavior and EEG. And then you take out the seizure focus, you take out surgically. Now, in some patients, 30% of the patients, you can't do that because you don't know what the seizure originates, the MRI and the EEG. Doesn't tell you. And then what you do, you implant electrodes into the brain. So you have up to eight or 10 of these macro probes.
Here, you see one in the hippocampus. This is the human hippocampus. And you can then-- these things are monitored. Essentially, it's like intracranial EEG. It's like EEG, except it's inside your cranium.
And these patients are monitored 24/7 for three to five to seven days in the clinic, and you can essentially triangulate and find out where the seizure originates. Now, what Itzhak Fried did, and his collaborators, he hollowed out this electrode, and then he adds nine wires. They're just conventional micro collectors, just like Earl Miller is using when he and monkey or Matt Wilson is using them in them in rats and mice. So now, we have roughly on the order of 100 microwires in these patients.
And then, what we can do, we can-- so here you see one of these patients. He's a conscious human. You can ask him what he's seeing? So here's the turban. And here's the preamplifier here, and then the wires go to the amplifier here and then go into a computer.
And then, so you can show him different images. So here, what we show him images. We ask him what the patient likes. We ask him what movies they like. We ask him what a movie actress, or et cetera they like, or what is he interested in.
So this is something you can't do in a monkey of course, so it's biased. And then, here, we show him-- so we show them all these images. So the image is present. Oh, it's really you can't see it very well. So this is three seconds, and here they're supposed to be-- you can barely see him in sort of vertical bars here.
The two vertical bars mark at one second. And we show the image always between those two vertical bars here, here, and here and here and here. That's why we present this image of the image of the Eiffel Tower, of the image of-- it's an actor, it's called Jennifer Aniston. And what you can see here-- so this is a new one in the medial temporal lobe because that's where most of the seizure happens, so that's most of our electrodes where they are.
It's a single unit in the parahippocampus left. And what you can see here, so this unit, three milliseconds after the image is flashed on the neuron response here. These are six different trials. These are all these six trail to these phase since they all happen to be the same actress, Jennifer Aniston. Jennifer Aniston used to be married to another famous actor called-- I don't know-- and then--
[LAUGHTER]
And then, so here, the neuron doesn't respond, although she is present. But she's also present. And you can do a lot of psychoanalysis here why this neuron doesn't respond.
[LAUGHTER]
The patient knew Jennifer Aniston, personally, and it doesn't respond to any other thing. So here you have another neuron that respond selectively to the Sydney Opera House. You have another new and that responds to another actress called Halle Berry. You can see here, so these neurons are very invariable.
This is a very high-- this is not a purely visual area anymore. It's the end point, sort of if you want, of all the sensory processing. It's a lot of that information gets sent to the hippocampus. And so you have neurons here that respond to Halle Berry, whether it's a photograph or whether she's dressed up in a cat suit or there's a line drawing of her or even the text. So Halle Berry, the neuron also respond very strongly, doesn't respond to other text, doesn't respond when another woman gets dressed up in a Catwoman suit.
[LAUGHTER]
It responds very selectively to Halle Berry. You can also see here Mother Theresa. I mean, there's very little rhyme and reason to this. It's in a different patient-- or Pamela Anderson. Again, the text Pamela Anderson. So here, we have sessions from-- this is roughly 1,000 neurons. And a small fraction of these neurons have these very selective and invariant responses.
Of course, it could be that a very large fraction of neurons respond this way, but we only have typically in session, half an hour before the patient gets bored or gets tired or the nurse comes. So we don't have a lot of time to test for such neurons. So what this test is-- if you're going to remember anything about the symposium tomorrow, okay, if you're going to remember anything about the presentation this morning, it's because that information was visual processed by your, let's say, inferotemporal neurons in IT, and then was sent onto hippocampus.
And so that's the neurons we tap into here. So this sort of suggests it's a very sparse explicit environment representation. And we've done other experiments with Gabriel Carmen where we ask questions that directly pertain to consciousness. Like you can focus on, ask the patient, close your eyes and imagine, for example, the various images I showed the patient. And then, a subset of the neurons ties as selective to either the real image presented on the retina or to the imagined image.
You can do other experiments where you can, again, manipulate. Just like before I showed you, you can manipulate the relationship between having this stimulus present on the retina and actually seeing it. And again, you can see these neurons only respond if the person is conscious of the input.
These are very reminiscent of [INAUDIBLE] hippocampal place cells. In fact, there may be a sort of generalization into a cognitive domain to these cells like Matt Wilson has them. They represent only if the rat goes to this particular part of the physical layout. And you can imagine that these neurons are sort of something similar in a high-level conceptual space. Let's skip this.
