Hey everyone, welcome to the Drive podcast. I'm your host, Peter Aia. Lisa, it's so great to see you and I know how busy you are. So, I'm uh incredibly appreciative of you making the time to sit down with me today. >> I'm delighted to be here. So, thank you for the invitation. So, as we just discovered a few minutes ago, um we both have this thing in common, having grown up in the same city. Um but it sounds like, uh we've both lived in our current
home longer than our previous home. >> Indeed. Indeed. >> Um I want to kind of understand when the first time you appreciated the idea of emotions as a as a topic for study. Well, when I was in graduate school, I so I I uh went to graduate school to I decided not to go to medical school and like all my other friends and I decided to go to graduate school to be a clinical psychologist. And so obviously right away you're starting to learn about emotion and
how they work and what they are and so on. And so I thought I had a pretty good idea. Um and I was um doing research for my PhD. um actually on a topic that was something other than emotion, but I had to measure emotion. Um and I did it in a pretty standard way, which is to ask people how they feel. Um and um what I realized was that um I wasn't able to replicate the experiment. So, usually in graduate school, what you do is
you take a published a a a research study that's been published in a peer-review journal. You try to replicate it just so that you can show that you know how to do an experiment and then you go on and you do your own stuff. So, I was preparing for my own line of work and I tried to replicate his experiment and I realized pretty pretty early on into it that the findings were not replicating. So, I'll spare you all the gruesome details, but uh over a
three-year period, I tried to replicate eight different eight different times I I tried to replicate experiments, probably a total of three or four experiments multiple times and wasn't able to do it. And um my first response to that was, well, obviously I'm not cut out to be um a scientist. Good thing I'll train to be a therapist. Um, and then I took a closer look at my data and I realized that while I was not replicating the published work, I was replicating myself over and over
and over again, which to me was an indication that something was interesting was happening. and um and luckily I was working with research adviserss who were willing to be persuaded by um my observations. So, I just can't resist pulling on this thread though in that I want to hear some of the details of this because I this was clearly such a pivotal moment in your career that it would be helpful I think for for me and I think the listeners to understand the nature of these
experiments, what was published and why you couldn't reproduce it and perhaps more importantly what it is that you kept reproducing that was distinct from the published findings. >> Sure. So usually when you ask people how they feel, you give them a set of adjectives or sometimes it's a set of statements, but some usually set of adjectives um um which are labels for emotion categories, angry, sad, happy, whatever. And you ask people, you know, how frequently have you felt this in the last week or how intensely
do you feel this now? Um and in particular I was really interested in anxiety and depression. Um and so I was also using clinical scales. Um in this case I was using uh a number of clinical scales but including the back depression inventory and the back anxiety inventory where you're asking you're giving people statements about um symptoms um and like I feel sad, I feel tired, I have no appetite. and you're asking people to rate the extent to which those symptoms describe um their experience. And
in the the studies I was trying to replicate were um basically this was a these were a set of experiments about the self about your self-concept um where if you compare your beliefs about who you are right now in relation to say who you ideally would like to be or who you think you ought to be maybe because your parents told you this or there's some you know standard that you think you have to meet that. If you fail, if you compare your experience of yourself
as you are to um who you ideally would like to be, you would feel depressed. And if you um were failing to live up to the shoulds, the ought as then you would feel anxious. This of course requires that people can distinguish between anxiety and depression as feeling states where I wasn't really dealing with clinical samples at this point. Um but uh one thing that I found was that um I wasn't able to replicate this effect because when people reported feeling anxious, they also reported feeling
sad or depressed. And when they reported feeling um happy, they reported an absence of anxiety and they would report feeling calm. So they were basically reporting I feel like or I feel okay even though the instruments I was using gave them the opportunity to be much more granular, much more specific. And now here what I was doing was what most scientists do which is to test a group of people and then look at the averages right of of the group and make inferences about it about
you know what you think is true based on what you observe um at the group level. Um, and the way that I phrase the question at that time was very much dictated by what I had learned up until that point, which is that, you know, anger, sadness, fear, depression, anxiety, you name it, emotions are biological categories that are distinct. And that people were having trouble identifying what emotion they were experiencing. And so my first way of phrasing the question was, well, I wonder what features of
emotion they're ignoring that, you know, is causing them to fail to recognize the emotional state they're in. And very quickly, once I actually started um I did more than just dip my toe into the emotion literature and read other people's summaries. I actually started digging into the papers myself. I realized actually that the phrasing of that question was in error, which meant something really interesting was going on. Um that an entire field seemed to be missing. >> So Lisa, I want to just recap something there
to make sure I'm understanding it correctly. Um what you just said actually answers a question I was about to ask, so I'll ask it only to make sure I understood. It seems to me that part of the difference was in the way you posed the question. Is that is that a fair statement? At first I posed the question the way that I had been taught to pose the question which is this assumption that there's an assumption right that this assumption is still very much with us
in certain circles that you know human infants are born with circuits embedded in some ancient part of their brain for anx for sadness for fear for anger for you know there are a handful of categories and that if you could find objective ways meaning not self-report. So, and we can come back to what this means objective because I think this is actually a big um this is a big >> a huge question. Yeah, >> it's a huge question. But the assumption was that objective meant not
the experience of the subject. >> Yeah. >> That you could measure something else about the subject. Maybe their blood pressure or their heart rate or you could look inside their brain or you could look at how their face moves or whatever. But that you could in some way determine because these are biologically distinct categories with separate psychological features that you should be able to um diagnose what emotional state a person is in. And if they don't report that state then that speaks to the fallibility of
it suggests that self-reports are very fallible um because you know there are these objective states. So the assumption is the there these objective states. They're biologically distinct. That is that the boundaries between these categories are in nature, right? They've been discovered by humans, not placed there by humans. And that there's some all instances of fear or anger or sadness, any category, but let's just take fear. All instances of fear should have some set of features that um make it fear and make a response fear and
not something else like anger or sadness. Um that is that that that all instances of that category should have an essence. So of course I wasn't phrasing the question with that degree of specificity but basically I was like well we know that these are universal inborn states. So if people aren't identifying them distinctly, there's something wrong with the identification process. And that's where I started. And in the process of trying to look for an objective measure to diagnose someone's emotional state, I realized there are no
actually objective measures. And that scientists have been debating this for 150 years. And I think at that point it would be more like a hundred years since I've been at this for a while. And in science it's often the case that when a debate um persists for a very long time and even before you know even really I would say even before the scientific revolution there were philosophers who were debating this. Um, when a question has been debated for a really long time, it often means
the question's wrong >> and that something else is going on. And because I hadn't been educated, schooled in theories of emotion and in the research, I came to it with like a blank slate, you know, I and I also came to it with clinical training. And this turns out to be, I think, maybe important um uh to the story because one of the things you learn to do when you're a therapist is identify when people are lying to themselves in your presence. >> That is when
they're telling themselves a story that just doesn't quite add up. And what I found was a massive story, a massive case of a story not quite adding up that there was a I think at the time I would have said the dominant view, but I think at this point I might say a dogma, a belie a a set of beliefs about emotion that were in the research literature. So what I discovered was that, you know, if you read the introduction sections of a paper and the
discussion sections, they say one thing, but when I looked at the evidence in the results section, to me it said something entirely different. And this set me on the path to um where I am really today. I changed my dissertation topic. I took an extra year to finish my dissertation because I completely changed my entire trajectory and did not become a therapist for very long. I was so in enraptured really by the research process and the discovery that I just focused um exclusively on that. >>
So how did the previous scientists explain variance in response? >> Well, they they do a couple of things. um with variance. So um I mean think about >> like you alluded to it earlier. You said basically they just took an average >> about yourself when you're angry when you're angry. What do you do when you're angry? Do you make a scowl when you're angry >> sometimes? I mean, yeah. Right. Exactly. Actually, research meta analyses, you know, statistical summaries of many, many studies of thousands and thousands
of people show that in western cultures, people, adults in urban settings scowl about 35% of the time when they're angry. That's more than chance. So, it gets you a publication in a good journal, right? But that means 65% of the time people are doing something else with their face. And half the time when people scowl, they're not angry. >> So if you use scowlling as an objective measure of anger, >> horrible predictive value. >> Yeah. And I might just add that's a lot of what AI
uh emotion AI is doing right now, but you know, that may be another topic, but you know, you'd miss you'd you'd miss 65% of the cases and half the time you'd be wrong. >> Call it. Yeah. >> Yeah. Actually, I what I would say is Yeah, that's right. Yeah. you you'd have both false positives and false negatives. And every category works that way of emotion. And it doesn't matter what measure you're using. It doesn't matter, you know, does the amygdala show an increase in activity during
fear? Yeah, more than chance. Yeah, about 30 30% of the time it does, but it often doesn't. And it you see amydala activation in many many many uh instances of emotion that are not fear and also across a slew of other phenomena. So so the the way that people deal with variance right what do you do when you're angry? Well sometimes you scowl but probably sometimes you frown and sometimes you might even smile or laugh. Sometimes you might remain silent, you know, and plot the demise
of your enemy. You know, what does your blood pressure go up, go down, stay the same? Yeah. It all depends on what you're about to do next, which depends on the situation that you're in. So emotions, when it comes to emotions, variation is the norm. People do a lot of different things in the same category. They feel Anger can sometimes feel pleasant. It's not always unpleasant. Anger sometimes distances people from one another but other times it brings people together and then there are cultural massive cultural
differences that we haven't even this is just like within one person o over situations right so I think part of what scientists do the way to think about it I would say is that science the world is complex reality is super complex and scientists simplify in order to study things and this is normal normal and one of the things that they they they simplify is variance. They're they try to often control variance as opposed to measure it. So one thing you can do in an experiment
is try to capture as much variance as possible and try to model that variance, try to figure out its structure and um and how to predict it. Another thing you can do is you can try to limit variability and that's what experiments typically in emotion up to that point had tried to do. So they would limit the number of categories being studied. They would limit the number of instances being studied. So the sampling of people, the sampling of um the stimula, the triggers to to provoke
emotion would be uh limited. Um the modeling specific the specifications of the statistical modeling would be designed in such a way as to remove variance or to or to get rid of variance as error variance. Um and when you start to look a little more closely um at the evidence, you can uncover that variance and you can even design studies to cultivate that variance actually to to try to observe it. Um, and in the end that's, you know, that's what I turn to were were methods
like that. >> So Lisa, I'm hearing two things that are related in question, but they seem distinct to me, and I want to make sure that I'm understanding this because it seems to be the foundation of what was the widely held view at the time you were doing these experiments. The two things are we should be able to objectively measure an emotion. And again, I think that's a very straightforward proposition um based on people's understandings of what's objective versus what's subjective. We should be able to
not we should be able to know your emotion without you telling us. Is that a fair way to say that? >> Yeah. What I would Yes, it is. But I just want to point out that always objectivity is within relation to the subject. Meaning you should be able to measure something in a subject that the subject doesn't tell you. Mhm. >> But I think it's important to also realize that measurements are never objective because scientists a scientists there's the synonym for a scientist's observation is the
scientist's experience. I mean sure you know you can quantify numbers. >> Well that's why we have to that's why we do blinding. That's why we want double blinded studies whenever appropriate. Yeah, but you know what? You can have double-blinded studies, but if you think that a double-blinded study actually removes your theoretical view from your study, you're deluding yourself because all experiments are theory laden. They're theory laden in the way they're designed. They're theory. That's not a bad thing. It just means if you recognize it's that
you can't you can never have an experiment that that samples um in a theory neutral way. You can never have an experiment that runs statistical analyses in a theory neutral way. This is a mistake. It's a mistake that scientists often make. Um once you realize that you yourself have a point of view and even though you try really hard, you can double blind things. You can, you know, there are a lot of things you can do to try to control the influence of your own point
of view, you can't remove it. And as a consequence of that, you can take other steps, additional steps to try to control it um that maybe aren't necessarily the norm. You know, >> can you give me an example? So, let's take something um let's take the following experiment. I have a hypothesis that I give you if I give you this pill, it will lower your cholesterol. So, I'm going to take a hundred people and without knowing which ones are which, half of them are going to
get the pill that I think will lower their cholesterol. Half of them will get a placebo pill. They'll take the pill for 3 months and before and after they start, we will measure their blood level of this cholesterol. So, what steps what would what steps would one have to take to remove as much bias as possible from that besides the blinding etc. >> What color is the placebo? How big is the pill? What are the other factors that are involved? You know, are your subjects what
are you controlling? Are you controlling what they how much they sleep? Are you controlling what they eat? Are you controlling? >> But does randomization take care of that? If I have a large enough sample, >> no. So, here's the thing. This is the thing that this is now, you know, fast forward 30 years now. Okay. But >> does randomization take care of it? No. Randomization takes certain variance off the table for you to model. But that doesn't mean that it removes the variance from the causal
complexity that is causing the phenomenon. And as a consequence of this, you can randomly sample a hundred people. You can run a study. You can have get an effect. that effect will not predict anything about any individual subject in your sample. The way that we get rid of variability by modeling by sampling large samples and large numbers of people and modeling them and then we try to replicate the average or the the summary. An average or a summary is a statistical fiction. >> It doesn't actually
let you predict anything about any individual. >> Correct. Yeah. And this I think is a super important point that of course people are discovering this um you know in personalized medicine and so on. But you know this this knowledge has been around for a good hundred years for sure. And there are reasons why scientists appeal to averages and summaries when what's real is variation. And what you so what you're doing when you randomize is you're removing variation as opposed to taking a hundred people and following
them over some period of time in their own lives and then you see whether the effect you get this whether the pill works you know by doing you can do a set of experiments where um you can you can still have control or I would say you can run your hundred subjects but you can quantify the levels of the the other factors that you think matter to whether the drug will work and then you realize what you're learning in that case is how the drug works
in a situation where those factors are held at that value and then you run another experiment where you set them at a different value. So, in this Makebelie um study, maybe the um taking the pill on an empty stomach versus a full stomach matters to its effectiveness. Or maybe um um you know, >> well, maybe there are genetic differences within the subject or you know, maybe it matters how much you've slept or maybe it matters. So, let me give you an example. The to me this
is like the like a great example. So there were these very very clever studies that were done um maybe 15 20 years ago um looking at um the response of people to um viruses. And so these scientists there were immunologists, there were psychologists, there were physiologists that they did is they took samples of people and each person they stuck them in a hotel room and then they quantified everything about that. how much they slept, what they ate, you know, basically how much, you know, mucus they
blew out of their nose. Like really, they were quantifying everything. And then they took the same amount of virus, viral load was identical, and they just placed it in each person's nose. And then they watched what happened and they observed that depending on the experiment there were about five I think five of these experiments done or something like that um between 20 and 40% of people showed symptoms of respiratory illness. That means the majority of people did not get sick from being exposed to a virus.
