Neuroscience is the endeavor of helping understand the human brain, and, by extension, a map of the thoughts and emotions of our lives. With the advancement of new technology like fMRIs, we can now have a clearer window into the way our brains work, yet these findings can also be oversimplified and misinterpreted. So what are the opportunities — and limitations — that neuroscience gives us when we think about questions of morality, punishment, justice and empathy? How do brains and culture interact? What findings are universal and which are more particular?
In this episode of Sacred Science, we speak with Rebecca Saxe, Ph.D., who is the John W. Jarve Professor in Brain and Cognitive Sciences at MIT, and studies how we assess the mental states of other people.
Geoff Mitelman: I’m Rabbi Geoff Mitelman. I’m the Founding Director of Sinai and synapses. And sitting in the next Zoom box over with me is Professor Rebecca Saxe, who is a professor of cognitive neuroscience at MIT who focuses on questions of theory of mind and morality. And right beforehand, I was telling her – she’s a little bit of a celebrity to me because she’s given a really fascinating TED talk. She was featured in Alan Alda’s Brains on Trial, which was a terrific PBS series, a few years ago. So, Professor Saxe, Rebecca, thank you for taking some time to talk with us here today.
Rebecca Saxe: Thanks so much for having me.
Geoff Mitelman: So I want to start by actually unpacking a little bit about what cognitive neuroscience is really about – because people talk about neuroscience, and they talk about the interplay of the brain and the mind, and what we’re able to actually find out. When do we actually understand, and how do we understand, this intersection of the brain and the mind here? And how do you unpack some of these questions in your lab?
Rebecca Saxe: Sure. Well, yeah, those are big questions.
Geoff Mitelman: Big questions.
Rebecca Saxe: Yeah. I can start by saying a little bit about what the work we do is, but I agree with you – it’s also important to say what it isn’t. And some of the most interesting questions for thinking about it go way beyond what we’ve already done and maybe what we could do.
So the work that we did in the lab, the stuff that you’re talking about, from the Brains on Trial – we were interested in what happens in your brain when you’re thinking about the motives for somebody else’s actions. By analogy, I would say it’s kind of like what you would ask a jury to do, or somebody watching a mystery show on TV. You’re not, yourself, involved in any of the actions, in any role. You’re not affected by what happens. But still, it’s really important to figure out who thought what? What were their motives? What were they trying to achieve? What did they know? What did they not know?
So we were interested in how people do that, and inspired by some previous research on kids, and some debates about chimpanzees – like, could chimpanzees do this kind of thinking? So when I started this research – which is over 20 years ago now, wow! – I knew the question we were interested in is, okay, well, when human brains do this. Obviously we do it, right, when we use our brains and in the surface of our minds to figure out what somebody else is thinking, how does that happen in the brain?
And back then we could ask pretty simple starting questions. And so, the first question that we started with is: Is thinking about other minds special at all in the brain? And by “special,” what I mean is it’s an instance of any kind of abstract, logical thought. So you could think, “Well, we have lots of our cortex devoted to abstract thought, to solving complicated problems. Maybe figuring out what someone else is thinking is just another logical, complicated problem.” Right. Hercule Poirot model. Right? Or Sherlock Holmes. If you’re smart and can figure out complicated problems, then you can figure out who had the motive to do it.
Okay, so on that model, thinking about other minds is just a hard, abstract problem. But there were already hints when we started that suggested that might not be how our brains do it – that it might be that in our brains, special, spatially organized resources are devoted to figuring out other people’s minds, and then we would be able to see them using fMRI. So, back in the day, more than 20 years ago, when we started doing this research, we’d have adults in an MRI machine. They’d lay on their back in a tube, reading little short stories, about a paragraph long, about a character. And we could describe whatever we wanted about the character. We could describe their physical environment. We could describe what they looked like from the outside, what you’d see if you encountered them. We could describe even their visceral sensations, like feeling exhausted or feeling nauseated. Or we could describe their thoughts – what they wanted, what they were disappointed by, what they hoped for.
In those early experiments, what we found is a group of brain regions – of which I usually focus on one, the right temporoparietal junction – where metabolism, and so blood oxygen, go way up just when you start thinking, or start reading, about somebody else’s thoughts.
And then the first decade of research was like testing alternative hypotheses – does it have to be thoughts? Does it matter that they were stories? So we used movies, we used commercial movies. We’ve done this now in many languages, my favorite of which was sign language, and across now, many, many experiments, that basic thing is true – that this group of brain regions, including the right TPJ, show a lot of activity whenever you have to think about somebody else’s thoughts.