You can also do decoding. You can also ask in an objective way. Forgetting about what this person says, you can just ask the following question-- given the pattern of neurons-- and see here? Give that I have 12 neurons to fire to these different images in different ways, can I decode them using a computational algorithm? Can I infer in an objective way independent of me as an observer? What input was present? And I can do that very well.
So to finish, what sort of the long term strategy for this sort of research to study consciousness and its particular, its neural correlates? So you want to do this. And this is now a sort of competitively big activity. You want to study neural correlates in humans using FMI. I showed you EEG, MEG, or single units.
Always paradigms [INAUDIBLE] that associate the percept from the stimulus. So you want to use always these various tricks and there are lots of them in the toolbox of psychologists. And you can also do it for other things like smell and for audition. Partly, this is very interesting clinical, very relevant because there are interesting questions. You want to ask fetuses, I mean, or newborn or people who-- or babies who are born prematurely to what extent are they conscious.
And of course, if you have patients like-- remember, the huge debate this year we had about Terry Schiavo? To what extent are people in-- there are 50,000 people in the US alone who are in PVS state, and a somewhat smaller number of patients who are minimally conscious state. Are there some more objective ways we can tell whether that patient in front of me who sort of barely has sleep wake transition actually has still some sentient, some consciousness left? We would like to develop some objective test.
But ultimately, what you really need to do, you need to go to animals, because anything you can do in humans is very limiting for obvious reasons. The tools we have are very good. The tools we have, by and large, measure bulk activity. They measure hemodynamic activity, or they measure bulk EEG activity.
And you've got remember that the smallest voxel in most advanced human imaging device includes roughly three million neurons. And so that's very, very good. And the time cycle we measure has a-- it depends on blood flow. It's three to five to eight seconds, so it's totally mismatched to the signals.
We really want to be interested in which are neuronal signals. And of course, then, you want to interfere with the signal, and you can only-- you can't really do that in humans. So you really need to develop behavioral assays. And together, with David Anderson, we kind of do this for mice. So you can try to use-- so there this strategy is you use paradigms.
And in humans, has been shown where you have some behavior that requires consciousness, and you slightly change the behavior. And for a slightly different task, you don't need consciousness. So you have this contrast of strategy in humans. And then you try to develop a very similar paradigm, let's say, in monkeys or in for vision or in the mice using associative conditioning or whatever your model system happens to be.
So again, you want to have a battery of these assays so can determine which species have behaviors that require consciousness. Then, of course, what's really absolutely essential-- and this is true for all of the systems neuroscience, no just for studying consciousness-- you want to deliberately, transiently, reversibly, and delicately interfere with the processes that lead up to and following the end the Neural Correlates of Conscience and model organism. So let's say, you take a monkey. You train the monkey like Keller was trying to do.
You train the monkey to see yellow and to tell you by doing two a turn of false choice behavior whether he's seeing yellow or not seeing. And then, you want to interfere and take out those neurons, let's say, using a specific promoter in a specific part of the visual system that's expressed in neurons that we have some reason to believe are involved blue-yellow perception and then turn those things on and off in order to go from correlation.
Because most of what we do, certainly in cognitive neuroscience, most of what we do is correlation. We still don't have very efficient tools to move from correlation to causation. And you can also focus on a search for random mutations in, for example, in mice or in other genetic organisms. And in fact, it may well be possible that we saw among the thousands of alleles of different mice, there may be already some that are sort of unconscious that are like zombie mice.
In a sense, that half complex behavior, but not those behaviors said that we think are associated with consciousness. And also what's important from a theoretical point of view, you want to develop a theory of consciousness. You want to develop a theoretical, probably information-based theory. And Giulio Tononi is doing that with Edelman to have a more justified understanding which systems and in which conditions show consciousness based on which criteria?
What tools do we need to understand consciousness? What's absolutely essential, and Francis. of course, always emphasize is we need to acquire a much more particular for the-- if you care about primates, humans, and macaques, or other non-human primates, you really need to much better and faster methods to acquire new atomical data, wiring data. It's really essential to know I'm not just recording V1, but I am recording from a layer of six cells that projects back to the LGN and only a five cell that project back to the [INAUDIBLE].
And I want to have a tool that lights up all those connections so I can then inactivate the [INAUDIBLE] record from them. We really need those tools. We need to develop electrical tools and optical tools to record not just from-- so in humans, we can now record from 50 neurons, and in rodents, you can do it from 200 using [INAUDIBLE], but you really want to have tools where you can sample 10,000 neurons, the activity at the millisecond time scale.
Then we need the mathematics. This hasn't really been mentioned at all. We need to understand how very complicated heterogeneous networks. So all the neural networks that people have done like Copperfield and other people, they're really very simple neural networks. They're all very homogeneous. And you have either things that excite inhibit.
