 Okay, let's get started here. So we have a short week, as you know, since it's Thanksgiving, no class on Friday. And Emotion, I think on the course website it says Emotion and Social Cognition or something. It'll be pretty much just Emotion today. And then we're going to talk about some diseases of the brain and mind, starting with autism on Wednesday. You already heard about Alzheimer's disease and neurodegenerative diseases last Friday. And then there'll be, I think, two more on depression and schizophrenia. One other quick note I had here, especially since I've been teaching some of a few more of the lectures towards the end here, which will therefore also be represented more on the final. If you guys have questions and want office hours, I have them again when, ostensibly on Sunday or something, or Monday. But nobody ever came. Just email me and we can set them up anytime. So if you have questions that you want to go over, in particular as you're going over some of these, the lectures in the last half of the course, in preparing for the final exam, send me an email if you have questions, can set up a time, anytime, including weekends, if you want to have office hours with me, go over stuff, okay. This is the list for the readings as on the website. So, and the usual list of things to take a look at to make sure that you understand what these mean for today's lecture and the one on Wednesday. And then no lecture on Friday, as I mentioned. So what we're going to do today is cover some theories that people in the past and psychologists as well as some neurobiologists have had about how to think about emotion, talk about some emotions in particular fear, and some of the brain structures that are listed here that are involved in processing emotions. So here's the warm up question for you guys, that you're now old hands at. What's the answer to this question? Anybody want to take, anybody other than Olivia, I want to take a stab at answering this question either from the reading or just even if you haven't done the reading, what would you say an emotion is who wants to take a stab at it? John, somebody called John and the great okay, I figured that would be somebody that's had that name. How about you? Okay, that's a state of mind regarding some situation. Okay, so that incorporates two important ingredients. So state of mind would be something like a conscious experience or something like that. And then a situation would be whatever it is that you're evaluating. That's certainly one line of answers that people have given that it has to do with feeling with conscious experiences. So many people would think you can't have an emotion if you're not conscious of it, if that doesn't make any sense. And it typically has to do with evaluating some situation that's making you afraid or sad or happy or whatever it is. Anybody else want to take a stab at that? What kind of answer would you give to this question if you were investigating emotions in species where you can't tell if they have conscious experiences because they can't tell you about them like animals. So say you're working on emotions in rats, then state of mind or consciousness are not probably concepts that you want to use. You would want to come up with a different explanation. Anybody want to try and take another stab at a complementary answer to this question if you were studying emotion in non-human animals? Try it, somebody. Good. Okay. So both of you guys talked about situations that induce or trigger the emotion. And in one case, the idea was that maybe we would be focusing on states of mind and what it's like for the person or the animal to experience the emotion with an emphasis on sort of conscious experience. In the other case, it's more operational. And it's like, well, what emotion would we infer or attribute from just watching the behavior which we do for animals? And of course, in many cases, do in the case of people as well. And as you'll see, those are basically the along the lines of theories that people have provided for emotion. Before getting to those theories, one way, if you read the literature, especially in psychology, that people have defined emotions, not to define it at all, but to or to define it by exclusion, by contrasting it with cognition, which of course doesn't make it any easier because now you have to say what cognition is. But the intuition behind this distinction is one that goes back a long ways. It's basically the distinction between thinking and feeling so that there would be sort of more rational, effortful, thinking-based information processing that we would call cognition, and there would be more automatic, intuitive, feeling-based information processing that we would call emotion. It's vague, but people have fleshed these out psychologists, mostly in so-called dual process theories. And so a lot of the concepts that people, in particular psychologists, but not only, use to think about, theorize about, and investigate emotion derives from these. And there are many of these dual process theories. They tend to share this basic sort of these attributes in common. So I'm putting this up as a sort of example. Some people have specific names for these. The original distinction, which came from psychology, was between controlled or effortful processing on the one hand and automatic processing on the other. But the idea is that there would be a system for processing stimuli quickly, automatically, effortfully, effortlessly that would often involve emotion. So it would have like a quick reaction that, you know, you feel angry at somebody and want to punch them. But then there would be another system that to some extent could override control over behavior of the system one. And you might not punch the person because you reflect on it, you exert some control, and you're thinking that this is probably not a good idea because it'll have bad consequences, etc. So in many circumstances, it seems both intuitively and in the case of experiments that there is dual control over behavior by these two systems. One which is sort of pre potent, and if you do nothing, it'll just be automatic and you'll go and punch people. And another one that is more controlled slower. So it takes a little time to kick in, but that can override this system. And of course, one big impetus for this is that you see the relative weights or exertions over behavioral control of these two systems change across species and indeed across development. So in young children, they tend to have a lot of this system one, they will punch their peers and cry and have a lot of behaviors that you would interpret as consequences of an emotional state. Whereas in adults, those can be to a large extent controlled by other systems. So that's a big distinction, primarily from psychology. Another way that people have thought about emotions is to simply list them. So in the absence of a definition, you can certainly give examples. Here are examples that people have come up with for emotions. These ones, so-called basic emotions, as defined by Paul Ekman, a psychologist, and other psychological researchers working on emotion, are these six listed here. These are thought to be emotions that would