And then, to connect it to morality, when I first started my own lab, I was incredibly lucky to have brilliant, highly motivated students join right away. And one of them was Liane Young, and she wanted to study how people think about thoughts in the service of judging moral blameworthiness. So, under certain circumstances, having a false belief can be very exonerating. That is, you’d say, if you hurt somebody and you really had no way of knowing that what you did would hurt them, then we see from outside that you’re not to blame for that action.
So, the story that I often start with in these examples is like, if you were making coffee, and next to the coffee machine is a jar of white powder, and it’s labeled “sugar,” and so you put some in another person’s coffee – but it turned out that that sugar had been contaminated by a poison. And you had no way of knowing that it was, and so you harmed them, maybe even killed them. We would still say that’s way less blame-worthy.
So she wanted to know: Do you use your right TPJ to figure out what somebody was thinking specifically in this kind of situation, where you need that information to forgive them for harm they caused, and, it’s reciprocal – when you thought you were going to cause harm and did it anyway, but by luck, didn’t cause harm?
So those were the experiments we did. And she did a whole bunch of super cool experiments showing that the right TPJ is active when you make those judgments – that the pattern of activity inside the right TPJ predicts which judgment you’re going to make. And possibly the most impressive piece of this was if you interfere with function in the right TPJ very briefly, by scrambling the signals with transcranial magnetic stimulation, then we can bias moral judgments, and so make you say accidents are more blameworthy and attempts are less blameworthy.
Okay, so that is the research that we did. And it’s all looking at how do you figure out who’s to blame, or how much somebody is to blame, for an action, from a third-party, unaffected perspective, taking into consideration what they didn’t know – their false beliefs of fact.
And one of the things that I think is most important to say that we didn’t do is, in that research, we never studied making your own moral choices. So the perspective you have of a third party, which says, “Given what they did and what they know, how much do I blame them?” is very different from the perspective you would have, in the first person, of deciding. And this is an observation that I think Philippa Foot and Judy Thompson, who are philosophers, had made 30 or 40 years ago, that a false belief is – you can never consider your own ignorance as morally justifying in advance. Right. You can’t be like, “Well, it’s okay for me to put this sugar in the coffee, because I don’t know it’s poisonous.” That thought makes no sense, right?
So the kinds of processes that we could use to decide what to do have to, in some ways, be fundamentally different from these processes that I’m talking about, which are explicatory or explanatory after the fact. “I see what you did, and I explained why.” I think that’s important to start out with, because for many moral philosophers, religious and otherwise, they dispatch with third-party blame pretty quickly. They usually advise us to be as generous and forgiving as possible. You know, Kant says, “Don’t think of the evildoer as entirely defined by the evil that they did” – and then move on. Like, that’s not the hard problem. The hard problem is not blaming other people. The hard problem is figuring out what to do, and who to be, and how to live.
And I’m saying that because the glib read of the research that we did that we sometimes get is that we used transcranial magnetic stimulation to make people less moral – we could “turn off morality.” Actually, I was just reminded, just the other day, that a sci-fi TV show called The Expanse put our research into one of their episodes. They have somebody who’s had transcranial magnetic stimulation done to their right TPJ.
Geoff Mitelman: Wow.
Rebecca Saxe: But they’re using this to explain how this character has become immoral. But they mean she makes immoral choices. And I just want to be clear – we never tested anything like that. We never asked our participants to make a moral choice or to be a moral person. We just asked them to do this kind of social-science explanatory job of saying, from the third party, from a distance, “How do you think about what this other person did?”
Geoff Mitelman: There are so many different things that are going in my head now as you’re talking, because it’s fascinating – and so many questions that it raises. One of them is the relationship between the cognitive ability and emotional ability, and how they feed into each other. And I’m thinking also of some of Jonathan Haidt’s work on morality – that a lot of moral judgments are sort of instantaneous and “snap.” And I remember seeing from one of your talks, also, that transcranial magnetic stimulation – that’s not something that you can just sort of do surreptitiously. And you’re not going to change someone’s mind or their emotions or their morality. It sounds like it’s – think more of, “How much do they judge the intention?” But morality is – there’s an interplay of “what’s the intention” and “what’s the impact.” And also, particularly in Judaism and in the legal system, of how much of it is “what I meant to do” versus “what was the impact that it had on somebody else.”