Neurons are very heterogeneous. They spike. They burst. They have a neurotransmitter. They have have dendritic patterns. They have calcium-based spikes. They have sodium-based spikes.
And so we need a mathematics. What happens if we have 10,000 neurons or 100,000 neurons or 10 billion neurons? Very complicated ways and into acting at the millisecond time scale. And we don't really have such a mathematics right now.
We need-- this has been mentioned before. We don't have this at all in humans. We want to be able, non-invasively, to sample the human brain activity at the millisecond in micrometer. We want to do it both for clinical purposes because we want it to-- we want to help people who are paralyzed or locked-in syndrome without opening the skulls.
And we also want to do it for scientific reasons. And right now, this is really a huge hole in the spectrum of our technologies that we have. And again, we need ways to perturb and intervene in the mammalian CNS in a very delicate and kind of manner.
I found this great quote in a review of the Beatson vote during the First World War when you reviewed Thomas and Morgan's book, The Mechanism of Mendelian Heredity. So here, Morgan, at the time still at Columbia, I believe, made the point that genetic information is stored along one dimensional. It's stored along one-dimensional strength in modern language.
And at the time, people didn't understand. They didn't-- even people who are materialists like Beatson didn't understand how chemistry could be used to store all the prodigious amount of information that makes up an individual and pass it on the descendant. People thought they understood chemistry, but they didn't know how this could work.
And so people like Bersaint developed new philosophy. We need a long vital, vitalis or people like Schrodinger, as late as 1944, said, we need new physics in order to explain this. And so he has this quote. "The properties of limit--" Well, you can read it. So the problem here was people didn't even-- at this time, they didn't even have the idea of a specific micro molecule.
They didn't realize that hemoglobin in me is not just a [INAUDIBLE]. It's not just a collection of different molecules that have this huge spectrum of molecular weight. No, but it's all one and the same molecule replicated a gazillion times. He didn't understood the prodigious power of strings of nucleic acids.
And so when I go a lot to these consciousness meeting, and people constantly talk about quantum gravity and quantum computation and sort of other fancy spooky stuff, and we really haven't exploited the power of the brain's, by far, the most complex system, and we don't know it. And we really have no idea how-- we just don't understand the power of the brain yet. And so we shouldn't sort of call up for all sorts of fancy new laws before we sort of really understand the complexity of these biological networks. And so I'd like to leave you with this last point, which fancies my entire emphasis that, "consciousness should be treated as an empirical problem to be tackled by the biological sciences." Thank you.
[APPLAUSE]
PRESENTER: Thank you. He will be back when he comes up here. Next, we're going to turn to someone whose questions about consciousness may be coming from an angle. Dr. Patricia Smith Churchland is a Canadian American philosopher. Currently, she is chair of the University of California at San Diego Philosophy Department, an adjunct professor at the Salk Institute for Biological Research, and Associate of the Computational Neuroscience Laboratory at Salk.
She is surrounded by philosophers, as her husband, Paul Churchland, also pleads guilty to that profession. Her recent work focuses on neuroethics, the attempt to understand choice, responsibility, and the basic of moral norms in terms of brain function, brain evolution, and brain culture interactions. Please welcome Dr. Churchland.
[APPLAUSE]
CHURCHLAND: I'm not sure. Oh, yes, I am wired. Great. Has anybody got any glutamate left? It's very late in the day, and it's been a very, very rich and really very wonderful day. Although my brief is to talk about consciousness, I'm actually going to say very little about it. What I would really rather do is to sort of take a very broad picture about some of the open questions in neuroscience and just to say a little bit about what we don't know, given the background of really this stupendous discoveries that have been made in neuroscience over the last couple of decades.
I begin with this particular slide, really, just to remind ourselves that our brains are as they are because they are the products of natural selection. And as we now know, the conservation of structures is really very remarkable across species. And it's unlikely that we are going to find in ourselves structures and even functions that are absolutely unique and have no precursors in other animals.
All right. The other point that's useful to make from the background of evolutionary biology is this-- that we differ from plants. That is, animals differ from plants fundamentally in having a nervous system that enables us to move. We don't have to wait and take life as it comes. We can look for food, look for mates, see to it that our offspring are nurtured, and so forth.
Motor control, or more broadly perhaps, behavior is really the fundamental thing for which nervous systems are designed. And I think it sometimes is forgotten that as we explore the nervous system from its sensory aspects or from the aspects of learning and memory, that whatever those functions are in their nature, they have to be serving behavior and motor control, so that perception is not there for the sheer wonderful beauty of it. It's there only because it serves behavior and motor control in some fundamental way.
An organism that has absolutely stupendous color vision, but that-- where that color vision in no way informs its behavior is not going to survive and consequently will not pass on its genes. Oh, let me just go back. I'm sorry. It was really Paul MacLean who put the matter very succinctly. The basic function of nervous systems is to enable a body to move so as to succeed at the four F's-- feeding, fleeing, fighting, and reproduction.