That means a virus is a necessary cause of a respiratory illness, but it is not a sufficient cause. >> But that's most of biology, right? >> Yes, it is. But if you if you stand up in a medical school and or at grand rounds and you say to a bunch of physicians as I have, do you think a virus is the cause of respiratory illness? They will like 90% of people will put their hands up. Why does that matter? And it's not physicians only. I mean
like it's like people doesn't really even matter what you're what you know unless you are somebody who studies complexity theory you probably think in terms of these simple causes but in fact I mean you you can't get I mean you can't develop with a the you can't develop the symptoms without the virus okay but the virus is if you're controlling viral load the virus is is not a sufficient cause there are other factors that are causal and th Those factors are also necessary, but they're also
not sufficient. And here's where it's a complex cause. >> Yes, for sure. So, there's there's two things there. Maybe the virus example is a bit tough because there's immunity, but maybe they normalized for previous immunity or previous exposure in the 60% that didn't get it. Um, but I still think this this line of inquiry around causality is relevant because if you can establish that the virus is necessary but not sufficient and that it plays a causal role, then the removal of the virus removes the condition.
Correct? >> For sure. But here's what I would say. in a large clinical trial. If you do a large clinical trial and you find that um um let's say I'm going to make up some numbers, but let's say um 30% of people get uh or have show improvement on a placebo. And so um that means that you need some other much larger number in your in your treatment group to to show an effect, right? Um, and let's say you only find like 40% of the people
um that the the drug worked for 40% of people. Your your and it's not statistically different from the placebo. Your conclusion would be that the drug is ineffective, but that could be wrong. >> Yeah. >> It could be that the drug is effective for 35% of the people. >> Yeah. and it's not effective for other people because there's complex causation and there's more there are multiple factors. So I do think that there is a problem when you're reasoning about the outcomes of individual people based on
>> averages >> group or averages. Yes. And this is you asked me you know how do people manage how do scientists manage variance? One way we manage variance is we ignore it. >> Yeah. We collapse it. Yeah. And instead we should be modeling it. And you give me any example of you know some I mean I I think there are lots and lots and lots of examples where you know I have my own personal examples um that I've talked about on other podcasts you know where
um uh you know which I could go in you know but where it matters basically. >> Yeah feel feel free to share one. I mean this is this is a fascinating discussion. A number of years ago uh I was diagnosed as many people are um with Hashimoto's disease and I went to see an endocrin so that's like you know where your immune system is destroying your thyroid and I went to an endocrinologist who was a great endocrinologist who's super super knowledgeable and he said I said
so what's the treatment and he said well we just wait for your immune system to destroy your thyroid I mean he wasn't being flippant about it but you know he said we'll wait till your thyroid is destroyed and then you start taking synthetic hormones and I thought Yeah. No, I don't think so. And so I did something that I often do when I'm dealing with my own physicians. I went to PubMed and I just started to read about thyroid diseases and a group admittedly I'm not
an expert but you know I have some ability to read papers and I found a series of studies which showed really clearly that if you give low doses of levothyroxine which is a synthetic thyroid >> you can arrest the progress these aren't clinical trials they're not large and they don't it doesn't work for everybody so I took I went back to him and I look at these studies. And he said, "As you would expect, as I expected, yeah, but there are no clinical trials." And so
then I explained the thing about averages. And then I said, "Yeah, but it seems like it works for some people." And maybe I'm one of those people. So let's do a little experiment. And he was game because, you know, I mean, it's not terribly dangerous to take a low. I mean, there are some things you have to work out, watch out for. Um um but we did it together because you know he listen he was very we were we had an open dialogue and sure enough
not only did it arrest the disease it actually reversed some of the damage >> and and sorry just to make sure I said does that mean that your antibodies went from positive to negative? It means no. It means that like nodules disappeared from my thyroid and you know so what it means is that the damage was the the the antibodies went down the damage was actually at first arrested and then eventually started to reduce a little bit and now um you know I'm kind of like
a borderline case. I was not a borderline case to begin with. It should have progressed but it didn't. And I'm not saying it's magic. What I'm saying is that um this is not something that everybody who has thyroid disease should do. It's not going to work for everybody. Studies already show that. But it will work for some people and it worked for me. And this is one example in many many cases where when you set aside the averages what the averages tell you and you look
for variation and you look for meaning. I mean it could have happened that it did nothing but that's not what happened in this case. That doesn't necessarily mean that um like I said that this will work for everybody and it won't and it might not work for me forever either. But the thing is though, now I'm on a low dose of levothyroxine and I'll be on that low dose for the rest of my life. But that's better I think to still have a thyroid and be
on this low dose than to be on a much larger dose and not have one for the rest of my life. I guess I just would to me this is just an example doesn't prove anything but but there are many many many many published studies which do make the point that averages hide meaningful variance and this is the case this is absolutely the case when it comes to the science of emotion averages or statistical summaries hide meaningful variance that is important in everyday life. So if
you construct your experiment in a way in a lab in a way to get rid of variance, you are creating a fiction that will not predict anything in the real world. >> Yeah, I couldn't agree with you. >> Let alone replicate, let alone replicate. >> No, I I mean I I agree with everything you're saying, Lisa. I probably on a personal level frame it a little bit differently and I'll I'll frame it back to you, but I I completely agree. Um I I have sort of
a way that I talk about this in the context of my world, right? So I talk about this thing called medicine 3.0 and I contrast it with medicine 2.0 and I say one of the limitations of medicine 2.0 is it homogenizes data which is just basically saying what you're saying, right? We the the the the bedrock of medicine 2.0 O is the randomized control trial where we take a very heterogeneous group of individuals, we apply a clean homogeneous treatment and then we homogenize what comes out
and that becomes the recommendation and on average that is an excellent way to practice medicine. >> Well, I would say on average it's a better way to practice medicine than the other ways for sure. >> Yes. Yes. But and then where I think we would agree is the challenge with that is the variability gets lost. There's nuance that get lost. So I'll give you one example from from my world. There's a drug called zedia or isetamide. It is a cholesterol-lowering drug and it has a very
specific mechanism of action. It works by blocking uh the one of the transporters on the cells of the gut and the liver that reabsorb cholesterol. Now when you look at the clinical trials of Zedia given as monotherapy it is rather underwhelming and we could without getting into the nuances of it there was a huge debate as to whether anyone should ever receive zedia and monotherapy um and it was viewed as not an evidence-based approach because it didn't have the hard outcome trial that said if you
give zedia and monotherapy you will save lives but once you understand the mechanism of how isettobe works And you understand that we can measure a person's cholesterol absorption using other biomarkers. We could selectively target people who are very high cholesterol absorbers and those are the people that would do very well in monotherapy. And lo and behold, we do this now. This is what I would call precision medicine or evidenceinformed medicine, right? And I think there's still a lot of push back to that approach. And I
would argue that what you've pointed out is another example of that, which is to say on average giving zedia as monotherapy probably doesn't save lives, but if you could target the people who are super responders. >> Yeah, for sure. For sure. For sure. So, let me just be really clear. I'm not saying we should throw out clinical trials. >> No, no, I didn't. And and I wanted to make sure that the listeners hearing you and I are on the same page. >> I think what you're
saying is very important. And to me, this is exactly where AI could make an improvement in how we interpret clinical trials because we wouldn't have to hide, we wouldn't have to try to compress variability. We could actually explore it more fully, >> I think. So that you're talking, you know, there are two things that you're talking about now. You're talking about, you know, what scientists, how science should be conducted and and then what AI can do or not do. And those to me are are somewhat
separate because it it all matters how you it matters a lot how you use AI. >> Well, let's put a pin in the AI. Let's come back to it because I want to talk about it in the context of your work as well. >> Yeah. But I what I will say is this that um in the science of emotion, there are no biomarkers. There are no biomarkers. There's nothing analogous to the example that you gave. Y >> and I but I would also say that um
and maybe this is what you know medicine 3.0 0 is I don't I don't know but you know but but I would say um we there are a lot of very brilliant people who if they put their minds to it could figure out other ways to do science that didn't require um using group averages. So what you describe here with this example is a great innovation to um patch up what is wrong with a clinical trial or use a clinical trial in a way that um
that um that minimizes the problems with it or the costs. And I think that's great. And again, I would say I wouldn't throw out clinical trials just now. But I do think it would be worth thinking about other ways to do science. We do have opportunities and options to do things differently if we decided to put our minds to that technology and the money that it would cost. You know what I often say to people is listen do you think that do you think that what
did it cost to build the Hadron Collider? Like how many billions of dollars did it cost? Yeah. Okay. How many countries participated to discover the Higs Bzon? It's because they thought it was important. Do you think I mean it I I would love to see what uh a bunch of what what medical science could do um with that kind of cooperation and that kind of money focused on one problem. I mean they built this m I mean I'm not saying it's it's good what they did.