Rebecca Saxe: Yes. Again, the study that we did showed that in very simple, carefully designed problems, where there are exactly those two factors – “what did you intend” and “what was the impact” – that for the stories we wrote – and we wrote them on purpose, to be like this, our participants – so, Western, well-educated, wealthy, right? Our participants initially make their moral judgments much more based on the intention than based on the impact – again, in stories that we wrote so that the intentions were very well justified. These are not negligent, or – there are many ways you can make an intention cavalier or frivolous, and we didn’t do that. These were careful, justified false beliefs. In those cases, the participants that we studied go with the intention in the balance between intention and impact until we put TMS on their right TBJ, and then we shift them more to impact.
Now, it’s true that you saw, we promised it – because people reacted very strongly to that study sometimes, and I do want to say all of the things you said. It is not surreptitious. You can’t do it from far away. Also, the effects are tiny, and they last very little time. We could make a difference of, like, half a point on a scale of 1-7. And it lasts between five seconds and 20 minutes, depending on how you do the TMS, and the one that lasts 20 minutes takes 40 minutes to set up. These are very small effects that you cannot induce without somebody knowing it.
And the other thing I was going to say is: It’s weird to have a panic about our ability to shift moral judgments with TMS, because we know you can shift moral judgments with talking. (Laughs) That’s how we shift moral judgment. We persuade one another. Of course, moral judgments can be changed – and so the tiny effects that we can have with TMS are scientifically very exciting, but they’re not practically very threatening. Okay, so that’s the caveat. You don’t have to be worried that somebody’s TMSing your right TBJ without knowing it.
Okay, so then you also asked about cognition, about hate. And was there something else in that question?
Geoff Mitelman: Let’s talk about this a little bit, about the cognition and the emotion, because I think we tend to separate out the head and the heart. But emotion is – that’s also a cognitive element, right? There is a way in which we need the emotion to be able to make judgments, to make any kind of decision here. And so, I’m curious as to how you’ve seen this intersection of “what’s the immediate emotional reaction,” which isn’t always a bad thing. It’s not always necessarily a good thing, that emotion just sort of – it is.
But one of the hard things, I would think, for some of your work is, how do you measure some of these questions – of “how much do you think someone intended to do this”? Because there’s a self-report, which isn’t always going to be accurate in responding when people talk about it afterwards, what’s happening in the moment, versus reflecting on it after thinking about this.
So, how do you gauge, how do you measure these questions of how much did somebody intend to do this? How much is it of something blameworthy or not blameworthy? I mean, I’m assuming there are different numerical scores that people will give, but you can’t just cut open someone’s brain and say, “well, that’s why they made that decision.” That doesn’t work that way.
Rebecca Saxe: I mean, it is kind of amazing about fMRI that it sort of does let us look in their brain and say, “this is the source of that decision.” I mean, that’s what a causal experiment is about, is showing, “I think this brain region caused that decision, because if I change this brain region, I changed that decision.”
Again, there are so many things going on in your question that are worth separating out. One piece is – again, we never studied first-person decision-making in our research. When that character put the powder in the coffee, what was really going on in their head? Did they really not intend the harm? That question is out of scope for our research. These are hypothetical narratives that we use – or even when they’re true stories, we’re trying to get at the observer’s perspective. So we can say the observer really thinks that they knew what they were doing, but there’s no real way of saying, “And was the observer right?” Like, what would that mean? So, yeah, there’s no fact of the matter. It’s mostly about the consensus of the observers.
And then the next question that you ask is: Okay, well, that thought that the observer has, which is like, “that person should be blamed, and I would like to see them punished. They put poison in her coffee, and maybe they didn’t know that that sugar was toxic. They put in the coffee, and look at the impact, and I’m just having this really strong reaction that something should happen to punish them.” Right? And there’s a question there of, like, is there a balance of emotion and cognition? So, I don’t normally think about this as “emotion versus cognition.” I don’t find that distinction very useful. But I think part of what it’s mapping onto is something more like “intuitive versus reflective.” Like, we have a first instinct which tells us, “I really want to see someone punished here,” and then maybe a second instinct, which we develop in conversation with one another explicitly, in which I might try to tell you why I want someone to be punished, and you might try to tell me why you don’t think it’s appropriate for anyone to be punished. And then we might come to a reflective and balanced decision that I could even articulate. I could say “the reasons why we institutionally chose not to punish anybody are like this,” while still feeling like – I know there’s still this gut instinct, but that just feels wrong.