[LAUGHTER]
And I think that that does essentially capture the important point. Between the sensory neurons on the sensory periphery and the motor neurons on the other end or what, to a first approximation, we can call the interneurons. And the basic function of interneurons is to bring it about that behavior is informed.
Part of what the nervous system needs to do is to have increasingly good predictions about what's going to happen in the world. And to the degree, other things being equal, that is motor equipment, and so forth to the degree that an animal has better and better predictions about where food is. Who constitutes a good mate? What's wrong with this? Will the wasp sting? To that degree, it's for survival is promoted.
Now, much has been asked, especially in the morning session, about nature versus nurture, And do we have a gene for sociability? Do we da, da, da? And several people that I chatted with over the break rather mourned the fact that people are still, shall I say, burdened with the idea that there is a fundamental distinction between nature and nurture, so let me just make a couple of obvious points that I'm sure most people here will be familiar with.
One of the great things that I think we have learned over the past couple of decades is the distinction as described in the 1970s and '80s was really deeply misconceived. I think it was Sydney Brenner who once said to me something like this-- well, look, suppose that mother nature got to a point where she said, well, look what should I do now? Should I put into the genome, synapse by synapse and neuron by neuron, how to make a brain? Well, that would be really very difficult.
Or should I, as mother nature and wanting to get things done quickly, should I do the following? Should I rely on the fact that there are many regularities that obtain that the genes can take advantage of? So regularities and, for example, chemical milieu of the conceptus and the fetus or regularities that exist in the post birth situation. So that the genes can count on the existence of those regularities to turn on other genes, which turn on other genes, and then we get gene expression in this wonderful, intricate, regulatory cascade.
So it means, really, that mother nature doesn't have to build at all in. But in a certain sense, it's built in, by which I mean that, there is this dependency on the environment being in a certain way. And if the environment isn't that way, then something else will happen. And I think that gives us a completely different picture of nature and nurture, and it really does mean that everything is much more intertwined than we used to think when we talked about genes for this and that behavior.
It came as a surprise to me, and probably to many other people, to be told that at least in mammals, there are no genes for neurons. I thought, well, how could you know that you're supposed to be a purkinje cell or a stellate cell or a basket cell, unless there were genes for those? But, at least, there should be genes for neurons. It turns out not so.
It turns out that what there are, of course, are genes for precursor cells, which given a particular chemical melia, will become a purkinje cell or will become a stellate cell and so forth. And once you see that, it gives you a completely different sense of how to use the word, hard wired, how to use the word, predisposed or determined and so forth. And I think that's all I really want to say about that.
But given that mother nature has decided to build cascades of regulatory genes and to take advantage of regularities in the epigenetic milieu. It turns out that learning is really a very cheap way to improve both motor control and predictive capacities. You can rely on the existence of certain regularities in nature to alter the neurons, which alter the gene expression, which alter the neurons even more. And so learning is a very, very deep and very, in a way terribly clever way, to manage to increase your predictive capacities.
And much of what we call a cognition, and this, I think really has come to the fore as a result of work in psychology as well as neuroscience, much of what we call a cognition is really skill. I think we went into the early stages of cognitive science using language as a model for knowledge in general. And we thought of knowledge as being essentially propositional, but mostly it isn't like that. Mostly, it's a skill-- spatial skills, skills for navigating the causal world to know, to understand the causal structure of that part of the world that's relevant to my survival.
I don't think I need to say anything very much about this, except I suppose, we are amongst a small minority of Americans who do think-- or a small minority or humans who do think that mental phenomena are nothing other than phenomena of the physical brain. There is no non-physical soul that does the thinking and the feeling and so forth. It's an illusion of the brain to think that we do. And it's very powerful and, perhaps, a very useful illusion.
And the self, both the sense of self and the self-representation of the body, the self-representation of one as having a past and as having a future, the self-representation that's involved as animals conceive of goals and plans and carry them out, they are all, one and all, just constructs of the brain. I'm not really going to say very much about consciousness except that I have-- although, I like that the take on that Christof and Francis have. For some reason, I guess, I'm just kind of a stubborn person. I just kind of look at it in a slightly different way.
Okay, am I pressing the button? Or yes, I'm pressing the button, but there we go. And this came up early this morning. And I think Sydney and I were, at least in some degree, of agreement. I tend to think that the problem of consciousness is really analogous to the problem that people face at the turn of the century about what is life?
And as you know, of course, some people thought that the answer-- well, what's the difference between a dead thing and a live thing? That the answer could only be élan vital, that there is some sort of vital spirit. And perhaps, some people looked for the vital correlates, you some say, of livingness. And I don't think the vital correlates of livingness strategy, with all due respect, I don't think that was the strategy that taught us about what life is.