It's just that that kind of ingenuity and cooperation and financing is not available c maybe it's available in some circles in medicine. It's certainly not available in behavioral science or in neuroscience frankly. And so they did something very clever. I think there are other clever things that we could do now. Um there is the technology that's available um now to start doing experiments differently. Um you just have to have the ideas, the will and the money to do it. >> Okay. I want to go back
to one other thing that you said at the outset that I I think is important for us to ground ourselves in which is part of what maybe we would call this foundational belief you came into. So one of them we've just talked about which is um the objective measurement of emotion. But then there's something else you talked about which is a belief that emotions are universal that emotions are hardwired into our nervous system. And and it sounds like you were saying, "No, I can't I don't
think that's the case." These two observations that emotions are not hardwired into our nervous system and that we don't have an objective way to measure them with a biioarker. Those are not those don't require each other to be correct. Those are independent observations, right? >> No. Exactly. And also, I mean, I I think it took me, oh jeez, probably a good 20 years before I was willing to say that there was no biioarker. I mean, you know, because I I had to learn a bunch of
literature. I had to retrain basically three times in order to um be able to make that claim. You know, I I think one of the things that I became very clear in my own mind about was that I wasn't going to be able to rely on the summaries of other scientists. I wanted to be able to collect the data and observe myself and I also wanted to be able to dig into their studies and understand for myself. So basically I was taking a very empirical stance
towards published studies. When graduate students sometimes ask me well how do you what do you do when you read a when you read a paper? I say well I read the few sentences that in the introduction that outline the hypothesis. I read the methods and the results and I throw the rest away. I don't read the introduction. I don't read the discussion. I don't care. I want to make my own judgment about what I think that shows. and then I will go back and you know
maybe read or maybe ask colleagues who I you know trust um because I'm very aware of how theoryladen publications can be. And so it's not something it's not like I ran a couple of studies in graduate school and came to this conclusion. I ran some studies in graduate school and discovered I had to learn how to measure facial movements objectively. And then when that didn't pan out, I had to learn physiology so I could measure physiological patterns to see if there were objective biomarkers. And then
when that didn't pan out, I had to retrain in neuroscience to be able to do brain imaging um to check that. Right? I mean, so I'm not these are not conclusion. Same thing about universality. We didn't I didn't just read a bunch of papers and test a bunch of college students. Our lab went to Namibia >> and we tested the Hima, you know, who live in um northwestern Namibia. Um we went to Tanzania and tested the Hudza, hunter gatherers. We went to rural China. We went,
you know, so of course we used other people's experiments, too. We used their data. We used their studies to draw conclusions. But we didn't only do that. We also went and did our own studies um um using where we used multiple methods and it's very very clear what you find. Um the findings are really really clear and they're clear across domains. It doesn't matter what you measure. Variation is the norm. There's structure in that variation. It's meaningful. It's not that we can it's not that it's
random. It's not that we can't say anything important about emotion. And also, I didn't come up with my own views on what emotions were or or how they work right away. I really spent really many years trying just to figure out the simple um answer to the question of are there biomarkers and is and you know are is there univer is there universality like and what is it that you know infants actually can do or can't do um and then I started to think about well
we have a real paradox here right I've experience emotion and I'm pretty certain about what I experience and I'm sure you do too and you and I could be um you know at a coffee shop having a conversation and I would I would feel like I was reading your emotions and you would feel the same. So how is it that this is that this is the case and then there are no biomarkers like this just felt like a really big puzzle. If it makes you feel
any better, Lisa, in in in the world where I spend all of my time, >> we don't have a biioarker for aging. >> Right now, >> think about that for a moment. Is there anyone listening to us that disputes that we age? >> No, of course not. We We are aging. Do we all have a gestalt of how somebody is aging? Sure. I can look at a person who tells me that they're 60 and I say, "Gosh, they don't look a day over 45." Okay, they're
probably aging slower at least in some dimensions. Of course, I can't see what's happening inside their body. Um, and yes, on average, we can measure things that are changing as a person ages, their bone density goes down, their muscle mass goes down, they become a little bit more insulin resistant, etc. But to actually suggest that at the individual level, I could measure anything in you, Lisa, and then do an intervention and then measure it again in a year and definitively conclude I have slowed your rate
of aging. Absolutely not. Now, there there are a lot of people that tell you they can do that. It is absolute nonsense. >> I mean, this is challenge. I mean, this is brutal. It's a it's a brutal reality. >> Well, it I don't I don't think it's a brutal reality. I think it's a reality and I think that that reductionist science isn't the method to get at that brutal reality. >> So, so does that mean you've given up on the biioarker? Do you believe that in
10 years we will have um biomarkers for emotions? >> No, I don't think it works like that. >> Okay. What I will say is if I measured you personally, Peter, >> uh, many many many, many times over many, many, many situations and I used a personalized AI for you, I could probably learn to reasonably predict the vocabulary of angers and sadnesses and fears and so on that you experience. That's what I think is plausible. >> You can't use that in a in an experiment you do
with a hundred college students, right? That wouldn't help you as much. >> Well, you what you would have to do is my guess is that that there would be some you could assess the degree of similarity between people. So, you and I grew up in Toronto probably approximately around the same time. And that means we're similar at least on that dimension. We both grew up around horrible Toronto Maple Leafs. Harold Ballard owned the Leaf. Like literally the worst team in the league. Okay. Yes, that's right.
We share a number of things in common. >> We share a number of things in common, you know. Um there are other aspects to, you know, that we share like we're both melanin challenged, right? We don't have dark skin. Um well, there are differences obviously, you know, you're a man, I'm a woman. um we have you know so but the point is that there are some features on which we're similar and there are some features on which we're different and we it's an empirical question as
to which features matter um to be able to use to be able to generalize to find you know well you know you can look for similarities in the algorithms that predict individual people and and potentially inductively discover um how uh whether that similarity exists and and and how to capture it and model it. So what I'm saying is that it's possible that that there could be more generalized algorithms, but you would not start with the simplification. You don't start with the simplification of getting rid of
individual variability. You start by modeling individual variability and then you look for places where you can reduce it on the basis of empirical evidence. that is a justified way I think it's just a much more cumbersome and expensive way to do the research and it has huge implications for um the incentive structure in science would have to change and a lot of things would have to change if we did research that way on a large scale but it all it all comes down to what people
care about and what they want to invest their research dollars in and you know that's not for me to say that's above my pay grade. >> So what what can we say about an emotion if you were now trying to explain to somebody what emotions are? We've we've said two things that we know on the negative. We know that they can't be measured objectively and we know that they are not necessarily hardwired into each of us as identical states. Fear is >> Let me ask you
this question. Yeah, let me ask this question, Peter. What is hardwired into the brain? >> What is hardwired? >> Yeah, it's a great I mean I think certain reflexes are clearly hardwired, right? So, for example, if I shine a bright light into your eye, I think it is hardwired that your pupil will constrict. Right. >> It will. But how much will it constrict? And you know, even reflexes which are hardwired are are not they don't always >> the degree to which Sure. Yeah. >> Yeah. No.
Well, but that matters I think. So what I would say is that there is very little in it's just like genetics right I mean there sure there are some cases where a single gene predicts a single characteristic you know with a single alil right I mean but those are really rare most of genetics doesn't work like that and I'm certainly not the first person to point out that you know NIH spent a tremendous amount of money um trying to map the human genome on the assumption
that that you know >> would actually if we just knew the genes we'd we'd be set a lot of the devil is in the details of how of the mechanisms right because of this complex causation um that happens and I would say the same is true um in the case of emotion and pretty much the case of anything a brain does. So the first of all I would not when someone asked me how do you define emotion I say I don't you can't how do I
def I could the way I would define it would be generic it would be like well any mental event basically could um you know they your brain's most important job is regulating your body coordinating and regulating your body in the most efficient way that's why brains evolved they evolved for they evolved at the same time as um invertebrates. I'm only speaking about vertebrates. I don't know anything about invertebrates, but invertebrates, they evolve around the same time as you get visceral organs and and you know, a
head and distant senses and eyes and and ears and so on. >> Let's probe that a little bit more. So, so this I think you you said something in passing that that is a pretty big idea, right? You said that the the purpose of the brain is to support the body. Can you double click on that a bit more? >> Yeah. So I wouldn't I'm I'm trying not to be ter I mean I was a little sloppy and a little teological about it. So I I
would prefer to back up and say like the brain's most important job. I'm not saying it's it's that's its sole purpose, but I think >> um that um its most important job is regulating and coordinating the systems of the body and everything I think everything it does psychologically speaking it's doing in the service of that and that means that metabolism writ large is important to things like feeling and thinking and remembering and and so on and so forth. So that's a key assumption in our lab
that we use when we're designing experiments. Now, um, and so what I would say is the brain is always receiving signals from the sensory surfaces of its body. Your brain isn't running a model of the world. It's running a model of your body and the sensory surfaces of your body. And of course, some of those sensory surfaces take in signals from the world. some of the signals from the world, but it only knows the world through the body. There's no other way to know a world
than through your own body. So your brain is running a model basically of signals from the sensory surfaces. It's attempting to predict those signals and give them meaning because those signals that it's receiving are the outcomes of some set of changes that the brain doesn't have access to. It only has access to the outcomes, the signals themselves. And that's an inverse problem. It means the brain has to infer what caused that loud bang, what caused that flash of light, what caused that tug, what caused that
change in glucose. It has to make inferences about that. And sometimes those inferences are emotions and sometimes they're not. >> Okay. I'm going to ask a question that >> I I never yeah I I this is a question I personally struggle with which is >> how do we define consciousness and how does consciousness overlap with emotion. >> So yeah >> I so I could see the look in your face. I intuited your emotion when I asked that question and the veilance is negative. Why? Well, what
I would say is that why is it because I think um I think that there are a lot of questions that you could answer about emotion without having to take on the big question of consciousness. But um but I would say um the following I guess that um I don't know if it's that we're asking the wrong questions or asking questions in the wrong way but um a lot of people have a lot to say about the nature of consciousness. None of which actually explains um
how it is that we are sentient and how it is that we have content that there's content that we get content out of out of electrical and chemical changes. And um I just prefer not to speculate. It doesn't mean that I'm not interested. It doesn't mean that I I'm not curious. I'm I'm very interested and curious, but I also am not a fan of talking out of my ass, you know, like I just um I think that we don't know enough maybe yet to ask the
right questions or whatever. So, we should continue to explore and debate and whatever, but you know, like people publish books and they make claims about consciousness. So, here's what I have to say. Emotions, um, are there aspects of emotion that you're not conscious of? For sure. That has to be the case. You're not conscious of most of what goes on inside your own body. Thank God >> because if you were, you'd never pay attention to anything outside ever again. Just think about the last time you
ever had like gastric distress. I mean, you just can't, you know, you can't focus. It's there's so much going on in there, right? And we don't have access to most of it. Most of the details are compressed out of the signals before they reach the brain. What the brain is receiving about metabolic and interceptive signals from the body is more like a gist. It's not getting specific details about everything. It might compute details like it might infer those details but like there's no visceral there's no
visceral map of the body in the brain that anybody has found. Some people claim they find it but when you look closely they don't. It's not like a retinopic map. Right? So your your brain has V1 primary visual cortex has a map of the retina. There's array to array mapping meaning you have an array of cells in your retina you have an array in your brain they map the mapping changes all the time right so uh what a like we would say the um the um
receptive field of a V1 neuron a neuron in primary visual cortex changes actually with the context of with the neural context of the brain but it's always the ma it's always a map right but there is no map of the body like that in your brain. Just like there is no map of your olfactory system. There's, you know, you have old factory receptors in your nose. There's no map. It's no spatial map in your brain. That information is compressed out of the signal before it ever
reaches your brain. There's temporal information, but not um not spatial information. In the body, the the brain has like it can tell there's some spatial information like is is a signal coming from the thor upper thorax, is it coming from the gut? You know, a lower it's not even gut so much as like lower part of your abdomen, upper part. There's like a little bit more detail here, but it's nothing like a retina or a coia where you have array to array mapping. This turns out
to be super important. Um um you know we in science people talk about extraceptive versus interceptive like signals that come from the world versus signals that come from the body. That turns out not to be so important. I think I think what's more important is is it a proximal sense or a distance sense >> meaning the the distance senses the best examples are vision and audition. they're processed the the signal processing the compression of those signals is very different in the brain than say olfaction which
is a proximal sense or gustatory or anything interceptive. Gustatory is usually considered um interceptive um taste is not but gustatory you know signals from the gut are interraceptive. So and then there are a bunch of signals that kind of don't fit the extra receptive in the internal external distinction like propriception is that internal or external it's kind of both right so if you just look at the signal processing the signal processing properties then other distinctions become important and I would say it's the same thing about
cognition and emotion or thinking and feeling there aren't a lot of distinctions that you can find in the brain or anywhere else to distinguish what is a thought and what is a feeling. But there if but if you look at the signal processing you know architecture and dynamics other distinctions become important for explaining how a thought arises or how a feeling arises. None of those explain how you get from you know, um a very very highdimensional set of chemical and electrical signals to a very lowdimensional
um you know um mental feature like um pleasantness or unpleasantness like nobody has that. Um and if they're claiming to, which some people claim they do, then they're lying either to you or to themselves. probably to themselves in your presence. >> So let's try to disentangle without defining an emotion some of the broad differences between say the positive and negatively veilanced emotions. So again we'll go back to this idea that there's a belief that god some of these are really primal like I'm going to anthropomorphize
this as well right? So fear is so essential, right? Without fear, we could not have evolved as a species. We couldn't have figured out a way to avoid predators or things like that. Um, so let's just start with fear. What is it that that you could say today we know about that emotion? >> I'm just going to let the point go about we couldn't have evolved without fear. It's a primal emotion. >> And by the way, feel free to attack that point. By by the way,
>> I don't want to attack that point. It's just I don't think that's going to take us down a productive avenue. But here's what I'll say about fear. There are no pleasant or unpleasant emotions. You think fear isn't pleasant? People pay money to go on a roller coaster. They pay money to go to a horror movie. I used to run a haunted house in my basement for charity for 13 years, an auspicious number where we scared the out of people. [laughter] We donated all the money,
but we scared we basically I used to [laughter] tell people I scare little children for for I scare, you know, I I have their parents pay me and I scare little children for money and then I donate it all to the Boston Food Bank, which is what we did for years and years and years. And we had three levels of scary, you know, um so people could choose and and um and all of my graduate >> is that a uniquely is that a uniquely human feature?