Okay, so I do think that distinction between a kind of intuitive judgment and a reflective, articulate, introspective judgment – that feels very real to me. But those intuitive judgments, I don’t think of as non-cognitive, because they involve really abstract calculations. And you can show, using hypothetical scenarios, that the initial intuition depends on all kinds of things. It depends on the exact causal path – was it an omission or an action? – the exact degree of negligence involved, how justified the beliefs were. For our intuitions start out very cognitive, and actually, some of the work of talking to one another (this is obvious in religious discourse) is just trying to articulate our own intuition, right? Like, the reason it’s so hard to talk about morality is often that I have an intuition, and I just am trying really hard to explain it to you.
There’s a big gap between my intuition and my ability to articulate or introspect. But it’s not because the intuition is less cognitive. Once you spell it out, you realize it was incredibly cognitive. It was incredibly rich and structured. That’s the reason I’ve resisted calling those judgments “emotional” or “non-cognitive,” if “emotion” is the alternative to “cognitive,” though, obviously they carry a lot of weight in terms of motivation or what feels important, what feels wrong. They carry a lot of that evaluative weight. So that might be what you mean by emotion.
But to me, it’s just very weird to call them non-cognitive, because they’re so cognitive? And for those, the way we study those, the way we measure them, is we try to ask people questions that seem easy and intuitive to answer. And so that’s what I like about asking about “how wrong was this,” or “how much blame did it deserve,” or “how much did the person need to be punished?” – is that for the most part, we feel like we do kind of have easy, intuitive answers. We also know that upon reflection, we could deviate from or change those answers.
But it’s pretty easy for people to just keep giving moral judgment. Like, I give you a scenario, you give me your instinct, I give you another one. And that’s very useful scientifically, because in order to do these experiments, I need tons of repetitions, right? I didn’t emphasize this, but those fMRI and TMS experiments involve many dozens of moral judgments in half an hour. And so it’s not your careful, Talmudic consideration of each case – it’s your quick, like, “I need another judgment every 12 seconds,” because that way I can get many, many repetitions, and I can study the average activity in your brain across all of these cases.
So, yeah. It is very worth thinking about. What we’re measuring is cognitive, but fast, intuitive. And it’s also only the parts of your moral judgment that are very repeatable. Anything that is unique to a specific situation, I can’t measure, at least not with the ways that we did it before. I guess these days, if I really wanted to, I would know how. But for the most part, the cognitive neuroscience strategies that we’re looking for are repeatable cognitive processes. And therefore, it’s the simplest parts, right? Like, again, you know from real life, and from religious discourse, that the repeatable parts aren’t the hard parts. The really hard parts are the exceptions and the details that arise in a specific case. And that’s usually not what we’re studying.
Geoff Mitelman: Well, what’s interesting is that you brought up a really important point, which is that we tend to separate them out when, in fact, they’re integrated, right? So kind of an emotional – or, like, they’re together. Right. They’re not separated out, they’re thinking.
And another thing that’s often presented as a binary, when, in fact, they interact in a lot of ways, is the “nature versus nurture,” and what’s universal and what’s particular and unique. And you had mentioned that a lot of the people that you talk to, understandably, in Boston, are sort of the Western, educated, industrialized, rich, democratic people. And I’m wondering, how much do different cultures or people from different cultures respond to these questions in different kinds of ways?
Rebecca Saxe: Yeah. So Clark Barrett, who is an anthropologist that I admire tremendously, asked exactly that question. So he took – I think, actually, this was partly because he saw a talk Leanne and I gave in 2005. Or he just thought of it himself, I’m not sure. But anyway, so Clark Barrett and his collaborators, who are anthropologists, wrote versions of these same stories that would be comprehensible anywhere in the world. So instead of poison and sugar and a cup of coffee, it was like, accidentally spilling fertilizer into a well. Or another one of his examples was picking up a bag at the market that isn’t yours. Or another example was hitting someone in a market, which you might do because of misapprehension of the situation, like just being wrong about what was going on, or because it seems necessary to an end, or in self-defense.
So they wrote this battery, again, of little hypothetical stories where somebody does something that causes harm, where you could vary what they knew. And then, Clark and his colleagues took these questions to nine different societies around the world and asked adults in each of those societies these basic questions, like, “How bad would this be for their reputation?” “Should this person be punished?” And found that specifically, accidents based on false beliefs of fact – So, “yes, I poured that into the well, but I didn’t know that it was toxic,” or, “yes, I took that bag home from the market, but I thought it was mine,” or, “yes, I punched that guy, but I thought he was threatening my son,” – so, these kind of false beliefs of fact – the degree to which those are treated as exonerating – is hugely culturally variable.