be reliably signaled across all cultures, according to some theories, by human facial expressions. And they're also thought to be emotions that have basic programs for executing them in the brain. And moreover, that those programs arise relatively early in development, and for many of, for most of these emotions, are shared with other species. So the idea would be that things like fear, anger, disgust, arises early in development. We share that not only with chimpanzees and monkeys, but also dogs, cats, and rats. By contrast, people have come up with another set of so-called social or moral emotions that arise later in development. So children develop them later on. And these seem like we share less of them with other species. So rats probably don't have any of these. Other social species, like dogs, would have some precursors of these. But these emotions here only make sense in a social context, if you think about them. So there are emotions that are thought to serve adaptive functions in relation to lots of circumstances, both social and non-social. You can be afraid of falling off a cliff, etc. So there's a bunch of things here. And then there's the social or moral ones that specifically evolve to regulate social behavior in groups. It goes without saying. Emotions are important. Psychologists point this out and have done lots of studies of this. It's worth just being aware of this. According to many views, emotions are not necessarily just only experienced when you have a particular inducing stimulus or situation. But to a greater or lesser extent, you're always in an emotional state. So according to many theories, emotions are always there, only more or less so. And psychologists have mapped this out. Here's one example where they asked people just throughout their day to indicate how much, when they were doing these different things here, how pleasant or unpleasant this was. So neutral is this hard to see vertical line, and less pleasant than average is to the left, and more pleasant is to the right. And how much time during the day you spend feeling this way when you're doing these things is indicated by the size of the circle. But the point is that there would be many things in here where you wouldn't normally say this is particularly strong emotion, that still share with emotions the feature of having some valence that is being pleasant or unpleasant, which psychologists think is the sort of fundamental dimension of emotional experience. Just by way of aside, because we had social behavior or social cognition in the title of this lecture initially, there is a big field maybe larger than the study of emotion per se even, is the study of how we attribute emotions to other people from looking at their facial expressions, body posture, etc. And there's a lot of neuroimaging studies, so studies in both psychology and neurobiology looking at that. And the only point to make there is that this is similar to things that you've heard about in the classes on sensory systems like vision. So the idea is you would have social perception, you would perceive faces, body postures, hear somebody's voice, etc. at the at the perceptive end. And then you would make various attributions about people's internal states from that, how they're feeling, etc. So in a sense, it's very similar to this that we saw before for the visual system, where your brain is making inferences based on a context and everything else that you know about something that you can't directly observe. So remember in the visual system, it was necessary to make those inferences to make sense of complex environments like the visual illusion of this overlying triangle. Same thing, the idea is with respect to other people, you can't observe their internal states directly, but you infer it to make attributions from watching their behavior, etc. And so there's a big, just to say it, there's a big literature, a lot of work, here's a little cartoon of it, on so-called theory of mind, broadly understood, which is the set of mechanisms of abilities whereby we attribute internal states like emotions to other people. Okay, we're not going to talk about that anymore, but just go on, as scientists, how can we study emotions in humans and animals? One way to think of them is to think of their control over behavior, like this dual process scheme that I just put up, in relation to more reflexive and more volitional kinds of control over behavior. So if you start with the simplest kinds of control mechanisms, you would have just like a simple reflex, like the knee-jerk reflex, that doesn't require any, you don't have any control or thinking or planning about that. It's very rigid, it's very directly, the motor behavior, the reflex, the knee-jerk is very directly linked through not that many synapses to the sensory input, the stretch of the patellar tendon, and of course you can have these at many different levels. So you don't need emotions for this, you don't need any thinking or control, these are the simplest kinds of control over behavior, and then as you would go up to more and more complex kinds of behaviors, like mating rituals of birds and so forth, all the way up to long-term plans and actions, the idea is that emotions are kind of in the middle, and what distinguishes this sort of rough dimension here is the extent to which behavior is closely coupled, both mechanistically, neurobiologically, and in time, to the stimulus that caused it or is decoupled from that. So if you formulate a long-term plan to study for the exam, that's a behavior and it's triggered by some stimulus, us telling you the requirements for the course, etc. But the action is very decoupled from the stimulus, there's lots of intervening time and you have to hold these long-term plans in place. Whereas if you go to the reflexes, it's an immediate very direct connection, both mechanistically and in time, between the stimulus and the behavior. Emotions are intermediate, so people often think of them as decoupled reflexes. Like reflexes, they have some kind of potency over behavior, but unlike reflexes, they're not rigid as reflexes are. So you can be in a state of anger, but not act on it. You can be predisposed to punch the person in the face and you can remain in the state of anger for some time, but to some extent you're able to decouple the action punching the person from the stimulus that might have caused that anger state. Okay, so here's just one way of going through a simplified causal scheme for how a stimulus would be evaluated, resulted in an emotional response, and then eventually a feeling. And this is kind of one of the basic flow that one of the theories we'll talk about in just a second, the one by William James, envisioned. Things are not this simple, they're much more complicated in good part because all of these are bi-directional, so all of these are feedback, but the basic idea is that to start to induce any stimulus, any emotion to sensory stimuli under normal circumstances, you need to perceive the stimulus, snake or spider that causes you to be afraid, the other person that causes you to be angry, you need to evaluate them. So there's typically lots of contextual processing, like with