Is there any evidence that an that anything else in the animal kingdom will voluntarily put themselves in a situation that produces fear? >> Yeah, absolutely. I mean, there are an there is a there's a whole literature on how animals play and um and they're often playing in ways that would, you know, that could be considered practice. Like it's a way of foraging for information and for like innovation basically. um whether they are feeling fear I can't comment on that but what I would say is that
you don't have access to whether an animal what an animal feels you you're if you experience if you look at an animal and you say that animal is fearful that's your experience of what those signals mean out of that animal and that doesn't necessarily mean that it tells you anything about what's actually going on inside that animal from a feeling perspective however I think you can say Like if you watch an animal work hard to avoid certain stimuli um that will cause it damage, right? Or
if you if an animal is willing to like expose it so you there are these studies, you know, where they'll like damage an animal in some way and then the animal will actually work hard and even expose itself to the possibility of a predator in order to get an opioid. That's a pretty good indication, I think, that the animal is feeling something like pain. Maybe not the way we feel pain, but it's definitely feeling something unpleasant because it's willing to incur a cost, a significant cost
in order to um um to get a drug that will reduce presumably its awareness of the maybe awareness is the but you know it's feeling in relation to some kind of tissue damage. Um that's very different than what like a fly will do you know where it does you know you damage the leg of a fly or a wing of a fly and it will still continue to behave in exactly the same way. Um then you know you you know you can sort of make inferences
that way which I I so I'm not talking about pleasure and pain. I'm talking about the specific the specific categories and distinctions that we make about humans that come from human experience. You have to be super careful when you impose those on other creatures to which for which we have no access to what they experience and not so I'm not advocating to assume that they experience nothing. I'm saying that you just have to be careful and there are lots of reasons for being careful both for
them and for us. But what I would say about fear is that you know what do you think the the um the what do you what is the um what is the what would you think is the prototypic response in fear um to freeze or to run away um well you know >> I think it varies and and by the way >> well sure it does but I mean the theory is in the literature if you ask most people fight or flight you know they're like
flight, flight, they'll freeze. An animal will freeze, you will freeze or you will um flee. But there's a um you know a culture where what people do when they feel afraid is they go to sleep. That seems weird to us. But you can look in the animal kingdom and you can find out that in invertebrates in the water animals when there is a predator will play dead. They will slow their heart rate so sharks can't find them. They will not move. They're not freezing. they're reducing
their um kind of um detectable output so that the predator can't find them. So I I think um I guess what I would say is when it comes to fear, what we could say is that variation is the norm and that that what you feel in fear, what you do in fear depends on the situation that you're in and um what you're going to do next. Because your brain doesn't react to things. It predicts and those predictions always begin always begin as motor predictions as visceral
motor predictions predictions of what to do with the visca to support skeletal motor predictions whether it's to move your limbs or just to move your eyes and your head or to not move at all. All predictions begin as motor predictions and copies of those motor predictions become the sensory the um the expected sensory consequences of those movements. So your brain is preemptively so what is an emotion? It's a a set of predictions that tells a story about what a set of sensory signals mean to you
in that situation. But that's not a specific answer to emotion versus cognition versus perception because that definition encompasses all of those. I could say, well, emotion is when there's some really big change in um it's it's what I just said plus there has to be some really big change in um you know, your level of energy or arousal. Um has to go up or down a lot from where you are right now in order for you to experience something as an emotion. And that would hold
that that aspect would also hold for many many people around the world but not everybody because in some cultures emotions and cognitions are not separate. In some cultures emotions and bodily states are not separate. They are considered to be a single set of categories. >> Can you double click on that a little bit so I understand what you mean? um the bodily state and the emotion for anger would be the same for every person. >> No, no, no. What I'm saying is that in some cultures,
so in our culture, we make a very firm distinction between is this an emotion or is it a physical state like is it you know do you have a stomach ache or or are you angry? >> Right. But when I'm angry I can feel something in the pit of my stomach. So what is the difference there? >> Sure. But you wouldn't look I mean there's look there's neurology there's internal medicine and there's psychiatry and what's the origin of those of those disciplines the origin of those
disciplines is where they think the cause is of the problem and there's there are diagnostic categories that that you use when you can't figure out medically what's wrong with someone but you know they're having some what you know difficulty in with mental features or what we would call psychological features. Right? And um the assumption is that it's all in their head, meaning it's not in their body, right? I mean, I'm not making this up. This is this is how things were for a very long time.
Um and what I'm telling so um so um you it's very sure it's the case that um like when you're anxious you you might have um you know um you know your heart might race or um or you might have gastric distress or whatever. But if you think gastric distress is the problem, you'll go to an internist. And if you think that your mood is the problem, you'll go to a psychiatrist when in fact they are related. So I'm not talking about what I think is
the case or what you think is the case. I'm talking about what a lot of people, including a lot of physicians, think is the case. Um, you know, physicians, I've had physicians talk to me about how depression causes heart disease and heart disease causes depression. Or another possibility is that they're both rooted in some met set of metabolic problems and one is not causing the other in the way that A causes B, like a mechanistic cause. It's that they have a common cause that makes both
of them more likely. That's a very different way of seeing things. And so what I'm saying is that there are some cultures who wouldn't say like in our legal system we have thoughts we are you can be culpable because of what you feel or you could be have culpable because of what you do and you there are different punishments for that. But there are some cultures where thinking and feeling are the same. They are features of the same state. So you wouldn't call something a you
wouldn't distinguish between a thought and a feeling because they're both it's like you have thought feelings. There's just it's not one or the other. They're they're both um you know it's like blue and green together make turquoise. So you wouldn't say it's blue or it's green. You would say it's turquoise. It's both. Um they're not distinct states. Similarly, there are cultures where there are somatic descriptions for emotion like that. That is that is the that is the nature of emotion in those cultures. So we need
to have a theory of how a brain and a body together in a world full of other brains and bodies uh create a mind and that has to work for everybody on earth. It c it's not just to explain western categories and that may mean that emotions and and thoughts and and perceptions are things we have to explain but they are not themselves explanatory categories. They're not categories that explain anything. Are you saying then that the that the universal theory that is culturally agnostic is that
the goal of the brain is to well there's two things I've heard. One is the a main priority of the brain is to support the body physiologically metabolically etc. And two, the brain is a predicting machine and emotions are arising out of predictions of physical sensations or the consequences thereof. Is that directionally what you're saying? >> Yeah. I mean, I would never claim that it's um it's un culturally neutral because I'm, you know, can't claim that. I'm I live in a c I was pickled in
a culture just like you and there are other but you know I think um I think that it's a reasonable um it's a reasonable >> but if it's biological it should I mean it doesn't matter what culture you live in there are certain biological things that are true across the board. I'm being a little humble here, Peter. Biology in western science. I'm not I'm not immune to I don't I don't think that I'm just being I don't think it's I don't think it's relevant to say
I don't I wouldn't claim trying to think of how to say this without being too negative. Um I'm very aware of the fact that some of the beliefs that I hold as a scientist um come from the fact that I'm was raised in a western culture and I use western concepts and methods of science and not everybody you know adheres to that. But I would say that I can defend that statement by looking at anatomy, by looking at physiology, by looking at brain evolution, evolutionary biology.
I can defend it by looking at um developmental neurobiology. I can defend it by looking at signal processing in engineering. I think that um when multiple fields come to the same conclusion using completely different sources of evidence and they don't talk to each other all that much, then you know that you're probably on to something. The idea that the brain predicts and doesn't react is is actually a concept that is in other cultures and it's been in our culture for many many many years like like
centuries. >> But this is very counterintuitive. So I really want to spend some time on this because I've always felt that my brain is just a reaction machine, right? like I'm riding a I'm riding a bull and that sometimes that bull is just walking along and everything is fine and sometimes it's bucking like crazy and my vertebrae are getting crunched. >> So yeah, >> let's maybe in any way you like let's go through the different lines of evidence as you've laid them out from different sort
of empirical disciplines to how we can say that the brain is is not reacting. >> Sure. Sure. I mean I just before I do though I just want to say two things. One is so for you it's a bull and for John height I think it was an elephant and for Plato it was two horses and a charioteer you know like that idea that is that you know that we have an inner beast that we have to tame and kind of control is wrong it feels
that way. It feels that way to me but it's not defensible. It's not a great metaphor because it doesn't really reflect doesn't really reflect the evolution and anatomy of the brain for one thing. Um, so that's my f that's called the try brain and it's just you know it's really really popular but the evidence against that has been in molecular genetics for at least 50 years if not more. So that's just we just have to take that off the I just finished doing a video actually
with a videographer just kind of do like um to explain um why this is not the case and show for the first time you know a depiction of what um of how um a common kind of manufacturing plan for a a vertebrate brain um could produce all different kinds of brains in vertebrates. really it's just mammals but you could make inferences to vertebrates I think in general the data is just about mammals but but you can get a mouse looking brain and a humanl looking brain
out of the same um exactly the same manufacturing plan in an embryo um that's not my work that's work by Barbara Finley and her colleagues um so that's the first thing I want to say and I can't even remember what the second one is >> well tell so but tell me why My feeling of >> Oh, your feeling >> of this of riding the bull is >> a question. That's a super interesting question. Yeah, that was my second Thank you. You that was good. Thanks for
that assist. Yeah, my second point is that um your feeling isn't evidence of anything and neither is mine. >> Yeah, could couldn't agree more. Yeah. >> So, so I think brains are amazing, right? Because they're like masters of deception. They create experience kind of seamlessly. that experience reflects nothing about how it they work. But yet, you know, we're they they're very tempted to use it that way. I was like, so I think that why is it that we experience things that way? That is a great
question. That is a great question for science. Somebody should study that because it isn't really good evidence but it is a phenomenon that is worth studying like where does our so some people study like where does a memory come from and where does the certainty about memory come from because they're not they're not the same they're not the same source um and I would say the same thing um about this um so why do we take our experience um to evidence of how things work. I
don't know why. It's an interesting question, but it is naive realism. It's naive realism in the sense that your experience doesn't reveal to you how many things work. And all science, all scientific progress is a slow march away from naive realism. And the same has to be true for neuroscience and the brain. Same has to be true for neuroscience and the brain. >> Let's go deeper into this idea that the brain is not a reacting machine. What is the what is the what what are the
threads of evidence? >> Well, I think first of all um if you look at how a brain is structured. Okay. Um so let's just go back to um primary visual cortex. I think this is a really good example. Um, and I should say I just published a paper that summarizes a lot of this evidence in one place that was I published with Earl Miller from MIT in nature reviews neuroscience. And most of this is not my own work. This is, you know, me summarizing with Earl
other people's work and also some of his electrophysiological work. >> We'll we'll link to that paper in the show notes. >> Yeah. But so let's just take V1. I think this is this is really interesting. So if you look at the synapses, so what? So primary visual cortex is called primary visual cortex because um fibers from the retina pass basically they they run along the optic nerve. they pass some of them will um terminate in a a subcortical area called the superior caliculus but most of
them proceed to the phalamus and then to V1 to primary visual cortex. The assumption is that those signals are what drive the neural activity in V1, drive those neurons to to activate to to have action potentials to to produce electrochemical changes. Except that 95% of the synapses in V1 neurons come from other places in the brain and not from the phalamus. other places like other cortical areas and not from the phalamus. So most of the signals are not bottom up from the retina to the brain.