So, in some cultures, like ours – the example he used for that was in Los Angeles. Adults measured in Los Angeles say, “those false beliefs are fully exonerated. You are not blameworthy. You shouldn’t be punished,” whereas in some of the other cultures that they measured, those kinds of beliefs were either less exonerating, or in some cases, not exonerating at all. So people said that was exactly as bad for your reputation, exactly as deserving of punishment, as if you had known what you were doing.
And so Clark and I just wrote a paper together this year, after I’ve admired him for almost 20 years, trying to say, “Okay, so what does that actually tell us about diversity of morality?” And should we think of what level of diversity is that? So the most kind of lumpy, big-picture, headline version of that would be, “some cultures don’t think it matters what you intended. They don’t think it matters what’s in your mind.” And that claim is actually defeated in Clark’s own data, because he also asked about other motives and other exonerating factors – so, for example, he also asked about self-defense, like, “what if you did it because you thought you were in danger?” And he found that in all cultures, self-defense would be this exonerating.
And the other case he asked about was, “if you think it’s a necessary means to a greater good” – so, the case where you assaulted somebody, but only because –you harmed them, you pushed them out of the way, and they broke their nose, but only because you were running to put out a fire that was going to take over the whole market. So in that case, where you think that this harm you cause is necessary for you to put out the market fire, even though it turned out not to be necessary. So in that vignette, in that story they used, actually, “by the time you get there, someone else has put out the fire.” So you thought it was necessary, but it turned out it wasn’t necessary. Even in that case, people all over the world say that’s exonerating, that the harm you caused by pushing somebody out of the way to get to a fire to put it out – that that is exonerating.
And so we can’t conclude that – at least there’s no evidence yet, no one has measured a culture yet where it doesn’t matter at all what’s in your mind. But there do seem to be differences, nevertheless, in which the kinds of things that might be in your mind, we think are more or less exonerating.
So yeah, in that paper, Clark and I kind of tried to lay out what this might mean about cultural variability. We wrote that paper this year. I think the answer is, “We don’t know exactly what is and is not variable,” but the position, how I read the literature right now, is: Thinking about minds and relationships as relevant to morality in some way probably is universal, but in what way probably isn’t?
Geoff Mitelman: And what’s fascinating – there’s so many different questions to come up of this interplay of intentionality and impact. And all the examples that you’re sharing feel like they could come right out of the Babylonian Talmud, of “I’m running to save [a life], but this person was mistaken.” And, well, what would happen if – and there’s a reason, we sort of joke, that so many Jews are scientists – because it’s the same kind of thinking, of, “Wait, what if we change this.” It’s interesting to think of all the differences here. But it also plays itself out in our society right now, of, “Did someone mean to do this? Did they not mean to do this? Who’s the person who’s making that decision? Who is it? What’s the appropriate punishment? Should there be a punishment? Are there ways to be able to talk about restorative justice rather than retributive justice?”
And because a lot of the conversations are very filtered by, at least in America, we say it often on the show – generally, basically, white, Protestant Christian America, that when there are different ways of thinking about these questions from a Jewish perspective, from a minority perspective, there are different ways of saying, “Wait a second, I’m going to approach these questions of morality in a different kind of way.” And that can make people very uncomfortable, I would think, to say, “Wait a second, my morality may be different than your morality here.”
Rebecca Saxe: Yeah, I think that is very interesting. And yeah, I want to leave a lot of room for that. I think it’s really important that whatever we say scientifically about how the mind works, first of all, doesn’t prescribe one way of solving these problems as better than another. I think it hugely depends on what problem you’re trying to solve, starting from what I said right at the very beginning – which is, the first question is: Are you trying to decide what to do and how to be? Because in that case, all of these questions are irrelevant. You can’t consider exoneration prospectively. There’s that. But then, even if we think about it retrospectively or reactively, what do we want to have happen?
I think the questions of blameworthiness are deceptively simple, or at least they’re much simpler than problems about punishment – because, at least what I want to say is, I think the question of, “Was that bad?” is a much simpler question than the question of, “How should we respond to it?” Because, in responding, a lot more is at stake than just whether that action was bad. We can say that action was bad – the impact was bad, the intent was bad. Saying that, yeah, I think that’s much easier. But then saying, “And we are going to be in the business of causing harm to remediate it?”, or of judging how bad it is relative to other forms of badness, or of judging how much it’s exonerated by the bad things that were done to you, or where we are in a cycle of vengeance – is this the instigating act, or the responding act?