the visual illusion of the canisa triangle that I showed you. So you need to go beyond just the stimulus, you need to figure out what that means in a particular context, etc. And then that is sufficient to induce an emotional response, which is typically highly multivariate. So let's say in your state of fear, you could measure changes in the body, I could measure changes in the way that you allocate attention to a stimulus, the way that you encode it into memory, of course I could ask you how you feel and presumably you're reporting about your internal representation of all those changes, and there would be some what we call feeling. But so the idea is that there would be this whole layer here and as I said context would be very important in order to interpret the stimulus. A lot of the emotions have to do with preparatory responses that involve predictions or expectations. So you might be anxious about giving a talk, or you might be scared of a dog that's about to bite you, but the dog hasn't bitten you yet and you're not yet giving the talk. So the emotion is an adaptive response that serves to prepare you to run away from the dog or to to give a good talk or whatever it is. So there's a lot of expectation there. And then it's important to point out that there are big individual differences and much more so than would be the case with reflexes. So some people are afraid of snakes and spiders and others have them as pets. So they're big individual differences and if they're systematic and extreme, of course, these constitute much of psychopathology in psychiatric illnesses. For mood and anxiety disorders, which are the biggest mental, biggest proportion of mental illnesses that they are, all arise from big individual differences in emotional responses to stimuli and your ability to regulate those. How do you organize emotions in some kind of framework or some kind of dimensional scheme? And psychologists have put a lot of work into this and there's lots of evidence from many different studies that indeed for all emotions, but in particular for the six so-called basic ones that are shown here, you can represent them in a lower dimensional space. So these are the six basic emotions, but you don't need six dimensions to represent them. Each of these dots here is one stimulus, which could be a story or a facial expression that is about that emotion, but you can map the variance in people's ratings of these onto a two-dimensional scale. So this captures most of the variance in emotions, is captured by two dimensions, valence, how pleasant or unpleasant it is, and arousal, how intense or not intense the emotion is. And you can kind of see how it maps onto this. So surprise would be a highly, highly arousing emotion and if it's a positive surprise, it would be over here on the pleasant side. If it's a negative surprise, it would start merging over into fear and fear could merge over into anger, disgust, and so forth. There are alternative schemes, but again typically two dimensions that people have come up with. This is from a researcher who wrote a book called, I think it was called Emotion Explained or something like that by Edmund Roles. And he conceived of emotions as responses elicited by reinforcers. And so if you have a reward up here or a punishment down here, these would be the emotional states that he thought would mediate your responses to rewarding or punishing stimuli. And then the withholding of punishment over here on the right or the withholding of reward would elicit these other kinds of emotions. So there are a bunch, but the main one that you should know is this one here, of dimensional theories of emotion that don't disagree that there are distinct emotions like this, but they think they all share something in common and that you can represent them as sort of XY coordinates on a simpler, typically two-dimensional space. Okay, so this is the first historical emotion theory you need to know about, formulated by the psychologist William James who wrote a paper in the journal Mind with this title, What's an Emotion? And his idea was that basically that scheme that I showed you previously, that emotions that initially you respond to a stimulus, there's a change in your body, your heart races, and your breathing increases, if it's fear, etc. And that the feeling of the emotion would be an interception, a perception of the changes in your body of all those changes. And he has this quote up here that exemplified the intuition that he had, maybe I'll just read it, if we fancy some strong emotion and then try to abstract from our consciousness of it all the feelings of its characteristic bodily symptoms, we find we have nothing left behind, no mind stuff out of which emotion can be constituted and that the cold neutral state is all that remains. So his idea was, you know, if you imagine, say feeling really sad, really depressed, he tried to analyze this and say, well, what are you really feeling? And it seemed to him that really there was, well, you know, you have like a lump in your throat and your chest is kind of clenched and something queasy in your tummy. And it's all these bodily feelings. If you get rid of all of those, there wouldn't be any feeling of sadness left. So to him, this the actual sort of quarrel of feeling sadness was constituted by interception of changes in your body. That might not be all of your intuitions. So other people's intuitions diverge from this. But that was the idea that that he had. Something that people do practically in emotion research is in fact, take operationally an approach by this. So by far the most common way to measure emotions in human studies is to measure changes in the body through psychophysiology. And you remember this picture, we had a while back of the two main components of the autonomic nervous systems. So remember that this controls glands and smooth muscle, the release of hormones that often accompany emotional responses by distinction to the somatic nervous system, which controls striate muscle that you move around with volitional movement. And they have these two big branches, parasympathetic and sympathetic. And they do, they have kind of opposite, they have opposite actions on their effector organs. So sympathetic is activated when you're running away from a tiger and parasympathetic is when you're sitting at home petting your cat and relaxing. And so for the sympathetic, you would increase heart rate, increase blood flow to large muscles so you can run away. There would be secretion of epinephrine and so forth. And for parasympathetic, it's the opposite. So these things are typically measured in studies of emotion. And the most common one probably is measuring galvanic skin conductance, which you heard about briefly when we talked about papillovian fear conditioning. This is a pure measure of sympathetic autonomic arousal. And it just measures how active the sweat glands on the palms of your hands or the palms of your feet are. So there's special acrine sweat glands on the palms of your hands that will get sweaty when there's sympathetic autonomic arousal. And so if you're scared or there's some other kind of arousal dimension to whatever emotion you're