They are from the brain to other parts of the brain. They're top down. That's one um interesting piece of evidence. So just to make sure I understand it, the visual cortex is getting most of its input not from the optic nerve but from other areas of the brain that presumably are not receiving light. >> Where are those areas and what type of data are being transmitted? >> Okay, so here's one. This is like one here's one example. So, I'll just say, you know, like trigger warning,
if anyone is unhappy or upset by um discussions or descriptions of studies that harm animals, you know, they might not want to listen to this because it's kind of a brutal experiment. Um, but there was an experiment done where they took mice, I think it was mice, it could have been rats, but I think it was mice, rodents, and they put them they had them run on a wheel and they recorded from uh all over the brain. They they sunk electrodes into the the rodent's brain
and they recorded the neural activity, the electrical activity in the neurons in many places including in V1. Okay. Then they abladed the the um the retinas of those animals. They destroyed the retinas of those animals. So no light now is getting into the brain. V1 neurons quieten down for 24 hours and then they ramp back up to a normal level of activity. >> Sorry, let me make sure. So So but the animals are now blind. >> Yeah, they are. But V1 is is the recordings are
back to a normal level. And what is driving those neurons? >> And what is the output? Like what? >> Well, those the animals can still move around. I mean, they can smell things. They can still move around. >> Yeah. Because you just you just destroyed their retina. >> You destroyed their retina. So they're so look primary visual cortex is supposed to be it's name I mean it's an occipital you know it's named as a visual cortex that's its name >> presumably it's supposed to be processing
signals that are related to light from the retina but animals whose neurons whose um retinas have been ablated have normal levels normal patterns of activity in V1 after the brain adjusts and those are related to movement. There are parts of the brain in the cortex like for example um in in mid it's called anterior mids singulates in the in the um mids singulate which is um does many things but one of the things it is um viscero motor and it's a viscot motor regulator. It regulates
the viscera of the body and it also is an association area for the skeletal motor system and it projects directly to V1. So does this lend credence to the idea that when a person loses a sense they have a heightened uh sense elsewhere? It um lends credence to the idea that that um signals are predictive that most of the signals in the brain are actually intrinsic signals that are predicting things those signals. So what's happening in in that part of the in the singulate is neurons
are projecting layer five in the cerebral cortex layer five neurons are projecting down to the subcortical areas down to the brain stem to regulate the body and literal copies of those so so collaterals off those axons go to other parts of the brain as sensory predictions which are the expected sensory consequences of those movements. What it means is that an action potential doesn't have an intrinsic meaning. Its meaning is determined by which neuron sends it and which neuron receives it. Exactly the same signal if it's
received by the spinal cord is a motor signal. Exactly the same signal if it's received by V1 is a sensory prediction signal. It's exactly the same signal. Its meaning however physically it's the same signal. Its meaning is different depending on who receives it, who sends it and who receives it. It's relational. So what this means um I mean for example there very clever studies where they take humans and they put blindfolds on them for a week, okay? And they teach those humans to read Braille. And
as they're reading Braille, V1 is, you know, going nuts. It's like definitely processing signals. And so you think, aha, but are those signals actually important for like reading Braille? Like, do they have any, you know, any kind of spatial temporal meaning? Yes, they do. Because then the experimenters they use TMS to they they they inject electric you know magnetic electrical electromagnetic signals into the brain that disrupts V1 and their ability to read Braille is disrupted. So you've got the causal link right and then you a
week later take the blindfold off. V1 goes back to mostly processing visual signals. So what I'm saying is that um that first of all that what we think of as you know brain regions that have very very very specific meanings um from a psychological standpoint probably don't. They certainly have specific they certainly have specific signal processing capacities based on their architecture and what they're connected to and so on. But um a lot of the distinctions that we make like I met a woman once who had
no smata sensory system at all like it was just congenitally not there. Um but she was able to eat only by looking at herself in a mirror while she ate. So, she was using visual signals to control her motor movements. >> Meaning, if she tried to eat without a mirror, she wouldn't know how to move spoon. >> No, she wouldn't. Yeah. She wouldn't know how to swallow. So, sensory signals are not so that we know the world. Sensory signals are in the service of controlling motor.
They're in the service of motor control. They also let us know the world, which is relevant for motor control. So I'm saying that, you know, when you start to look at the architecture of the brain, and we've only talked about architecture. We haven't talked about any of the other types of evidence yet. Um, you know, you start to see things differently. You start to realize, oh, well, actually, so most of the synapses are predictive synapses. And that's interesting. and um that um that motor signals, copies
of those signals are going everywhere and they're beginning to construct experience before the sensory other sensory signals arise. Like if this were a different format, I could show you an image, an illusion that you would experience as red. You would literally experience red and there would be no wavelengths of light in that in that range that were hitting your retina. But you would experience red. So you would be like predicting red, remembering red. How >> your experience of red? >> What do you mean how >>
how would you make that happen? >> What would you do? You're you're saying I mean you could do that obviously by stimulating my brain in the right areas. Is that >> Yeah. No, I don't have to stimulate your brain um with electrodes. I could just stimulate another way. So, what you do is you you show people um well, I'm going to ruin it now for everyone who hasn't seen it, but you show people um a stop light, like a >> like a traffic light, red, orange,
green, except the red one where red is supposed to be is gray. >> There's no there no wavelengths of light. There are no long wavelengths of light coming from that spot. But because um you've had countless numbers of statistical like there's been a countless numbers of statistical regularity for you know facing this array of signals. Your brain predicts that that's red and you see red. You see red until I move in two gray boxes and sit them right beside and then all of a sudden the
light looks gray and then I pull the boxes away and now it goes back to being red. This is something I've never had. I've had anyone not experience this probably but they haven't told me. I mean it's one of the I have an I have an illusion where I have that I got from somebody who made that where it's um electrical towers are playing jump rope. So these like electrical towers are playing jump rope. It's just a video but you can hear the towers hit the
ground. You feel the vibrations in your chest. Those are all prediction signals. Prediction signals are not like abstract things. They are um they literally it's literally your brain is using its own intrinsic signaling to change the firing of its own sensory neurons in preparation for signals that it expects to receive. It's acting. You're acting in the world that will generate certain sensory signals and it's expecting those signals. And when those signals come, the neurons that are already firing in a pattern to um expect those incoming
signals, they stop firing and you have an experience. So the other interesting thing, this is this is Earl Miller's work. The only thing that's making its way from say V1 to the middle of your brain or the front of your brain are signals, visual signals that the brain didn't expect because the ones it did expect have already quiet have they've already kind of canceled out the um the neurons the prediction signals that were correct. So you don't react to in the world. stuff in the world
is there to confirm your predictions or correct them, not to react for the most part. >> I'm still struggling to wrap my head around this and I now realize why you you said that sometimes you're still struggling to understand your own work. Um, let's go from some not not to minimize the red light experiment. Um, but let's now take that to an emotion. >> Okay, before we do that, why don't we take it to baseball or hockey or something like that? Like every sport, every sport
would not exist if brains were not predictive. Like there's this really great example that's been worked out for professional baseball. I wish I had a hockey example in my pocket, Peter, but I don't. It's the unfortunately it's baseball. >> That's okay. We love We love We love us baseball over here. >> Okay. So, you know, >> hitting a baseball presumably >> hitting a baseball. Yeah. So, if you look at a major league diamond, the size of a major league diamond and you look at the speed
with which a major league pitcher usually pitches at, like 90 95 miles an hour, there is absolutely no way that if a batter waited to see the ball, he would ever be able to hit it. There's just literally not enough time. What's happening is that based on who that picture is, where they are, what the temperature is, what the wind is, how much he ate earlier that day, how much sleep he got based on all of these things based on a huge ensemble of signals. His
brain is making a prediction about where that ball is going to be in a moment from now. And that is where he swings. He's not doing it consciously. Where do predictions come from? They come from me. Predictions are just memories. You don't you don't experience yourself remembering anything. But basically, your brain is reimplementing signal patterns from the past for the purposes of doing something in the present. That's what a prediction signal is. So every experience you have, every action you take is a combination of your
remembered past and your sensory present. The sensory present is there to confirm the remembered past or to change it. And we have a special word in science for changing the remembered past. We call it learning. That's what learning is. >> What accounts for the differences between individuals? Why are some people better at prediction? So, um, let's use baseball as an example because it's a very it's a very concrete example. So, what made um what makes Show Otani or Aaron Judge better at hitting home runs? Like
presumably they're not aware of it, but but is it that they saw more reps? Is it that their eyes are better at picking up the queue sooner as the as they see the hand of the pitcher? like um what what do you think? >> So I I don't think anyone knows the answer to that question, but um certainly it has something to do with experience. It probably doesn't only have thing. It's probably not only sure um experience, but certainly lots and lots and lots and lots
of practice and repetition for sure. That's how every skill is built. Like think about exercising. So this is another example that I find really interesting. Like if you wanted to get faster at running a mile, you would run that mile over and over and over again. You would you would train to to do the same thing and you would do it over again as many times as you needed to to get to the speed that you physically could manifest. Basically, that means you're becoming more efficient,
more metabolically efficient with practice. That's what athletes do. But what if you're doing what if you're exercising not to um what if you're exercising to burn calories or what if you're exercising to remain healthy? Then you do something like interval training where somebody's calling out to you every 30 seconds. You don't know what they're going to say. It's changing things up every 30 seconds. That's much more metabolically demanding way of exercising because your brain can't predict what's coming next. So you're always adjusting, adjusting, adjusting, adjusting
learning um processing prediction errors is more metabolically expensive on all things considered than than um when you um your motor actions arise from your predictions. So every skill is built by practice. You can't become skillful at something without practice. But that doesn't really account for why some people need more practice than others. Like I no matter how much practice I ever had, I will never play baseball like a professional, you know, athlete would play baseball. Um so I think again I would say the value, Peter,
of um of a the new theory is not that it explains everything. It's that it it it asks questions that were have never been asked before and it sets aside some questions as you know sometimes progress is asking better questions it's not answering old questions so I don't know the answer to your question but um what I will say is that practice wi you repetitive feed you have to have practice and you have to have concrete repetitive feedback practice alone without without um concrete repetitive feedback
is not usually um useful for building a skill. And this is also true about emotion. You know, if you want to become more granular in how you experience the world or how you experience other people, you can learn to do that just like you would build any other skill. But it requires feedback. And that feedback is very hard to get in an unambiguous way. Who gives you unambiguous feedback? Maybe your spouse, maybe your mother, maybe not, you know, maybe your therapist if they're good and you
pay them enough. You know, it's that it's not like, you know, building other skills where there's concrete unambiguous feedback to be had, but it is it is a skill. It's a skill like any other skill that can be built. And that's I think the thing that why people find this theory really valuable. They find it valuable because it gives them agency where they didn't think they had any and it gives them hope. >> So so how do we transition to that? So there there are if
you look at certain certain areas of cognitive behavioral therapy for example we'll talk about the relationship between thought and emotion and action. Um, and I think those are attempts to also give give agency, right? >> I I think so. Let me just say that the the treatment strategies work, but they don't work for the reasons that cognitive behavioral therapists say they work because they couldn't they couldn't possibly work for that reason. >> So, that's that's kind of where let's go there because that's kind of what