It turned out it came from a huge amount of work from cell biology, understanding the nature of membranes, of ATP, of energy use. It came out of understanding molecular biology and how proteins get built and so on and so forth. And it's very interesting that it wasn't that people finally said, well, we now have the answer to what life is. It's just that question doesn't get asked anymore because this huge amount of basic cellular and cellular physiology and anatomy makes that question [INAUDIBLE].
And I have the feeling, and I could be totally wrong about this, that consciousness is a bit like that. The disanalogy is with something like, what is fermentation? Where there was a project, an answer was achieved, and we knew what the answer. I don't think that the question about consciousness is like that. There isn't, in my opinion, likely to be a single answer.
And although, I think in some ways, Christof probably agrees with me because we both agreed that you can start with what it's not. It's not consciousness. It's not in existence in any of those instances, and there are a range of phenomena, although we tend perhaps to focus on awareness of visual phenomena.
And we saw today the spectacular work of Richard Axel on sensory of olfactory phenomena. But there is also, of course, imagery. There are these rather more diffuse things like feeling well or ill or energetic. We're aware of spatial relations. We're aware of temporality. We're aware of ourselves as agents partly because we have [INAUDIBLE].
So the indirect approach is that we need to really understand a whole lot of functions, some of which I list here. But in addition to that, this is the part of the talk where I want to say that, I, of course, am one of the people who is very thrilled and excited by the kind of developments we've seen in neuroscience, especially over the last two or three decades because we really do see all kinds of places where the traditional philosophical questions are being ushered into the neurobiological laboratories, and that's a wonderful thing.
But I also want to emphasize, and this is something that also came out this morning, that there are major issues that are not yet understood and resolved within neuroscience. And I'll try and quickly go through these without pausing for a long discussion. We do not yet understand how information is coded in neurons. It is true that tremendous progress has been made, especially in sensory systems, but bear in mind that there are many instances where there is a representation in the nervous system, but there is no stimulus.
Let me give you an example from-- I'm going to go forward, then, I'm going to want to go back that we saw this morning. We see the brain. The visual system represents a line of the triangle. It isn't there. So there is nothing in the stimulus to which that line corresponds.
Similarly, in the case of so-called subjective motion where a light goes off here and comes of here, and you see a line moving, there is nothing in the stimulus to which the activity of an MT neuron corresponds when it signals that it's aware or that it's responding to motion. But more broadly, I suppose, and there's one thing Terry pointed out to me that as a result of the work that he and Terry [INAUDIBLE], that he did with Sydney Leckie, there isn't really anything in the world that corresponds to white.
And yet we have the representation of white. It's something that the brain constructs. Brains also habitually constantly represent goals, plans, projects. And the thing that is represented does not yet exist. So when these really, quite wonderful accounts from the information theory people are offered to us as an account of how neurons process and encode information that has to be remembered, those accounts, one and all, depend on the idea that the neuron is responding to an external stimulus and often, and perhaps importantly, almost never is so.
Oh, sorry. I want to go back, if I may, please. And back one more. One more back. Back. There we go. The issue of spontaneous activity is not unrelated. And I'm going to go very fast through the rest of these. But one of the things that has now been a question that people are working on is, so what is all that activity in the resting state?
So when you put someone in a functional Lamar and give them a task, of course, what you're doing to get those wonderful pictures is subtracting from the baseline. Or you're subtract in the baseline from the experimental condition. So what is all of that spontaneous activity?
And Mark [INAUDIBLE] group has discovered that there really are two kind of systems that we wouldn't really have known about save for this kind of work. And that is that there is one system that is more active during the task, where the other system tends to be less active during the task and then the reverse in the resting state. And there are certain parts of cortex then that are more active during the resting state.
And they also seem to be ones that one may speculate are involved with such things as self-reflection, planning, envisaging the consequences, imagining what might happen, considering what might be the spatial layout, and so forth. But there is a huge amount of spontaneous activity. And of course, we also see it at the single level, cell level, and we really don't know what that's all about.
We don't know how the organization of the motor response is achieved. We don't know how there is integration. This was already alluded to earlier as the binding problem, but we don't know how integration across sensory systems or integration across sensory systems together with memory and giving a motor response. We haven't really a clue how that's done.
And within memory, well, there's been beautiful work done on individual cells, the LTP work and beyond. But how is memory all orchestrated across a network? And as Sydney raised the question this morning, how is it that, as it appears, information is initially held in the hippocampus and somehow gets transferred to the cortex? I mean, that's not just a sort of one single cell to one single cell, it has to be a large scale maneuver. We have no idea how that's done.