experiencing, you can measure an increased conductance by just putting electrodes on the palms of your hands. It's easy to measure, people measure that. You can also look at the heart rate or variability in the heart rate. This is more complicated. It reflects both parasympathetic and sympathetic control. But again, this changes as a function of the emotional state you're in. So people measure that. You can measure facial, electro, myography in the face or you can videotape the face. If it's overt facial expression, then you can look at people's either very slight or overt changes in facial expression. And that is related to emotions. You can see if they're smiling or frowning, et cetera. And there's a whole vocabulary that people have come up with in terms of the specific muscles that are activated when you are in a particular emotional state. And then there's a whole bunch of others that you can measure. So these are all composite measures of changes in the body that put a company in emotional state. One that's quite fast and easy to measure that people use nowadays fairly frequently because eye tracking is a common measure is to measure the size of the pupil. So there are many things that control the size of the pupil. One is just the light intensity. So it's a reflex, if you'll remember, way back to our visual lectures that goes from retinal ganglion cells through pre-tectal nuclei that controls the size of the pupil in response to light. But if you're paying attention, concentrating on something, or feeling emotion, this is also altered. And basically, if it's parasympathetic, your pupil gets smaller. So if you're relaxed, your pupil is small. And if it's sympathetic, if there's sympathetic arousal, then your pupil gets big. And these two branches of the autonomic nervous system, sympathetic and parasympathetic, achieve this effect on the pupil by acting on different muscles that surround the pupil, as shown in this picture here. You can see this, you probably know this and have watched this in people. You can see it very easily in animals. If you have cats at home and you play with them, which is what I do often with ours, like a laser pointer or something to make a little mouse, when they're really interested in something and they're going to pounce, the pupil dilates and gets very big. And in fact, you can tell if you're being confronted by a mountain lion or something, when you go hiking up in the mountains here, if the animal is staring at you, the moment the pupil dilates, you know, it's going to jump. So it gives you some warning. A few seconds, one second. The second theory that you need to know about comes from Charles Darwin, who wrote a book in 1872 called The Expression of the Emotions in Man and Animals. And he observed that a lot of the expressions were interpretable in animals as in humans and seemed to have a similar form, so that dogs and cats, through their body language and to some extent facial expressions, seemed to have a similar repertoire of expressing emotions as did humans. And he came up with three principles. One, he thought that these behaviors arose from ancestral behaviors that were adaptive. So when your pupils dilate, when you're afraid, he thought there must be a story to tell about why they dilate, that you know, you want to be able to get more light in, to see better in the dark. It's not trivial to come up with an explanation for that. Your visual acuity would decrease, actually. And anyway, but the idea would be that there must be a story to tell about how that came around, how those emotional behaviors arose from something that was adaptive. And people have come up with many theories. It's a really interesting question. If you think about it, you know, why do you weep? Why do you have tears when you cry? Why do you sob? What's the, what good does that do? It's not obvious to come up with an explanation, but one thinks there should be one. You also thought that there, this comes back to the sort of two dimensions grid that I mentioned to you before, that there was a principle that he called antithesis that emotions that were approach-related and emotions that were avoidance-related fell along some broad dimension. That sort of approach and avoidance was one of the broadest organizing dimensions of emotions. And that they achieved this in part by activating opposite muscle groups, which is indeed the case. And then he had nothing specific to say, because nothing was, of course, known about how the brain does this back then. But he did think that it rose from mechanisms in the brain. And this then gave rise to this idea of so-called, these so-called basic emotions that had particular systems in the brain that would trigger them. So here are just some drawings from his book where he observed, in this, for dogs, in this case, mostly body posture and also some aspects of the face that would signal particular emotions that he thought we can understand as humans, in part because they bear some analogy to what kind of body postures we would show. And then what he focused on a lot and what people since then have focused on, as I mentioned, are the facial expressions in the elements of emotions, which most other animals can't do because they don't have the muscles. But for the six so-called basic emotions, here are the six expressions. And again, one definition of basic emotions, according to Paul Ekman, who came up with the idea, is that they are, they correspond to universally expressed features of the face. Okay, so there's a lot of debate about emotions these days. This is a paper and there's a series of these from one of the foremost people working in particular on Pavlovian fear conditioning in the amygdala of Joladu at NYU. And he and others have argued against the idea that we should use the word emotion in neurobiological research with which I and other people disagree. But his argument is basically that normally, like what John was saying at the beginning, when you think of emotion, you think of some state of mind or in particular some conscious experience. That's a concept of emotion. And you can't tell if animals have those conscious experiences. Therefore, it's unscientific to use the word emotion when you're investigating these processes in animals. So, and I think, so there's a number of arguments like this that he's put forth and other people. So there are papers and indeed books these days that argue against a sort of basic neurobiology of emotions in part for this reason. So I think this is not correct. And David Anderson here at Caltech and BBE and I wrote a paper recently in the journal Cell and are currently working on a book to try and come up with a way of thinking about emotions that would avoid this problem and would allow you to investigate them in animals as a human. So the basic idea is that emotions you should think of these as functional states implemented in brain systems that are caused by particular stimuli that cause the behavior and cause the feelings. So a big reason for the debates and or people's confusions about emotions these days, I think, comes because they, for number three and four, they think of the emotion as constituted in part by the behavior or by the