I was trying to understand is how does this explain an observation of CBT or DBT's effectiveness? Well, I think first of all, I mean, have you ever tried to not think a thought? >> Well, it's impossible. We know that. >> Yeah. It's really hard. >> Yeah. Yeah. >> However, if you change your actions or you change your circumstance, like you get up, go for a walk, or you don't even have to physically move. You can just >> virtually like move, meaning like change your attention. >>
Yeah. >> Yeah. Just shift actually move your eyes. You could you don't have to move your the rest of your body. You can just move your eyes. Shift your attention to something else. um that can that works sometimes. >> So, but let's let's pull on that thread. So, let's use the DBT example of opposite action. This is one of my favorite personal techniques, right? When I am really upset, I I 100% isolate. This is my default pattern. When I am upset, I want to pull myself
away from everything and everyone, including my family. So cognitively I know this is not a good thing to do but I do it and when I'm when I'm able to I can muster the energy to undergo opposite action. So I will say I'm really stressed out. I'm really angry. I'm really upset. I'm going to isolate myself here. No, Peter, get up off this chair and go and play baseball with your son or go and play soccer. And it guarantees you're going to feel better. But it's
like it's so hard to get up and do it. So >> for sure, >> but I do feel better. So why explain everything that's happening through the theory? >> Well, I can't explain everything. But what I can say is that every mental event begins as a motor plan. Every set of predictions begins as a motor plan. The remembered past begins as a motor plan. a plan for, you know, an alostatic plan for your brain is anticipating the needs of the body and attempting to meet those
needs before they arise. It's a plan for your heart and your lungs to coordinate and your gut to coordinate. It's a it's a metabolic plan that will support the skeletal motor plan, whatever it is you're going to do. And copies of those signals become your thoughts. They they create your thoughts. They create your feelings. they create your sensations. So the reason or a reason or maybe a part of the reason why um uh you know literally physically doing something different works to change your thoughts and
feelings is because thoughts and feelings are the result result from motor the preparation for motor movement not the other way around. And actually, if I were gonna be >> sorry, say say that again. Um, >> it's not that you see something, then you think about it, then you feel something, then you move. >> Ah, yes. Yeah. >> It's that you start with a movement. You start, right? If we were to stop time right now, what's happening is your brain has some there's some pattern of signals
of what just happened and it's making a set of motor predictions. So, it's starting it's preparing to change what's going on in your visca in your metabolism, the coordination of that. It's preparing to send signals and it's also preparing to receive signals that it expects to receive based on those motor changes. That's it's you start with a motor plan. Then you think and feel. Actually, I would say then you feel and think and then you see. Or maybe seeing happens at the same time as feeling
and thinking. But basically, it's motor first, mental second. It's not mental first, motor second. That's we experience the world that way. That is not how it works under the hood. And it can't work that way. I mean there's a beautiful book that was published um by a couple of physiologists um called the was it called the principles no what was it called? Oh it's n 2015 Peter Ster Sterling and Laughlin MIT press >> we're gonna find it we're going to put it in the show. >>
It'll come to me. >> Yep. >> This book is like this is a this is a masterpiece. This book is a masterpiece because they basically I think they go over a fly, a a mouse, and a primate maybe. No, maybe it's a fly, a worm, and I don't know. They take three organisms and they go over basically like the the um the energetic constraints on a nervous system, and they just show that if it didn't work by prediction, it couldn't possibly work. >> Do do you
think this explains, by the way, why our brains, you know, the famous statistic, right? It's 2% of our body weight and yet 20 to 25% of our metabolic demand. Um, is that in your mind a further line of of of of contribution to this? >> I don't know. Principles of neural design. It's called the principles of neural design. Um, I don't know. I don't I certainly haven't ever treated that as evidence before. Um and also I would say the estimate that your brain takes 20% of
your energy budget is based on the full um cellular respiration cycle. So it includes oxidative phosphorilation and the brain neurons don't always get ATP um by um the whole sig the whole um cycle they sometimes some neurons almost exclusively get their ATP by glycolysis with that doesn't require oxygen. So I I I think it's just more complicated. Um, but it is the case that the brain is a very it is your most expensive organ. >> So, so now let's let's think about that through the lens
of other animals, other mammals even. Um, would you agree with the statement that we as humans have the most developed brains relative to others? >> No. >> No, >> no, no. So, I mean, our brains are not more developed. Um our brains are differently developed. You know our brains are we have exactly the size um of a you know our cerebral cortex are not like larger than what you would expect for a primate of our size, right? Like we I mean our I mean primates have
a larger larger proportion of cerebral larger proportion of the brain is the cerebral cortex than for other for some other mammals. Um, but that's a that's an innovation that comes with primates. It's not like human specific. And there are other animals that have much larger cerebral cortexes than we do, like an elephant, right? Or a whale. >> So, what what is unique um neurobiologically about humans relative to others. >> Yeah. So, I think one thing that's unique is um it's probably not the only thing, but
one thing that's unique is our capacity for abstraction. And that you can see the evidence for or what the maybe not evidence but the like anatomical substrate for that you can see. So if you were to take the cerebral cortex off the rest of the brain and lay it out flat like a napkin what you would see is that there is there are there's an architectural gradient there. And the the architectural gradient is mostly in the upper layers and it it kind of looks like this
that like when you're at V1 so you you have like at the out at the outer part you have V1 A1 so primary visual cortex primary auditory cortex um primary semataensory cortex sort of and um there the neurons are very small and they have fewer connections to each each other very they're so there are um a lot of neurons but they're very tiny I'm talking about parameal neurons mostly and they're they're not very well connected so they're they're processing small details and at the other end
of the um gradient you know in the midline um in the areas that are closest to the subcortex which we would call limbic they are um there are fewer neurons but they're honking big and they have a ton of connections so the thickness of The cortex is the same but the constitution of the architecture is different. You go from a lot of little cells to fewer to fewer to fewer to fewest. Okay. There's also a temporal um gradient but we would call that dimensionality reduction or
compression. So signals are being compressed or you could say they're being decompressed. So what is unique about a human brain um or not unique but what is special about a human brain is that our ability to compress is greater. We have a lot more that gradient is sharper in a human than it is in any other animal even our closest relatives um like a chimp or or or a gorilla. Why does that matter? Because what it lets us do is it lets us see similarity in
differences. It lets us treat, you know, a cup and um uh and um >> you know, a pale and a and a and your hand as um something to drink out of. It lets us see um a kitty cat and a tiger and a cool guy all as cats. It lets us right. So what a chimp or what a what a chimp can do is a chimp can probably understand that um cl a monkey climbing a tree and a snake climbing a tree is both climbing.
They're both to climb and it might even understand like a human climbing like literally climbing up the stairs is the same as climbing whereas I think probably macak can't do that. They probably you know climbing a stair climbing the stairs and climbing a tree would be seen as different actions. But I'm guessing I don't actually know. I'm just sort of inferring. But what it definitely can't do is understand a metaphorical social ladder. Whereas a human can understand that. So a human can find functional differences in
things that look different, smell different, sound different. This is something that humans can do. Um, now on the other hand, here's something humans can't do that a chimp can do really well. If you do you ever have problems trying to find your keys? >> Sure. >> A chimp would never lose its keys. Like chimps forage for information. They forage for food. So their ability their what we would call episodic detail. Their ability to remember detail far outstrips a human's ability to remember the details. So you
know we would get lost uh in the forest whereas a chimp would not get lost. They would know exactly where they were by fine grain details that we would probably ignore which is also a consequence of our great capacity to abstract. And what comes out of this great capacity to abstract? >> Civilization comes out civilization comes out of this. I mean like basically there are many many many things that exist because humans collectively we impose functions on things that they don't have by their physical nature.
But if we impose them collectively and they dictate our actions then they do have those functions like money. I want to go back to this idea of chronic stress, right? Because at the end of the day, what I think matters is our our experience, our subjective experience. If I mean, we've already touched on the idea that I think we both believe, which is the state of one's mind can impact one's health or vice versa, right? We don't we I I shouldn't infer the causality like that,
but let's just say that there is a relationship between the state of mind and the state of health. And by the way, even if there wasn't, why would we want to suffer, right? Why would we want to suffer in our, as David Foster Wallace calls it, the tiny little skull-sized kingdoms on our shoulders, right? Like why why we we don't want to suffer? >> Well, I think there are a couple of things that that I would a couple of caveats that I would make. First of
all, there's a difference between discomfort and suffering. >> I can think of lots of reasons why I would I would deliberately engage in discomfort every goddamn morning. Exactly. Exactly. Exactly. Or, you know, I don't know, learning to play piano or learning doing a PhD or whatever. There are lots of reasons. >> Getting getting your getting your kid out of bed in the morning when they don't want to. >> For sure. Suffering though, I think is a different ma matter. suffering is is there's something additional to
suffering that over and above discomfort and in fact in chronic pain the goal is often to reduce suffering even in to dissolve it into just mere discomfort so that's the first point that I want to make the second point that I want to make is that um that you can talk about stress so what I mean you can talk about stress from a metabolic standoint point or you could talk about stress from um from a from a subjective standpoint and and I think it's important to
>> but the but the metabolic standpoint's difficult Lisa so for example like I can if I if I had 10 people lined up and I measured free cortisol levels in all of them and epinephrine and >> yeah yeah but I just want to make the point for the listener as well if I measure all the catakolamines and all the adreninocorticoids the stress hormones it tells me nothing about what they're experiencing. >> Okay, so cortisol is not a stress hormone. Cortisol, as you know, is a gluccocorticoid.
It's involved in glucose metabolism. That should be a clue. Like if you when you when you wake up in the morning and you go to drag your ass out of bed, you're you have a cortisol surge. Why? Because you're about to do something that costs >> energy. So that should to me that was a clue like when I realized you know somehow I mean I took introysiology in college I completely wasn't thinking about it until 2010 when I was at a meeting at NCI at National
Cancer Institute and they were talking about um gluccocorticoid um metabolism and I was like holy depression depression major depression has people have cortisol dysregulation It's not a stress hormone. It's just a hormone that is involved in regulating glucose metabolism. That means that stress probably has something to do with glucose metabolism. So when I thought about it more and over the years, this is what we've come to really a moment of stress metabolically speaking is just whenever your brain is predicting there's going to be a big
metabolic outlay. Allostasis. Your brain is anticipating the needs of the body and preparing to meet those needs before it arises. >> By the way, Lisa, you've used the term before and I know what it means, but I want you to contrast homeostasis and allostasis because we talk on this podcast a lot about homeostasis, which is the reaction to, but maybe you can put that in context. >> Yeah. Yeah. Sure. Sure. So, homeostasis is reactive. Allostasis is predictive. Homeostasis is about returning to a set point or
a range of set points. Alostasis is not about um returning to any set point. It's about efficiency. It's about metabolic efficiency. So I can be well someone could be in alostasis when they're running a mile and when they're asleep. I mean, I wouldn't be in alostasis when I'm running a mile, but somebody could be, but they would never, nobody would ever be able to maintain allostasis when they're running a marathon. That's just too much of a disruption. So, allostasis is predictive. Homeostasis is reactive. Alostasis is
um about maintaining balance or um efficiency metabolically speaking. It's not about trying to reduce cost by getting back to a set point. Different systems can work by homeostasis but they are all coordinated by the brain's alostatic efforts and there are you know global metabolic signals that the brain receives um GDF-15 um there's some indication cortisol actually may function this way there are a number of kind of general um signals >> I think even lactate probably classifies as well. >> Yep. Yep. It's possible. So, um, so
I'm talking about allostasis because I'm talking about prediction. So, when the brain is predicting an allostatic disruption, it's anticipating the needs of the body and preparing to meet those needs. That actually is the definition of allostasis um according to Peter Sterling um um that we use in our lab. That is stress. I mean, that's stress. That's when you see a cortisol surge. That's when you see other chemicals and so on. We don't always feel subjectively distressed when when there is a moment of stress. >> Correct.