But to decide all of those things typically requires a lot more social commitments that are much more subject to bias. And then also, we communicate other things when we make our own action choices than just judging somebody else, right? So again, now we’re starting to say: If I choose to punish, what kind of person do I want to be? What kind of example am I setting pedagogically? What kind of example does this set for other people when they think they’ve been wronged? What is the balance of deterrence versus being the kind of person or in the kind of relationship that I seek to be in? What will it imply about the role or relationship I think I’m in if I place myself above you in order to punish you? What does that do to how our relationship goes? What does it do to how other people see me? I just think there’s so much more at stake in choices about punishment. And I think choices about punishment are typically made very badly, and with very little introspective awareness. It’s so easy to say “that’s bad and I’m angry and I want to hurt you,” and not be introspectively aware of all the other implications of that choice.
Geoff Mitelman: And it’s almost why culture exists, why laws and religion exist as a whole. Because our natural inclination, I would think – I am somebody who thinks that our brains evolved on the African savannah. And one of the reasons our brains are so big is because we’ve got to think about all these theories of mind of different people – that our natural inclination, “if somebody hurts me, I want to hurt them back.” That’s the natural human inclination. And say, “well, wait a second, that may be my immediate emotional, intuitive reaction, but let me be reflective and think about this. I don’t have to act on this immediate element” – of, “okay, wait a second. What are the other factors that are coming in?”
And what’s so hard is – I quote all the time an op-ed from the Times, now almost 15 years ago, from Dan Gilbert, of “whoever cast the first stone probably didn’t” – that we feel the pain that somebody does towards us, but what we did, “I didn’t mean to, I didn’t mean to, I didn’t mean to.” But if somebody hurts us, we don’t care that they didn’t mean to. What we care about is that they hurt us. And so it ends up even just trying to think of a tit-for-tat, “I’m going to get you only as hard as you hurt me” – we feel much more pain than we actually inflict on others.
Rebecca Saxe: Yeah, totally. And you get accelerating cycles, ratcheting up. Totally.
Geoff Mitelman: What I’m curious about – I think one of the things that’s both, I think, so important and also challenging as somebody, because I know that you’re not just a researcher, but you’re also a science communicator. You bring your research out into the wider world. The part of the challenge is communicating that to the public, because it can easily get misconstrued. There may be things where you could say in your paper, like, “this is a very narrow definition, and it’s only in this place,” and the story becomes “your brain tells you that you don’t have to do it,” or whatever the story is. That’s what people will remember. So how do you ensure that the findings that you talk about – because when we’re talking about neuroscience, that can be something where it can really impact juries and decision-making process of like, “well, the brain says that this is what we’re going to do.” How do you think about and communicate these findings in a way that are going to reach the people that you want to be able to reach and not be misinterpreted and misconstrued?
Rebecca Saxe: Yeah. I mean, I think… I think I don’t solve that problem. I don’t know if anyone could. I try my best to be specific. I think, to some extent, the most important balance in science is to combine wonder and humility, and you need the wonder to keep going. But I think the humility is incredibly important, too, to not over-state or over-claim. The thing that I would most want to communicate in public communication with science is those same ideas – to be, on the one hand, amazed at what modern science lets us see about ourselves, that this is a tool for reflecting on what it’s like to be human in a different way, by peering inside our own heads while we’re doing some of the most human things we ever do. I hope that inspires wonder. But also to avoid hubris about it, right? It’s incredibly limited tool, a very narrow window. Some of the limitations are already obvious to us as we do it, and others will become more obvious as we go along.
So, yeah, I hope it’s possible to communicate science that way. And I’m just remembering – you mentioned I gave a TED talk. I have this sort of slight trauma memory about that TED talk because it was the first time I talked to a big public audience about this neuroscience-of-morality work, and it’s been seen many, many times. And the speaker immediately after me was also talking about neuroscience, and made much more aggressively ambitious claims – said we were within 10 years of simulating a whole human brain, that we knew everything we needed to know about how the mind works. And of course, that was incredibly popular at the time, and he got a billion euros for his research out of that idea. And I was kind of devastated by that. I think it’s incredibly sad to oversell science to the public that way, and to take public resources for science on that kind of false promise, which is proven, in retrospect – that’s more than 10 years ago, and we are nowhere near knowing how the human brain works. We could tell in 2009. And indeed, that project totally failed. And that feels terrible to me, the idea that you would sell so much more about science than you could give.