feeling. But as long as you separate these, then I think you can avoid these problems. So here's the basic scheme, very oversimplified, that you would have some stimulus it triggers a central emotional state. And then this would have a lot of consequences, like the behavior that is triggered, subjective reports, how the person feels, psychophysiology that you could measure. Importantly, lots of other effects on cognition, how your attention and your memory are modulated and so forth. But all of these are consequences of the emotion state that if you measure them, they give you some evidence that there is an emotion, but they're not the emotion itself. And so if somebody says, you know, I feel fear, well, that's just one piece of evidence. The feeling isn't part of the fear. And you don't need to say that somebody has to feel fear in order for there to be a central state of fear. And so you can study it in animals. And then, of course, this basic scheme gets much more complicated. So once you have that basic causal scheme I just had up, you would want to acknowledge that these are not static states, but processes extended in time. And so all the effects of them can feed back to the state. And there's lots of modulation that I already mentioned. So the particular context in which a stimulus occurs matters. If you see, you know, a snake in the forest versus a rubber snake that someone puts in front of you, if you look identical, these would cause different emotional responses. And there's to some extent, of course, conditional control over your emotions. And so what we did in this paper was to come up with a set of features that we thought would characterize emotions in humans and in animals that could serve to, as a starting point for investigating emotions. And so the features are listed here. So number one and two just say that there's this two-dimensional structure that I already showed you. Emotions can vary in intensity. And they can be positive or negative. Darwin's principle of antithesis. So this is this 2D similarity structure I already showed you. They persist for some time. I'm like reflexes. So they're states, they're there for some time. So they're in a position to influence subsequent cognition and behavior in a state of anger that'll influence how you behave and what you're thinking about and so forth for some time as long as you're in that state. They generalize, they have multiple effects. And then down here is the idea that they do have, going back to this dual process idea, they do have some automatic kind of priority over behavioral controls. If you do nothing, you will kind of be guided by emotions and it's typically effortful for your behavior not to be guided by emotions. And this is something that arises during development in humans for instance. And so babies cry and have tantrums, but humans are typically able to control that. Because of this priority over behavioral control in babies and in animals, not necessarily in adult humans, the emotional behaviors are also good for social communication because they provide an honest signal of the internal state of the animal. So you can tell by looking at a dog, if it's growling at you and the teeth are bared or if it's crouching down and looks unhappy, you can tell something over its propensity to behave in a certain way. And so these behaviors were co-opted for social communication. And of course, their intentional communication and regulated and you can deceive people in the case of adult humans, but that's not the case in other animals with emotional behaviors. And then there's various uniquely human aspects here as well. Okay, let me, so here's an example that some of you may have seen that illustrates the idea that although you can have a dimension, in this case of increasing intensity, that you can quantify. So it's the proximity that a diver that's going up to this animal here that's camouflaged currently, as you can just measure that distance and as the diver gets closer and closer, there will be an increase in the state of, in a threat or fierce state of this animal that then has very specific behavioral consequences when particular levels of threshold are reached. So that's, you know, the basic idea. You should be able to measure that in the brain. So here's an example, the driver comes up, he breaks camouflage, he gets even closer, swims away and makes an ink cloud. So there's a number of different behaviors, at least those three, breaking camouflage, swimming away and putting an ink cloud out, that all need to be coordinated and are triggered by the relative proximity of a threat. Here's another one that shows you the social communicative aspect of emotions. And again, many of these are very nice because once you have some of these features as a framework for thinking about emotions, it becomes possible to study them in quite simple and nice tractable model organism. This is from Roger Hanlon at Woods Hole who studies cuttlefish, which are amazing animals with very big brains, even though they're not vertebrates. And so what they do is they can very rapidly control the color and pattern on their skin, which they use for camouflage, amongst other things. But they also use it for emotional signaling. So over here on the left is a male, over here is a female on the right. And friendly behavior is signaled by this sort of reddish brown color, whereas the white color means aggressive behavior. And so what the male cuttlefish will often do is show one side of his body, the friendly side, toward the female, but show on the other side to all other males this anger aggressive side. And what's in this little video here, what's fascinating to watch is how flexibly this male cuttlefish can deploy this social communication of emotional state when he switches sides to the other female. So here they're swimming around and the right side is white on him, but now he goes to the other side and switches almost immediately. And the left side becomes white. So he needs to keep track of his relationship in space and who's there, these are other things doing stuff. So there are very clear behaviors and signals we can study in animals that seem to be linked to clear emotional states. And many of these have been co-opted for social communication. Okay, let's move on to two specific emotions, mostly one fear and talking a little more detail about them. So as I mentioned, if you were Charles Darwin, you would say that these of the expressions of these emotions, the behaviors associated with them are caused by them should have some adaptive story to tell. Adam Anderson at Toronto, formerly at Toronto, Cornell studied this in detail and looked at the facial expressions. And so the question here is, what good does it do you? What's the functional role, the value of doing this face when you feel disgusted and doing this face when you feel fear? There must be some story to tell why you make wide eyes and these kinds of features in one emotion and a different one in the other. And so you can list the particular features that reliably are caused by these emotional states. And then what he did in the study that's cited down there, if you want to look for more details, is measure