But >> right, it could be heightened. could be pleasant as you said, >> but when there's and as long as you replenish what you spend and you're predicting well and you're not like, you know, swimming in a sea of uncertainty where you can't predict very well. Um, by definition, that's what it means to be uncertain. You can't predict. um you're you're you know what you spend metabolically speaking and now I'm kind of like veering into metaphor here what you spend you replenish what you spend and
you're okay but if you don't replenish what you spend or you mispredict or you are in a sea of uncertainty Or you have costs that are um that can't really be recovered like maybe you have um systemic inflammation for some reason like anything which is basically going to met metaphorically make withdrawals from your metabolic budget that will very slowly could accumulate over time and then you will be distressed because fundamentally mood is related to metabolic status. Okay, let's let's take another example of um a child
that is being abused. Um so there's two things that I can at least imagine that are creating distress. The first is the unpredictability that you said. I don't know which day dad's going to come home drunk and I don't know what he's going to do, when he's going to hit me or when he's going to hit mom. And then there's a second feeling that I think is related, but I'd like you to comment on, which is the helplessness of the situation, which is I don't have
agency in this situation to do anything when dad comes home drunk. So that's there's there's uncertainty and then even when something's happening, I don't have agency. And to me, I can't imagine a greater um set of circumstances to produce distress in anyone. I even though I use a very pointed example with a child. >> Yeah. I would say though that there's an additional thing there that um there's an additional so there's uncertainty but there's also and there's lack of agency but there's also um you know
we're social animals. So we don't manage our nervous systems on our own. We we are the caretakers of each other's nervous systems under best circumstances. Right? The best thing for your nervous system, Peter, is another human. And the worst thing for your nervous system is also another human. And so one of the things that we do for our children when things work well um is we we maintain allostasis for them. They can't do it. Their brains actually aren't equipped to do it. They have to kind
of learn. They don't know. I mean, a baby is born, it can't even nurse. You have to teach a baby how to nurse. baby a baby can't even control its limbs. I mean babies the the wiring is there but it's not fully developed. Partly what we do when we feed them and clothe that keep them regulate their temperature and cuddle them and do all these things is we're actually regulating their their nervous system. We're regulating their metabolism and slowly slowly slowly they learn to take this
on themselves but not completely. We regulate each other all the time. So, we can basically create for each other savings and like to use that meta body budgeting metaphor like your brain's running a budget for your body. It's not budgeting water. I mean, it's not budgeting money. It's budgeting salt and glucose and whatever. You know, this is a metaphor that I've used before. It's a metaphor. So, it's not an explanation. It's just a metaphor, but it's a good metaphor for this purpose. Um, we can create
savings in each for each other. We we regulate each other. We literally regulate each other. You know, there was um, you know, you have your hand up here like this. And so, you know, if you and I were in a coffee shop, I might do this and then, you know, you might scratch your head and then I might like push my hair back. People coordinate their actions. They coordinate their breathing. They coordinate their movements. and you coordinate your breathing, you will, you know, more likely coordinate
your heart rates. This is something people do when they trust each other, when they like each other. Um, we can also disrupt each other, right? So, one of the things that's happening for a little kid who's in an abusive situation is that um they're not getting the benefit of that kind of savings that most of us who um are in safe environments, you know, benefit from. >> What how does that translate to the to the metabolic deficit? Right. So in that situation, >> well, it doesn't
it doesn't it just means everything is costing more. Everything is costing more. Like literally, if you're with a friend and you have like a a backpack full of rocks on your back like in an experiment, you're with someone that you like and that you trust. You burn fewer calories when you shle that thing up a hill than if you're by yourself or you're with a stranger. So what this means is it just means that like things are more expensive. That means that that kid's going to
just metabolically be less resilient. Plus, if you add to that the fact that they're going to be more vigilant and there's actual harm being done to them, it's like a perfect storm of um metabolic burden that doesn't show up necessarily then. It shows up in adolescence or in adulthood. And it doesn't just show up as mood disturbance or whatever behavioral problems. It shows up as like di likely increased likelihood of a metabolic illness, diabetes, heart disease. I mean ACE, you know, ACE scores is like, you
know, like the adversity in childhood >> predicts not perfectly >> bad health outcomes. >> Bad health outcomes. So it's like a slow drip on a pipe. One drip is not going to do anything. You know, a 100 drips not going to necessarily do anything. But eventually, if it goes on for long enough, that drip will bore a hole in a steel pipe and or you know, monkeys jumping on a bed or whatever, you know, like like eventually the bed will break. Maybe not the first time,
maybe not with the 10th monkey, maybe not with the 20th monkey, but eventually it will break. And so I think that's the way to think about it, that it's a slow, slow process. So chronic stress might make you feel like but it's not actually going to harm your health immediately, but it will inevitably harm your health eventually. >> How do we turn that into a prescription? So let's let's just now talk about a situation where uncertainty and helplessness take a different form. It's no longer something
as dangerous as an abusive parent. It's uncertainty about a job. It's uncertainty about a relationship. It's uncertainty about the world we live in. AI, you you pick, you know, there we could all come up with many things to be uncertain about and to also have some degree of helplessness about that certainly would never rise to the level of trauma or traumatic insult. And yet two entirely different people can be faced with the same set of external circumstances and have a totally different manifestation of this notion
of chronic stress. So one I'd like to understand a little bit why that's the case and and perhaps more importantly what can the person who is having a more negative response to that do in order to shift the balance and take that metabolic cost off their shoulders? Well, the first thing I would say is that I don't think we have to imagine what it's like to live in uncertainty. I [clears throat] think I think um modern life is designed to bankrupt a bo a human body
budget. And we're not equally all affected by these things, but we're all affected in one way or the other. Um and even people who think that they are immune are not immune. Right? So, in the popular press, you see people talking about, you know, fight andflight circuits and how they're overworked and that's why everyone's so stressed. There are no fight orflight circuits in your brain. You don't have fight orflight circuits in your brain. Your periqueductal gray doesn't contain fight orflight circuits. It contains circuits for regulating
your autonomic nervous system. And it regulates your autonomic nervous system throughout your entire life. And of course during moments of threat they're it's very active but we have done these experiments where we have people do a a memory test where they have to remember one thing [laughter] that modulates the perryqueductal gray the so-called fight and flight circuits too. You know it's like the task is easier than remembering a phone number. there's no stress involved um in the conventional um distress sense but yet you see these
circuits being active. So so I would say there's tremendous amounts of uncertainty um combined with you know the fact that we our niche is much bigger like the things that concern us that affect our metabolism is much bigger than it used to be because of the news and because of the internet and so on and so forth. So, we've got expanded niches, like an expanded potential burden, expanded um um uncertainty. So, a lot of expanded cost um and so people are more vulnerable to um to
metabolic illnesses of all sort. And my prediction would be, you know, people I I hate making predictions, but I will predict this because I won't be alive to see the result. But I think in a hundred years, if we still exist as a species, somebody's going to look back at this time and just see it as a massive public health crisis, right, where we're basically all bankrupting our body budgets at the same time as we don't sleep enough and we eat pseudo foods, you know, and
like, you know, anything that's not on the um periphery of a supermarket, you know, and we're do doing all these things, each of which is um sometimes a nudge and sometimes a massive shove in a particular direction. So that I feel very comfortable making that prediction. your question about like well why are some people more susceptible like you know I mean yeah why do why do some adversity adversity is the potential for something negative to you know but there are people who experience adversity and who
are not stressed and they're like subjectively and there are people who experience stress they're experiencing things as adverse that like the mythical average person wouldn't All I can say there is it has something to do with prediction and memory. I think that seems pretty clear. Um and also remember that you know you're you're I don't have to tell you this. I mean you know this that while there is no muscle memory really I mean in the sense that your muscles don't have a memory your brain
has a memory. Um, you know, nervous systems do adjust themselves to the environment that that they're in. I mean, your brain wires itself to its world. That world includes the body. From your brain's perspective, your body is part of its world. That's like a really important part. It's not just what's out there. It's also anything that it doesn't have direct access to. And that a brain wires itself to the body. So if there's chronic stress or if there's adversity and the brain's regulation of the body
changes how the body is, the body then sends signals back to the brain that the brain then has to adjust to. Right? So what I'm trying to say is that you Peter have a brain that is wired to see out of your eyes that are literally in the positions they are. If we magically took your brain and stuck it in somebody else's skull, you wouldn't be able to see the right if somebody something happens to somebody's ear, right? They they can't hear for a while. They
have to like adjust. I mean, think of it like uh when you get your teeth fixed and then your tongue is like constantly probing at the, you know, it's because it's trying to forage. It's learning and it will wire itself eventually to whatever is new and different. If you have an illness in the like say you have literal tissue damage in the body and that's a met that's a metabolic burden right there just the healing right that means the brain doesn't have necessarily the resources that
it needs so that when those signals disappear the brain is still predicting that it's going to receive them because it's not updating it's not learning because learning is expensive. That's chronic pain. That's pain that is not pain is real, but the tissue damage is gone. But you don't feel pain in your body. You feel pain in your brain. You feel everything in your brain. You see in your brain, you hear in your brain. All of your sensations are in your brain. You know, you pinch yourself.
You don't feel it here. You feel it here. So I think that one thing one source of individual difference is um you have an adaptable nervous system. You have an adaptable body. You have an adaptable immune system. That means it learns and we each of us come to these external circumstances with different learning histories. That means we're wired. I mean, we are wired by our experiences. Not I mean there I mean it's not I'm not saying that we're blank slates or anything like that, but people
come to a situation prepared to experience it in a particular way. Everything you experience, everything you do is a combination of the remembered past and the sensory present. If you don't have a remembered past and your brain can't cobble together bits and pieces of the past to make a prediction, you will be experientially blind in the moment. You just won't. It's just be noise to you. >> But how do we change that? So let's everything you're saying makes sense and I if I'm in a situation
where I I am just manifesting I'm I'm just making up this example. I just I get so anxious under a certain circumstance that you wouldn't get upset about because you bring a different history, a different memory, a different pattern, a different recognition to that system. But I I I bring this very negative experience and I have good reason to because of whatever happened in the past. But now I want to change that. I don't want that to be my experience anymore. What can I do? >>
Yep. So, one one thing that you could try to do which won't be very effective is you could try to reach back into your past and change your past because who are you right now? You're a combination of your remembered past and your sensory present. So, one way to change your current experience is to change your past. Well, good luck with that. That's a little bit but so Lisa change my past or change my interpretation of my past. Those are two different things obviously. >> Well,
they actually aren't different. Exactly. When you remember your past, it is a current experience. Memory isn't different from a perception. Okay. Like when you're remembering something, it is a current experience that you are having of something that happened in the past, but it's not identical. And it's actually >> Yeah. I I I sort of maybe I didn't say this correctly. What I mean is you could change your interpretation of it, couldn't you? >> Yes, you can. But the more effective thing well anyways that's where I
was going where I was going with this was to say you could try to change the meaning of what happened to you in the past >> which some psychotherapies attempt to do and with more or less success. But one thing that you could do is you could change you could change something now and practice that change so that you could predict differently in the future. It's like driving. You invest energy in the moment to learn something new and you practice and practice and practice and then
what be starts off being deliberate ends up being automatic. Every skill works that way. Driving works that way. So, for example, when you are feeling overwhelmed and um you have to force yourself to um go be go play soccer with your kid or whatever, you could practice that skill in advance um before the heat of the moment so that in the heat of the moment, your brain is predicting differently. >> So, let me put it this way. I don't know what you did in Canada when
you lived there. One thing I used to do was the first snowfall first, it's not really snowfall. It was like the first ice that appeared, I would get into my car and I would del I'd go to a um like a a side street or or a parking lot and I'd deliberately put myself into a tail spin so that I could remember how to get out of it. I didn't want the first memory to be the time when I really needed it. I wanted to practice.
Um sometimes you can take advantage of moments where you can practice actually making meaning differently. So, um, here's an example I often use. Um, I have a colleague who studies positive emotion and like the benefits of positive emotion and I'm inherently skeptical person. So, I thought, well, I'm going to try it myself. So I thought, all right, I'm going to experience awe every day. I'm going to like give myself a little exercise. And so I started with something easy. Like I would drive along, I lived
in Boston at the time. I drive along the Pike, Massachusetts Turnpike. There was like a billboard of a cute little baby orangutang with its mother. Adorable, adorable little baby. So, I would look at the baby as we were passing by it and I would think about the baby and I would have a moment of awe, you know, and then I moved on to things like, you know, the sound of the waves, you know, beautiful flowers, the crickets at night, you know, really loud crickets at night.