But I’ll tell you a different story that I feel more complicated about. So a project I did, now seven years ago, was, when my children were born, when my sons were born, I was scanning their brains because, of course, who wouldn’t want to scan their baby’s brains? It was amazing. I wanted to see how their brains worked. So we were scanning my babies’ brains the whole time that they were babies. And while we were scanning them, I would go in the scanner with them, so they’d be with their head in the coil, having their brain scanned, and I’d like inside, lying on top of them, making sure they were fine and trying to help them hold still and stuff. I spent a lot of time, when I was a new mother inside an MRI machine with my babies. And I started thinking about what that would look like if you could make a picture of it. And so, with help from our MRI Physicist, Atsushi Takahashi, we figured out how to take a simultaneous photograph of the two of our brains together. And we called this “the mother and child.” It looks like a Madonna image, right? A mother and baby together. So we have this picture of me, my brain, and sort of see my body and my baby’s body together. And then both of us, of course, have been scanned many times looking at valuable faces. So we then when they made an image where we kind of pasted over our anatomy, the picture of activity in our brain, when we saw a face that we liked or cared about. And we looked at that, and we’re like, yeah, that’s kind of fun. But we didn’t totally know what to do with it. I tried to get, like, a museum or a magazine to take it. And nobody got that interested in it.
And I don’t know how, but it ended up on, I think, Instagram. Somebody else posted it on Instagram saying that the activity you could see in the image was oxytocin, released by the bond between the mother and the child, which is totally false. It is not oxytocin. And the image went viral. So that image has been seen many millions of times, much more than anything I have ever produced as a scientist. It still circulates. I see it on Twitter in a new language every month or so. So it’s still circulating. It’s been the Valentine’s Day image from certain magazines, it’s been the image of the year in some European country. I mean, it’s an incredibly popular image, but it circulates with false information about what it depicts, saying that it depicts the oxytocin burst due to the mother-child bond.
So, early on in that process, I tried to correct that narrative. I wrote a thread that says, “here’s nothing to do with oxytocin in this image. It really is a mother and baby, and it really is activity in our brains, but it’s really not oxytocin.” And the dialogue that goes on on Twitter that I see sometimes has three stories. The very minority story is, “the person who made this image says what it is. Look here.” But that’s the tiny little thread. But the big story is, “on the one hand, this is a picture of oxytocin in the brain,” and on the other hand, “this is a fake.” This image is a fake, and it’s not a fake, and it’s not oxytocin.
But on the other hand, it is art. It wasn’t ever science. Right? We never made it to test a hypothesis. We made it because we thought it would look interesting. We thought it would evoke interesting thoughts about mother-child relationships and biology and the mind. And I think when you make art, you don’t get to decide what other people say about it, right? So some part of me decided – I still find it a little bit irritating that half the people think it’s oxytocin and the other half think it’s fake. But I tried to decide that when you make art, you just make art and let it go. And so the most well-known thing I’ve ever done will be misrepresented to almost everyone who ever sees it.
Geoff Mitelman: That’s fascinating. I love what you’re saying, because as you’re talking of this interplay of art and science, my mind’s going to the questions of religion and science, which is a question that I get all the time, which is, “How do you read the origin of the universe, Genesis 1, when we understand Big Bang cosmology?” And there are a lot of people who go, “Wait a second. But the text says one thing, but all the astrophysics says something else.” And the way that tend to [think of] science is – the Torah is not meant to be a science textbook. It’s meant to be read as poetry. Actually, it’s pretty clearly designed to be poetry. And so you engage in it in a different kind of way. And so trying to say, “How can I prove Genesis is” is like, “How can I prove what this picture is?” They’re approaching it in different kinds of ways.
Rebecca Saxe: Interesting. That’s a really interesting analogy. I like it.
Geoff Mitelman: Yeah.
Rebecca Saxe: Cool.
Geoff Mitelman: Well, the last thing I just want to talk about, we have just a couple of minutes, because I know you did a study about isolation, which I originally said was going to be, “Why would this possibly be something useful?”And then, lo and behold, a couple of years ago, it was quite crucial. Is that something you can talk about in just a couple of minutes?