since the acquisition of sensory information. So the idea was, the idea is that in the case of fear, you want to maximize sensory information acquisition because there's a potential threat out there. So you want to open up wide and gather all possible information. In the case of disgust, there's something toxic or noxious or poisonous out there and you want to kind of shut everything down and not sample what's out there. So in this slide here, what he looked at was just the widening of the eyes and he mapped out the visual field that people could see if they had these facial expressions. And in the case of disgust, you can see less than in the case of fear where you, because of your wide eyes, you're opening up your visual, literally opening up your visual view more in order to be able to detect events at the periphery, like potentially dangerous tiger or something stalking there. And something for smell, you measured flaring of the nostrils and the flow of air, the air velocity through your nose, which is increased in the case of fear. So you're able, humans pretty aren't too good at this, but you're able to smell anything that might be dangerous out there. And in the case of disgust, you scrunch them up so you don't smell whatever is disgusting. Okay, let's talk about brain structures. So there are many that have been involved in emotions from a variety of neurobiological studies. And we'll just go through three or so of these here quickly. One scheme here that goes back to the William James view, and again, need to emphasize this is not, this is way oversimplified. Nonetheless, the idea would be that you would perceive a snake or a spider or something out there, so that it has to be involvement of sensory courtesies at the front end in order to perceive a potential threat. And then at the end here, there has to be structures that can coordinate emotional responses, changes in hormones, changes in blood pressure, all of this. And we know what those are. We know that those are nuclei and the hypothalamus, the periaqueductal gray, and so forth. And we know what these sensory ones are, like visual courtesies. And so what you need is a set of structures like these ones shown here that can bridge between these two that can link sensory perceptions of the tiger out there to activation of hypothalamic nuclei that would trigger changes in hormones and heart trait and blood pressure. And then finally, this is much more debated, the idea is that there would be other structures, cortical structures that could then, in the sort of William James' view, represent all those changes that happen in the body. And these would be courtesies that get interoceptive input from the body. And those would be response, both would be the proximal substrate for feeling the emotion. So starting with hypothalamus, Walter Hess, who won the Nobel Prize in 1949 for his work in cats and others, Philip Bard and others who did similar work using stimulation rather than lesions, focused on the hypothalamus and the sort of the experiment that you need to know that your book also mentions that he did was as follows. You take a cat and you make a really big lesion. It's amazing these animals stayed alive, but there's enough processing in the brainstem for this to work. So if you lock off essentially all of the forebrain here, including the hypothalamus. So if you make a huge lesion, everything to the left of the line is lesioned. It turns out that amazingly that the animal still can move around and walk around and react to its environment to some extent. So it walks around and behaves and there's enough subcortical processing. This would never work in humans, but in cats it does. However, they don't seem to show any kind of emotional response to stimuli. Okay, so it's not very specific because you lesioned all of this area here. However, if you make a slightly different lesion, you also remove all of the cortex here that you removed before, but now you tilt it slightly differently. So the hypothalamic nuclei here are still intact, but they're no longer connected to the cortex. What you now produce is very strong and in fact exaggerated in appropriate emotional responses, so-called sham rage. So these cats would exhibit strong anger-like responses or rage and attack things, even if they were very innocuous and stimuli that normally the cat would not attack. The interpretation of this experiment was that hypothalamus was necessary and sufficient for at least some emotional responses, in this case something like aggression and rage. So that normally you would need the cortex to provide the appropriate input about the context and the sensory stimuli that should trigger the attack. That's gone in both cases, but here you still have the centers that can cause the emotional response, but they're no longer regulated by all the normal evaluative machinery in the cortex. So you trigger the emotions inappropriately. In modern day, this is work by David Anderson here at Caltech. People have put electrodes into not cats, but mice and rats in the same hypothalamic nuclei and you can record their responses of these cells that correlate with attacking another rodent. And I'll just show you one experiment in the next video that shows it sort of direct causal manipulation. You can use optogenetics by very specifically activating a subset of cells in these hypothalamic nuclei. So you go in with a fiber optic cable and you've expressed a light-gated ion channel in a specific set of neurons in these transgenic animals and you can activate very specific populations. And it turns out that that is sufficient to induce a state of aggression and indeed behavior corresponding to that in these animals. So here's the video. Here's the mouse up there. There's a glove lying around here and you'll see when the mouse moves around it has this fiber optic cable that's tethered to the ceiling with the fiber optic implant going into the hypothalamus. And then whenever they turn the light on, these nuclei in the hypothalamus will be very specifically activated. So here it is just digging around, not doing too much sniffing, exploring this glove. Now they turn the light on and the hypothalamus becomes stimulated and you'll see that there's this strong aggressive behavior towards the glove. You turn the light off, it goes back to just exploring and sniffing and so you can switch between these behaviors, the normal behavior of just exploring an object and a very aggressive behavior that's selectively triggered by activating the hypothalamus. It's worth noting that as you'll see in the video, it's not instantaneous. So it takes a little while in order to induce this central state of aggression and to have an effect on behavior. It's not like a reflex. It's not like you just flip a switch, but it takes a little while for the hypothalamic stimulation to coordinate this overall central state. There are a number of other brain structures involved in emotions that we don't have time to talk about. In fact, given the time, definitely not. I will just show you these here and say a few things about them and then conclude by showing you a little bit about the amygdala. So here are the structures that were in the