And then eventually that became easy. So then I started to try like um well how about a weed or how about a weed popping out of a cracked sidewalk? A weed. A dandelion poking its ugly little head out of a cracked sidewalk can be a moment of awe if you understand it as nature's um power to be unconstrained by human attempts to control it. Now it's very easy for me to experience a moment of awe when I most need a break. Like when my nervous system
most needs a break. For example, you know, I'm about to take an important meeting. >> I'm stuck in traffic. I'm about to take an important meeting on Zoom and my computer fails. Um, you know, or um, you know, a meeting where I really can't lose my temper, you know, where I really have to be cool. um or um you know a time when you know I'm sleepd deprived and uh my you know little kid is doing something to drive me batshit crazy. Um you know these
are all moments where if I can just get a little distance for two seconds I can take a breath metaphorically speaking and then I can come back into the fray >> and and what are you doing? Are you picturing the dandelion? Are you going back to >> No, I'm finding a moment. Like for example, if our internet failed right now and um I would, as I often do, I would be like, well, it's miraculous that we can even talk to each other. I don't know where
you are, Peter. I'm imagining you're on the west coast. I'm on the east coast. That mean, no, you're not. Okay. Well, wherever you are. The point being that we could be thousands of miles away from each other and but we're talking to each other like we were in the same room. So yeah, when it fails it's frustrating, but the fact that we even get to do it is like amazing and awe inspiring. So no, I find something in the moment to be in awe about, >>
you know, >> and how long did that take? >> Well, the practice No, no, this is a very important example and I really like it. How many weeks or months did it take to rewire? >> It took a couple of months. It took a couple of months of daily practice. The thing I would say about having more control and more agency is that first of all, it doesn't look the way you expect it to look. You expect, you know, you expect to be able to just
in the moment do it. You can't. You requires practice. Like I was just telling somebody yesterday, you know, like who wants to just stand up and give a lecture? I'm like, "Yeah, I can stand up and give a lecture after I've practiced it 30 times." like it looks like I'm talking off the cuff and I am in that moment, but it's all automated. I've practiced it already a ton of times, you know? So, you can take advantage of that. The control is looks different than you
expect. It's harder to get than you expect. And not everybody has the same opportunities as everybody else. So some people have more options for controlling and for changing things. But even a monk who is being tortured can do it. That means that we all all of us even in the most dire moments we all have a little a smidge more control that over our how we experience ourselves in the world than we think we do. And in some moments that really that all all you really
need is that little smidge. You know, in some moments it's that's good enough. And in other moments when you have a little bit more energy, you know, you can practice lots of different you practice basically making meaning differently. You practice. So, for example, um I every morning I get up and I have to walk by my office to get to the gym and the temptation is so there. I'll just look for a minute. I'll just look for a minute. I know I know I won't just
look for a minute. The minute I start looking at my email, I'm going to be there for an hour, right? But every morning I am practicing not doing the thing that I feel compelled to do in order to do the thing that I know later I will be glad that I did that is you know in the in the moment it's trivial but the practice in being able to do it um >> and you think everyone is capable of that. I think everyone is capable of
that. I think everyone is capable of something like that. >> So what I'm not disagreeing with you. I'm not agreeing or disagreeing with you, but I want to know what you would say to the person who's listening to us who doesn't because you and I think are both cut from that same cloth. Lisa, um I've always chocked it up to um Well, I Yeah, let's just say this. So, someone's listening to us and they're saying, "I don't have that gene, Lisa. I I don't have the
willpower. What you're describing sounds like willpower to me." And how would you correct them in their thinking? >> Well, here's what I would say. First of all, no, willpower is this is designed to um this is just based on on empirical evidence, like how do you build a skill? It has nothing to do with willpower. Willpower is putting a bowl of potato chips in front of yourself and trying not to eat it. What I'm talking about is getting rid of the like you construct your niche
so that you don't have to deal with the bowl of potato chips. Um, you know, you you can if you want to change how you feel in the moment, you can either change your remembered past, you can change your um sensory present um or you can move in some way, which usually involves changing your sensory present. But like that's pretty much it. So you um but what I guess in the end I'm a scientist and what I would say is the following. I have emails from
like I have received more in excess of like 10,000 emails from people who have read both of my books and have found ways neither of those books gives people advice on how to use sci these scientific findings in their own life but they do. These people write to me with stories about their own lives and how they've used it and how hopeful how it gave them a sense of agency and a sense of hope. I've had people write to me from um from a battlefield like
on a battlefield like in the middle of a war. I've had people who've written to me like all walks of life, all kinds of people and I will also say that um I we used I mean how I came to any of this right how I came to any of this I mean I study emotion and I had a lot to say about emotion and the brain and so on and so forth but all of the work on metabolism and meaning making and all of that
came later. Have any of the people who have written to you um used any of what you've you've you've written about through the lens of grief? So you talked about your personal experience where you have a much better capacity to manage your emotions in distressful situations. And part of it is making sense of it, finding awe in hey like I know the internet went down. The fact that we have internet's amazing but now when someone loses their child loses their spouse I mean it's a it's
it's it's grief to the highest extent is is there tell me tell me how you you think about um impro reducing suffering in that state. >> Yeah. So I think that you know I think that when I'm talking about the brain's meaning making capacities I'm not talking about grand things like finding the meaning of life or explaining why someone dies or things like that of a you know of a of a teological nature. I'm talking about things like what does a set of sensory signals mean?
So the thing about grief, the thing about losing someone that you love is that you are losing a a source of body budgeting impact. Like we regulate each other's nervous systems. My daughter when when she was a baby was regulating my nervous system as much as I was regulating hers, maybe more. And when you lose someone you love, it can feel like you've lost a part of yourself. And you have you've lost body budgeting support, you've lost somebody that you are physically connected to in some
way. That's what love and safety is. If you feel safe with someone or you make someone feel safe, you create an environment for them to feel safe, you're physically linked to that person in a way that is not necessarily metaphorical really in that they had an impact on your metabolism and you have an impact on theirs. I have a friend, I can text her, she lives in Belgium. I can text her three little words. She doesn't have to see my face. She doesn't have to hear
my voice. What I text her will affect her heart rate. It will affect her breathing rate. It will affect her mood, right? I mean, powerful words and actions in relation to other people are powerful. So you when you lose someone, you've lost that and you feel it. Now, that doesn't explain why they died. It doesn't make it seem any more fair that they died, but it does help you understand what those sensations mean and that they eventually will um in like I would say the evidence
is something like 70% of people who don't experience complicated grief, you know, they will go away >> in in what period of time? Well, I think I'm thinking of George Banano's work now and your now your cat I mean I haven't looked at this work for a long time but I think it's like a couple of months and then complicated grief like loss of a child or something can go on for for longer but eventually within a year or two I think the evidence is that
people if it t even if it takes that long it's not that they're not sad it's that they're not suffering right so it'll still feel like occasionally uh over time, but um or maybe every day, but you won't suffer. Um it'll be discomfort. >> And are there things that an individual can do to shorten that period of time between suffering and sadness? >> Yeah. Well, the evidence suggests that you can remember times that um we're were happy. So the people who recover faster from um the
loss of a loved one um are people who um remember the positive meaningful time together. Um that's what the evidence shows. And it's interesting because if you think about there are cultural like traditions that get you to do that like sitting shiva um in a in a Jewish um when you're mourning someone Jewish or awake or you know that you're remembering things often with other people that um that bring the person make the person closer to you and bring them more alive like so here you
know so here's example this is sort of a negative example but I I think is an interesting example. So my daughter um doesn't remember herself when she was two. Of course, like who does, right? I remember her when she was two and my husband remembers her when she was two and we had we have a shared memory of her when she was two. She doesn't remember. So that means I had a relationship with someone who's not here anymore. And sometimes, Peter, I miss I mean, I
love my kid to I love her to bits, but I sometimes miss that exuberant little two-year-old. Like, I really do. >> I want to pause there for a second because this is so funny that you bring this up, Lisa. Two nights ago at dinner, we sat at the table, me and my wife, and our three kids. So, uh, 9, 12, and 18, and we all ran around the table explaining our first memory in life. So, what is the first thing each of us remembered? And it
for all of us, it took place at around three years old. Now, of course, it's much more I'm less interested in what me and my wife experienced, but I loved hearing what the kids experienced because I, to your point, like I was there. I was witnessing it. And there's amazing things they can't remember. Like the fact that my daughter doesn't remember the first two houses she lived in. No recollection at all. >> Right. Um, but what you said is so profound that I never thought about
it that way, which is, um, my kids get so annoyed with me because I can't stop telling them about how much I adore them. And I'm like, "When you were little, you were the cutest, sweetest, most edible thing on the planet." And they're like, "Dad, shut up." But what they're what I'm missing, I think, is that was they they don't remember that. They don't know what I'm talking about. >> But your wife remembers. >> Yeah. And so what I'm saying is that >> and we text
each other pictures and videos all the time. >> No, we do that too. Yeah, we do that too. And my husband was so great, you know, when when our daughter was small, he would always like write down these like perfect little gems of things that she would say like >> we have a journal each of them. We have a journal for all the funny things they said. >> Yeah. Exactly. And so, you know, we'll text each other things like just in the middle of the day.
But the point is that I can I can experience that feeling again with him, the two of us together, not with her >> with That's such an interesting observation. She's not there in that capacity, but you're still experiencing her. >> Yeah. And you know, being her mom was is a huge part of my life. And um and we're very close and I've you know we're very close and we talk to each other like text, you know, like all the time. Um and you know, we don't
live in the same city anymore, but you know, like this morning we had breakfast together on Zoom and whatever. So, we're close. And there's always been something to love at every age, like something new, something miraculous. But I do miss that. I do miss that little I do miss that little girl. I I really do. And sometimes I really I really miss that little girl. Um where you know she just wanted to be scooped up. She had her hands in my hair you know she would
you know and we'd be talking to each other like you know um sometimes I really miss that. >> So so tell me though what is that feeling is predicting what? How do we bring this full circle to the moment? Well, it's a memory. Everything is every the thing this is why thinking and feeling and perceiving are not that different from each other because they all are rooted in predictions which are memories. Predictions are memories. Predictions are your brain reinstating patterns from the past for the purposes
of the present. Like when you write when I'm reading a memoir, I'm not literally reading what's happened to someone. I'm reading how they experienced in that moment what happened to their experiences in that moment of whatever happened to them in the past. When you have a memory, you're not literally remembering what happened. You're remembering your memory is creating an experience in the moment which is informed by some set of signals from the past. And this is also I think a very powerful powerful insight that you
can use that you can use. Memories are not like files in a file drawer. They are not veritical. You are always reconstructing them. When you remember something whether you whether you have an experience of remembering or not, you are reconstructing something. You're constructing a set of signals that has bears some resemblance to the past or bits and pieces of the different pasts that your brain put together. That's how it works. And sometimes sometimes when something comes to mind like I have an intrusive thought or whatever,
I'll be like, "Nope, not that past." Like I literally will say that out loud to myself sometimes like, "Nope, not that past." >> It's interesting. you you went into psychology thinking you were going to be a clin clinical psychologist through this series of experiments you're unable to replicate. You completely find a new direction in your career. And yet your um your the work you've done probably has a potential for greater impact in the field of clinical psychology than you could have ever imagined or produced had
you become a clinical psychologist. >> I'll take that. Lisa, you've given me a lot to think about and I'm going to be I would be lying if I said I understand everything we talked about today. I think this will be one of the podcasts I have to relisten to. I don't do that often actually. Um but I I can I I think there's going to be a few bike rides I go on. Just for the folks wondering, no, I'm not going to be riding outside listening
to a podcast. I'll be [laughter] on a trainer. um where I'll be I'll be going over this again and I I'll probably come back to you with some other questions over email. So um >> that'd be great. >> I I really appreciate your your time and your your insight today as a fellow connect. >> Well, thanks for having me on your podcast. [music]