Okay, and then, we'll go to the next one. I'll leave off functional organization for the moment. Time management is a problem that is very deep and we've really only begun to address. And I won't really say anything about that now because I'm running out of time. We don't yet understand-- sorry, we don't yet understand why it is that we sleep and dream, although it's fascinating to discover that the humble fruit fly also shows patterns that are describable as resting and active that look in many respects like sleeping and being active.
And finally, we really don't yet understand many of the fundamentals of how information is stored and retrieved. An intense amount of work, of course, has been done on declarative memory and the role of the hippocampus. But skill learning is clearly a very, very critical part, as I suggested earlier, of how cognition works, and that remains very, very poorly understood.
My own sort of hunch about these things is that there are some really fundamental sort of-- I hate to use the word paradigm shift because it's such a cliche, but that we are really in for a very different way of looking at the brain and how it builds this model within this-- In this meet, it builds this model of the other world, of the external world. It cannot have direct contact with the external world. It's all mediated.
And somehow, it builds a model of the external world and builds a model of itself as acting in that external world. And I think there's something very deep about that. And I don't think it's philosophical. I think it's neurobiological. And I don't think we've quite got a grip on it yet. Thanks very much.
[APPLAUSE]
PRESENTER: Going to ask our last three speakers to have a seat. We'll have a nice 10 minute discussion on time management, on neural-- feels like today we've had a debate between the 1960s B.F. Skinner and Konrad Lorenz as about studying how to have the correct way to study behavior. I'll take questions from the audience if you want to get up to the microphone.
And I'll just start out of-- just let me open up the question to say that this morning, I would venture to paraphrase four out of five novelists who would agree that who said that exploring consciousness is a waste of time. That, sooner or later, you'll give up on it as being meaningless. I imagine you don't agree with that--
KOCH: No.
PRESENTER: --Christof.
KOCH: No, I don't [INAUDIBLE].
PRESENTER: Okay, just wanted to get that bit of bookeeping out of the way. We have a question in the audience. Up there. Yes.
AUDIENCE: [INAUDIBLE]
SHULGIN: There has been a very recently, quite a bit of increase of use of psychedelics in various therapeutic applications. The studies are going on in Los Angeles and one other area, I forget where, of the use of something like psilocybin or MDMA in administering to people who have terminal cancer to alleviate the anxiety of that. These have been quite successful. There have been studies in the post-traumatic syndrome, again, to relieve anxiety.
There are probably half a dozen such studies either underway now or in the machinery of approval to be done. The primary negative of this entire area is the public opinion, the legal status, the general attitude of the authorities that any work with these materials is probably, basically, evil, make the getting of permission from the authorities, means that be it health, be a drug, authorities virtually impossible to get, and hence will be a long struggle for any of these material these studies to become real.
The funding is no question. The funding is available from many sources. It's the machinery of the permission getting that has been difficult. I don't see it being softened at all until the-- as it happened in Europe to some extent, and more and more-- this entire area, research moves from legal control to medical control. And I think that transformation will probably allow many of these studies to be done
PRESENTER: Dr. Kandel?
KANDEL: Yeah, I want to come to the defense of Christof Koch, not that he needs it from me. But I have followed as an outsider with considerable interest the work that he and Francis Crick have done over the last 20 odd years on the problem of consciousness. And I think we all would say, including the two of them, that they have not moved the problem brilliantly in empirical terms. Nonetheless, if you think of how fuzzy our thinking was about those issues 20 years ago and how clarified it has become because of their continuous evolution in their thinking, you have to say, this is remarkable progress.
And I think that in addition to methodological advances, there were conceptual advances. And they have helped us think through a number of ways that couldn't, in principle, study consciousness which we didn't have before. And I'm a little bit reminded of another conceptual advance of this sort. I mean, everyone points to shredding important, but what is life?
And why did it influence so many physicists come into the field and even influence biologists? Because it pointed out the role of the gene in information transfer. So it changed biochemistry from being influenced by primarily being focused on energetics to the flow of information within cells.
And I think that you and Francis have accomplished this for us. You've given us certain ways of thinking about it, which we didn't have before. So in that context, I want to ask you a specific question.
[LAUGHTER]
I thought that you were going to devote your lecture in the claustrum and how the claustrum them brings together activity in the posterior parietal cortex and the prefrontal cortex. They give us consciousness. And I realized that this was only an idea thrown out sort of as you were evolving in your thinking, but I wonder what your current thinking is as to whether there is some group of cells that ties together the activity of the number of cortical circuits in order to give you this unifying function?
KOCH: Okay, so the claustrum for those who don't know, there's structure in mammals-- I think all mammals have it-- that's sort of underneath the insula, underneath the front part of the cortex. I think it's bilateral. It's unique in a sense that gets input from every other-- it gets input from every [INAUDIBLE] and projects back to every [INAUDIBLE]. It's very thin. It's a sort of almost a single sheet of neuron.