previous slide. On the ventral surface of the frontal lobes, you have this cortex here. So you're looking at the bottom of a brain, the top left here, and these red and green regions are medial and lateral sectors of the orbital frontal cortex. And so this is a brain region that gets input from all sensory modalities. So it gets input from vision, it gets input from taste, from smell, but then it connects. It has projections going from there to the hypothalamus, pericarductal gray, et cetera. So it's one of these structures just on the basis of its connectivity that's positioned so as to link sensory representations with the elicitation of emotional responses. And there's a lot of work on this recording from neurons here or doing fMRI studies in humans. You find that this region seems to represent the value or the valence of stimuli rather than their sensory properties. If you have lesions here, you will impair processing of emotions. Here's this region that I mentioned that if you were William James, you would think would be the proximal substrate for the conscious experience of emotions for feelings, the insula. So this is a cortex that you cannot see on the outside of the brain. It's buried inside, but it's cortex. So if you carve away at a frontal lobe here, then this purple thing on the inside is insular cortex. And this is an interoceptive sensory cortex. So it doesn't get information about vision or hearing, but it gets information from the body. So there's information from the body that feeds into the insula. And according to some theories, this representation of the body state in the insula is what generates how you feel, your sense, your knowledge of how you feel. There are other structures here, the amygdala you already heard about in the context of Pavlovian fear conditioning and other parts of the frontal lobes like cingulate cortex, for instance. So it's a lot messier and more distributed than for the other systems we talked about. It's not like the hippocampus for declarative memory or visual cortex for vision. The neural substrates for emotions involve many different regions. And it's not too surprising, given that there are many different kinds of emotions, and emotions consist of all these different components. Sensory processing, then an evaluative component, components that link this to psychophysiological responses, and parts of your brain that have to do with the subjective conscious experience of emotions. And maybe I can do this. Okay, so very briefly on orbital frontal cortex. So there are sectors of the prefrontal cortex, orbital frontal cortex, and also probably regions dorsal to that, ventromedial prefrontal cortex, that are complex. There are many different regions of cortex in there, but collectively they all seem to play an important role in reward processing, emotion, and decisions based on emotions. This has been studied in humans and animals. One famous case that you may have heard about from 1848 is this case of Phineas Gage who was working on the railway in Vermont. We have a skull still which has a hole in it because he had a rather bad accident that amazingly didn't kill him. He was tamping down a metal rod to tempt on gunpowder into a hole in the ground for blasting and struck a spark. And this metal rod, the gunpowder exploded and this metal rod was shot up like a projectile and went through his head, like it's shown in this drawing here. So it was shot through his head and was a, you know, a lesion of interest to researchers. Surprisingly, as I said, the guy didn't die, but he did have profound changes in his behavior subsequently. It's been hard for people to know how to make sense of these, so it's not the cleanest description, but it's clear that there was, there were changes in his behavior, in particular changes in his emotional behavior. And since then, people have studied patients who have had tumor resections or other neurosurgical resections and they've done experimental lesions in non-human primates. So one experiment here done by my former colleagues at the University of Iowa, Dantranel Antonio de Massio, goes as follows. You put someone in a chair, you hook them up so you can measure an emotional response, skin conductance response, and you present them with stimuli that would variably trigger emotional responses. So you show them one at a time, not all of these together, the stimuli like what's shown in this collage of them. Some of these are neutral, like somebody driving or some landscape, and some are strongly emotional response, like pornographic pictures of people with their faces blown off. Well, when you look at these, when a picture comes up and it's like some terrible accident, you will normally have a strong emotional response. And so there will be a change in skin conductance response that you could measure in a controlled subject or a healthy subject. So it's shown here x-axis is time, the arrows are when a slide is shown, and t are these highly emotionally arousing pictures. And here what we're measuring from the left hand, from the right hand, is skin conductance response, sympathetic autonomic arousal. So it's pretty straightforward. Every time there's a picture there can be a little blip, but for the strongly emotionally arousing ones, there are big changes that you can measure, very reliable to do in the lab. Another key experiment that they did was to do exactly the same experiment in patients who had lesions of the orbital frontal cortex or a ventromedial prefrontal cortex. And that's shown here, they're completely flat. And it wasn't because these patients couldn't see the slide. So when you ask them, they can tell you everything that's going on the slide, they can see them, they have declarative knowledge about what emotion should be felt by the slide. So they would say things like, you know, I know this is a terrible slide. Most people would feel very upset at looking at this. And I know that I should feel something, but I don't. And so it's kind of a nice experiment that illustrates the role of this region of the prefrontal cortex in linking sensory representations, which are intact in these people, to the elicitation of an emotional response. They just can't get there because that link is missing. Similar kinds of things. A similar story pertains to the amygdala that you already heard about in Pavlovian fear conditioning. So this just summarizes what I think we already talked about. Here's the amygdala. I'm going to skip this part here for the sake of time. Okay, here's another video. So this is fairly striking to see. Here's an experiment that they did where you take a rat and you just have it forage for food. So very straightforward. But you can measure the behavior of the rat. You can measure how long it spends out in the box. There's a little food pellet there in the middle that you see. The rat comes out. They don't normally like open places. It grabs the food and it goes back in. Now what they do is they have a fear inducing stimuli, the stimulus that they can experimentally control. This Robogator that's sitting over there in the back. And now you can measure how the rat responds if you have this fear inducing stimulus come at it. So the rat comes