Rebecca Saxe: Yeah, absolutely. Well, you know, you told basically the story, that five years ago, or six years ago now, Livia Tomova, a very talented scientist, joined my lab, and we were trying to figure out what would be a cool project for her to do. And she was inspired by research that a colleague of ours, Kit Ai, had done with mice on the effects of social isolation. And so we designed this experiment to ask whether people who normally have lots of social connection in their lives, but are suddenly forced to go without any social connection, crave social connection the way somebody who normally has enough food, but is suddenly forced to fast craves food. So we wanted to study: Is there a common neural mechanism of craving when you’re suddenly deprived of a need that is normally met?
And so it was an incredibly elaborate study. There were 40 people in the study who were each scanned three times. They had to be monitored for a whole day beforehand, whether they were totally in isolation in a room we designed for them, or were fasting the whole day. So it took well over a year to collect all of these data and analyze them. And in January 2020, we sat down to write this paper. We had found that being forced into isolation makes you crave social contact in one particular brain region, in the substantia nigra – so, a particular brain region where we had hypothesized we might find this event. Actually, the rest of the brain treats food craving and social craving quite differently. But this one key, core, deep-brain region treats them similarly as a craving. When I went to write this paper, I was like, the only problem with this paper is, “What is this paper really about? How are we going to tell people why this should matter to them?” This is not a disease. And although there is a lot of isolation and loneliness in the world. Most of that is chronic. It doesn’t suddenly happen one day that you lose all kinds of social connection. And when it does, it’s usually voluntary, right? Like you’ve decided to go on a retreat, or you’ve decided to sail across the Atlantic. Or it’s punitive, right? Like solitary confinement. And what we were doing was not punitive, but it wasn’t voluntary and it wasn’t chronic. And I was like, “This thing where suddenly you have to give up all your social connections… why would anybody be interested in that?”And then in March 2020, I was like, “Oh, I don’t think we have that problem anymore.” (Laughs) I think people are going to get why this is interesting.
Geoff Mitelman: It’s so fascinating, because when we think about – I don’t know if this is accurate or not, so I hope this is. But when I worked at a camp for a summer, one thing that the camp director said was that if a kid’s able to find one friend – like, one close friend – that’s going to make all the difference. They don’t have to necessarily be friends with ten kids. But if they don’t have any friends, that’s going to be the big problem. But if they’ve got one good friend, that’s going to help them get through.
And I’m wondering, of isolation – when people are isolated totally by themselves or isolated with family members, with a spouse, or isolated with people – Zoom. their stock shot way out there. We were able to have some sort of…
Rebecca Saxe: Exactly. By that point, by March 2020, suddenly everything about isolation was interesting. All the questions you’re saying – and the one that we had anticipated as interesting was specifically about “What kinds of virtual interactions are enough,” right? Is email enough? Is Facebook enough? We hadn’t really thought about Zoom, because I don’t know if you remember this, but pre-March 2020, we didn’t all do Zoom. So we were interested in what kinds of virtual interactions would make up – would eliminate that craving.
But here’s the sad story. We wrote a whole bunch of grants asking to continue that work, all pre-March 2020, and they all got turned down. And so Livia left, which is Cambridge University’s gain – she’s doing great work there. But we stopped that research program in my lab. And her questions there are also really interesting, and connect to what you just said, which is, does it matter how old you are when you’re isolated? Is it different if you’re a teenager? Is it different if you’re a child? Some of the work in animal models suggest that isolation during critical developmental periods has more chronic and lasting effects than isolation in adulthood.
So, anyway, Livia has gone on to ask many fascinating research questions, and my lab has gone on to do many cool other things as well. But we didn’t do anything else about social isolation after that.
Geoff Mitelman: Well, I mean, the questions that you’re exploring, there are so many different ways to be able to look at these questions. And, Rebecca, I can’t thank you enough for not just your time, but all the research and the work and the communication that you’ve done to be able to get all of these fascinating studies in a very nuanced and clear and compelling way. So thank you so much for your time and your wisdom here for today.
Rebecca Saxe: Super kind of you to say. It’s my pleasure.
Geoff Mitelman: And if people want to follow you on Twitter, it’s @Rebecca_saxe, correct?
Rebecca Saxe: That’s correct. Terrific.
Geoff Mitelman: So you can follow her and find her on her website. You can find us. I’m at @RabbiMitelman or @Sinaisynapses, and we have these conversations at jewishlive.org/SacredScience. We’re going to be taking a couple of weeks off for spring break and Passover, but we’ll be back at the end of April with a whole new slate of guests. So thank you all for joining us. We hope you’ll join us again. And Rebecca, thank you again for your time today.
Rebecca Saxe: My pleasure.
0 Comments