Nobody has any idea. Very few people study it. It's obviously there in us also. And so if you look at the structure, and you follow structure function, you merely infer that it must have to do something with integration. So we were talking to-- Axel was talking this morning about the binding partner. It seems like an ideal structure to do that.
The only relevant study right now is FMI. And they suggest it's involved in cost model integration when you have more than one sensory. It's a visual and auditory, then the structure gets very gets very active, so that would support that. But because it's very thin, it's thinner than a single voxel. So it's very difficult to separate it from the underlying basal ganglia and from the overlying cortex or the [INAUDIBLE] problems.
And I think what has to happen is what you did in the amygdala. You have to find a gene that specifically express there and then try to inactivate or activate it and then manipulate it. It's just too difficult to study in humans.
PRESENTER: Any other? Anybody? Yes, sir up in the cheap seats. Go ahead.
AUDIENCE: [INAUDIBLE].
KOCH: Yeah, there are plenty around. I mean, if I ask you to do a difficult problem, if you're unconscious, you're not going to be able to do it. So there was a paper by Larry Squire in Science a couple of years ago using trace versus delayed conditioning. And the evidence shows, and other people have replicated it, that if you do a delayed conditioning where the CS and US overlap, you don't need to be aware of the CS and the US relationship. But in trace, in order to be conditioned. But if you had trace conditioned, in other words, if there's an interval between the tone, and the shock, then you have to be conscious of the--
AUDIENCE: [INAUDIBLE]
KOCH: You would what?
AUDIENCE: [INAUDIBLE]
PRESENTER: He said, if you cannot use the--
CHURCHLAND: Said, if it's an animal-- is that because it's an animal?
AUDIENCE: [INAUDIBLE] consciousness, the way you're using that term, [INAUDIBLE].
KOCH: And so A, you can do the same experiments in bees and in flies. Like, you can do delay and phase conditioning. How do I know you're conscious? I don't really know you're conscious. I infer you're conscious, okay, because of similarity to me, similarity in behavior, because of the legal system. If I don't treat you as conscious, I'm going to be in trouble, okay? So--
[LAUGHTER]
--how do I infer that the monkey's-- How do I know my dog-- I mean, Rene Descartes makes the following outrageous statement. He says, if you ride in the carriage-- I mean, this is a French dualist, right? --if you ride in the carriage over a dog, and the dog cringes and yelps, don't believe he's conscious because he's just-- he's just [INAUDIBLE]. It's just behavior.
We, today, we think that the animal actually has pain, actually has all these subjective, the negative affect of pain. Why? Because of similarity in behavior, similar nervous system, and evolutionarily continuity. So it's-- I don't think it's that difficult to infer consciousness in other organisms.
PRESENTER: If an animal shows empathy for another animal, does that mean it has consciousness?
KOCH: I mean, it depends. I mean, it depends how you operationalize it.
PRESENTER: I'll ask this because I interviewed someone who worked with bonobos and chimpanzees. He said, there's a world of difference between a bonobo and a chimp. And bonobos really exhibit so many, so many aspects of that are much more human-like that if you study them, you'll get more an idea of what our closest relatives are like. And he talked very much about the empathy that they have to other animals. I mean, it's almost like they were talking about this animal being conscious of the feelings of the other animal.
CHURCHLAND: Well, there are lots of experiments, both on regular chimpanzees and troglodytes, as well as bonobos showing, so to speak, theory of mind, so that one animal will show a behavior that indicates that it has an understanding of what the other animal sees. So, I mean, one classic one consists of the subordinate chimpanzee and the high-ranking alpha male sitting apart from each other. And if food is put in a place where the-- and there's a barrier, a glass barrier-- but if he can see it, this guy won't take it.
But if it's put it here so he can see it, but he knows that he can't see it, then he will take it. And so that's the source of experiment, but there's lots done like that. So I think I'd look for a single criterion like empathy or being able to laugh or being able to look in a mirror and recognize yourself. I wouldn't look for a single criterion because it looks like the phenomenon involves a whole lot of stuff. That would be my guess.
PRESENTER: One last question on the aisle here.
AUDIENCE: Hi, I have a question, in general, about consciousness from the point of view of evolution. So what I mean here about consciousness is the feeling that you would say that I am doing this because I think all other emotions can be also correlated in zombies. Like, you laugh when something happens and something like that. But we have a feeling that we are doing this.
This could be an illusion of the mind. But my question is twofold-- one, is that, why would evolution would create this illusion of mind? And for all the other phenomenons like vision and everything, we've been able to create it in machines.
But so did evolution really have the ingredients to create an illusion in the mind that someone is doing this? I mean, evolution cannot create anything if it is needed, like something like predicting the future or something. There's still [INAUDIBLE] why it could do it? And can we say it could do it?