out. It wants to go to the food pellet. The Robogator comes up and it doesn't make it to the food pellet. Now you can measure it. So this is a nice system because you can measure another latency that the rat takes to come out. And it's longer than it was before. It's very cautious because it's scared a little bit. Scared again. And now you can keep measuring this and it's going to take forever to come out. It's going to take a very long time. So you can measure a clear effect on behavior of this long-lasting central state fear that has been induced by this potentially dangerous stimulus, the Robogator. So it's still trying to come out but it's too afraid. So now you lesion the amygdala and you do exactly the same. So it's the same rat that this performed after with the same animals. And you can lesion the amygdala and they've made it even harder now for the poor rat because the food pellet is way over there by the Robogator. So this is a rat with amygdala lesions. It can see the Robogator. It can see the food. So there's nothing wrong with sensory perception but you will see that it has no fear response at all. It grabs the food pellet and goes back in. So a very striking demonstration that is similar to the conclusion that we had for over the frontal cortex. It's not that the rat is blind and can't see the Robogator or can't smell the food. So sensory processing is intact. Behavior is intact. It can move around but it lacks a part of the brain that would normally link sensory representations to an emotional state that would change the behavior. Last couple of slides here the remaining minutes. A former postdoc of mine who was at the time that he did this study at the University of Iowa with colleagues there studied a patient that we've studied a lot who has lesions kind of like this rat. So this is a patient who goes by the name SM. She's still alive and has bilateral amygdala lesions from a rare genetic disease. So it's like similar to first order to what we saw in the rat. It's a human brain with lesions to the amygdala and in this particular study what this person, Justin Feinstein, was the first author of this paper down here and now faculty in Oklahoma did was to use a whole bunch of different dependent measures to ask the question can this patient experience fear? So they had lots of questionnaires where you would indicate how strongly do you feel different emotions. We tried to induce emotions with films like horror movies etc. have autobiographical recollection and then took the patient into a pet store to have her handle big snakes and spiders. So many different ways of trying to assess fear and across all of these what we're striking in this patient is that she not only lacks Pavlovian fear conditioning as you would expect from the learning and memory lecture because of her lesions to the amygdala but also seems to not have any central state of fear as evidenced in conscious experience of fear. So it doesn't endorse feeling fear and if you try to show her horror movies or try to have a recollect traumatic events from her own life or give her big snakes there's no fear. By contrast all the other emotions were there. So it's not that she's just emotionally flat and doesn't have any emotional response which was which is more the case what you find with these patients with damage to the orbital frontal cortex is not that specific. In the case of the amygdala at least in this particular patient and also in the rap video that I showed you it seems remarkably specific for fear. So one thing to take home is if you had to pick a single emotion and it's linked to a specific brain structure it's probably the amygdala and it's relation to fear that's the clearest so far. No by no means totally unambiguous but the clearest but as I said she can express she can measure all these other experience all these other emotions normally. Here to finish with is a video of her this is the patient with these amygdala lesions and prior to going to the pet store she said that she would be afraid of snakes and wouldn't want to handle them she's never handled them before but when we got there she was extremely interested in all the snakes even the poisonous ones to be prevented from handling them and instead of any kind of cautionary behavior that would normally be triggered by state of fear what her amygdala lesions seemed to unmask was this unusual curiosity about wanting to touch and handle everything and you see very similar behavior if you make amygdala lesions in monkeys they also don't are no longer afraid of snakes which monkeys innately are normally but instead seem very curious about them so let me summarize this so emotions are not you know just isolated states or states that arise in certain psychiatric diseases but the idea is that they play a role in organizing behavior and cognition all the time fear has probably been studied in the most detail and been studied a lot in rodent models um and the amygdala of the brain structure that has been linked to fear if you had to pick one other brain structure for in a particular emotion it would probably be the hypothalamus and aggression from the other video that I showed you but if you had to say you know love or awe or embarrassment the link between those emotions and brain structures is really not worked out it's unlikely that there would be any simple link it's going to be much more complicated um the final thing to point out is so you might think from this experiment that I just show you this video that you know the experience of fear is somehow in the amygdala and that's not the case the amygdala again is one structure that can link sensory representations to the induction of fear it turns out that you can get to fear even in patients who have amygdala lesions if you do it through uh throughout other than sensory stimuli out there in the world so if you use an intraceptive in that same patient and others uh people have done experiments where you can induce fear or panic through intraceptive stimuli what you basically do is you give people a little gas mask to put on and you have them breathe carbon dioxide if you take a breath of carbon dioxide like that uh many people have a panic attack and it feels like you're suffocating so you feel like you're suffocating for about 30 seconds or a minute because your blood pH drops from the CO2 that you inhaled and that triggers particular receptors uh so and that happens even in these patients with amygdala lesions so they they have this sense of panic and a full-blown kind of panic attack when you give them carbon dioxide to inhale the amygdala seems essential to link representations of external sensory stimuli snakes tigers out there in the environment to a fear response but there seem to be other brain structure that are responsible for linking intraceptive cues to fear but this experiment shows that the fear can't all be in the amygdala the amygdala is just a structure that coordinates representations of external sensory stimuli to a central state of fear okay we'll stop there we'll have autism on wednesday and again just to remind you guys if you want to have office hours with me please send me an email i don't have to to arrange anything