KOCH: And so the question of why has many answers. One answer is if you believe that you-- I'm the author of my action, if I believe I'm actually the author of lifting my hand, it makes me more effective actor. And that's how Dan Wegner argued in his book, The Illusion of Free Will. That those systems which have this perception of authorship, they're more efficient because they believe that their action actually makes a difference.
As to the question of machines, we have no idea because we don't have a theory of contents. So we don't know once machines, and they're getting closer can perform at our level of sensory motor positing and vision and our olfaction, et cetera, whether that means automatically they'll have consciousness, or whether you need something in addition to that, whether you need-- does it need to be a pilot machine?
Does it need to be a pseudo machine? Does it need to have-- what sort of architecture is it? We don't know because we don't have a theory of which systems, whether artificial or natural, have these states, these so-called subjective states. We don't know right now, and your guess is as good as my guess.
CHURCHLAND: But I think you're right. I mean, there might be another--
KOCH: It won't be Windows.
[LAUGHTER]
CHURCHLAND: There might be another--
[LAUGHTER]
PRESENTER: And that's forever.
CHURCHLAND: --another point, too, that you want to make here. And that is, we sometimes enter the discussion about consciousness as though it's all or nothing, that you have this great, huge, humongous, rich thing, whatever the heck it is, or you don't. And actually, when you reflect on it, and look at the data, it really looks like it's more likely that it comes in grades and degrees. It can be diminished. It can be enhanced.
And it may be that there are early mammals that have it in, and perhaps, reptiles and perhaps, for all I know, insects, that have it in a very modest degree. But it doesn't have quite the same dimensions or the richness, whatever that is, that we have. So I think that probably is an illusion as well, actually.
PRESENTER: All right, it's no illusion we've run out of time. And I want to thank you all, for the panelists for joining us. So I thank you all for a day of tremendous discussion.
I have one announcement to make, is that there is going to be the unveiling of the manifesto and portrait downstairs. And we're going to split up into two groups. We're going to have the MIT and Picower Foundation and MIT Picower faculty go downstairs to the unveiling. There they're not really enough places very high to the stand.
So the rest of us, if you don't have the invitation, please stay up here. We'll watch it on the big screen TV together. And this way, we'll have enough room for everybody to see what's happening. So thank you again for a wonderful day.
[MUSIC PLAYING]
I understand that we have a little bit more room downstairs. So the invited guests can sort of squeeze down there. If you have a badge, you get in, I guess is how it is. Thank you.
[MUSIC PLAYING]
SOUND TECHNICIAN: Check 1, 1, 2, 3, 4. That's good. 1, 2, 3, 4. Do you hear it upstairs?
[MUSIC PLAYING]
Check 1, 2, 3, 4. Blue microphone.
[MUSIC PLAYING]
[CHATTER]
SPEAKER: Okay, so it's still raining by the way. Bill, wait, wait, wait, wait, I'll tell you. All right, so when we come in from these doors every day, there are two things we should be reminded of. And one, is who made this entire thing possible? And the second is why we are here?
And the other two, who made this possible? It is obvious that if it were not for Barbara and Jeffrey Picower, we would not be even standing here. So as a reminder of their generosity and trust and the hope you press to us, we'd like to present something very special today to both of you. So please open it. And come over here.
[APPLAUSE]
So, yes, I hope you like it. All right. So I as to why we are here, that is also equally important. And Picower Institute was not created to tackle small problems, but to take on the most challenging. So the faculty over the Picower Institute led by Matt Wilson, is he here? No? He's too embarrassed to be here.
[LAUGHTER]
Drafted what we call a manifesto, which is basically to say our promise into the future, which every student and every faculty members, all their staff, every visitor will see when they come in and out of here, and I would like Ellie to read this. And he can unveil now.
ELLIE: We hold this our ultimate mission to understand the mysteries of the mind through studies of the brain. We dedicate ourselves to the principle that the advancement of knowledge comes through an unwavering commitment to the integrity of science. We recognize that achieving these goals will require the shared vision and the collective efforts of all members of our community. We embrace the beauty of truth and the power it holds to change the world. This we believe, and this we shall accomplish.
[APPLAUSE]
SPEAKER: And here's the [INAUDIBLE] and it's been photographed by people [INAUDIBLE]. Two of them are missing here, but we put them in the painting.
[LAUGHTER]
[APPLAUSE]
All right, so everybody, thank you for participating today. And so for invited guests, there will be a shuttle I think, right outside.
SOUND TECHNICIAN: Yes.
SPEAKER: Right outside? So you can go back to MIT hotel and have a few drinks.
BARBARA PICOWER: Thanks a lot.
SPEAKER: Thank you.
BARBARA PICOWER: It looks fabulous.
SPEAKER: Thank you.
[MUSIC PLAYING]