WEBVTT - Ep 53 " Can societies fight better? " 0:00:04.600 --> 0:00:09.040 Humans disagree on many political fronts, but we're finding ourselves 0:00:09.080 --> 0:00:13.760 globally in a period of higher polarization, and the question 0:00:13.920 --> 0:00:17.760 is is there anything to be done about that. Presumably 0:00:17.760 --> 0:00:20.759 we're not going to get humans to not have conflict, 0:00:21.160 --> 0:00:25.279 but is there any such thing as better conflict? What 0:00:25.320 --> 0:00:27.400 would that mean? And what does any of that have 0:00:27.480 --> 0:00:30.360 to do with brains or with the fact that modern 0:00:30.440 --> 0:00:34.040 humans only spread out across the globe sixty thousand years ago, 0:00:34.560 --> 0:00:40.640 or with social media recommender algorithms, or the Iroquois Native Americans. 0:00:43.200 --> 0:00:46.680 Welcome to Inner Cosmos with me David Eagleman. I'm a 0:00:46.720 --> 0:00:50.400 neuroscientist and an author at Stanford and in these episodes, 0:00:50.520 --> 0:00:54.000 I examine the intersection of our brains and our lives. 0:00:54.400 --> 0:00:57.440 And today's episode is about something that you know, I'm 0:00:57.480 --> 0:01:00.560 obsessed with from the point of view of brainience, which 0:01:00.600 --> 0:01:05.959 is issues of conflict and empathy and in groups and outgroups. 0:01:06.840 --> 0:01:10.760 We find ourselves in a time of conflict. Now. The 0:01:10.840 --> 0:01:14.640 things on everyone's mind are the war in Ukraine and 0:01:14.720 --> 0:01:18.440 the war between Israel and Hamas. And even though those 0:01:18.520 --> 0:01:21.679 have sucked up most of the media's attention, it's critical 0:01:21.680 --> 0:01:25.319 to note that we have lots of other conflicts also 0:01:25.400 --> 0:01:30.080 going on around the globe right now. There's ongoing internal 0:01:30.120 --> 0:01:34.600 conflict in Myanmar. There's an Islamist insurgency in the Maghreb 0:01:34.680 --> 0:01:39.280 region of Africa. There's the Boko Haram insurgency in Nigeria. 0:01:39.319 --> 0:01:44.080 There's a civil war between rival factions in Sudan. There's 0:01:44.080 --> 0:01:48.080 a multi sided conflict of the Syrian Civil War. There's 0:01:48.120 --> 0:01:53.120 the ongoing civil wars in Somalia and Ethiopia, and Afghanistan 0:01:53.200 --> 0:01:57.400 has been in near continuous armed conflict for essentially as 0:01:57.400 --> 0:01:59.920 long as I've been alive. Now, if you've been listening 0:01:59.960 --> 0:02:02.320 to this podcast for a while, you know that I'm 0:02:02.680 --> 0:02:07.800 very interested in why conflict happens between humans and why 0:02:07.880 --> 0:02:12.920 it happens so readily and consistently. So in this light, 0:02:13.520 --> 0:02:17.280 there are three main things to note from the neuroscience 0:02:17.320 --> 0:02:20.520 point of view. The first thing is that we all 0:02:21.080 --> 0:02:25.799 have the same brains on the inside. So despite cultural 0:02:25.800 --> 0:02:30.720 differences and geographic differences, your pink brain looks exactly the 0:02:30.760 --> 0:02:35.040 same as someone else's anyone else's anywhere around the globe. 0:02:35.080 --> 0:02:37.080 And I don't mean that as a feel good statement. 0:02:37.120 --> 0:02:42.760 That's just a biological fact. Anatomically modern humans radiated out 0:02:42.760 --> 0:02:46.200 of Africa only about sixty thousand years ago, and some 0:02:46.280 --> 0:02:49.080 turned left and became European, and others turned right and 0:02:49.120 --> 0:02:53.000 became Asian, and those who stayed remained African. But the 0:02:53.200 --> 0:02:57.120 timescale that we're talking about some tens of thousands of years. 0:02:57.639 --> 0:03:01.640 That is simply not enough time on the evolutionary timescale 0:03:01.919 --> 0:03:05.519 for brains to change in any meaningful way to change 0:03:05.560 --> 0:03:10.320 their operation or their algorithms. So we're all running the 0:03:10.360 --> 0:03:16.440 same version of the biological operating system, despite superficial differences 0:03:16.480 --> 0:03:19.800 in size and shape and culture and the costumes that 0:03:19.840 --> 0:03:22.359 we wear and how much melanin we have in our 0:03:22.400 --> 0:03:26.680 skin and so on. We can end up seeming pretty 0:03:26.720 --> 0:03:29.480 different as a result of these surface properties, but we 0:03:29.600 --> 0:03:36.160 have the same cognitive and emotional engine on the inside. Now. Unfortunately, 0:03:36.760 --> 0:03:40.200 one of the things that operating system does is it 0:03:40.240 --> 0:03:44.320 gives us all a very strong drive toward having in 0:03:44.400 --> 0:03:50.720 groups and outgroups. This tribalism is something that characterizes our species. 0:03:51.160 --> 0:03:53.880 We belong to groups based. 0:03:53.560 --> 0:03:56.520 On our neighborhood or the way we look, or our 0:03:56.560 --> 0:03:59.720 religion or our family or whatever it is, and we 0:03:59.800 --> 0:04:03.640 take to feel closer to our in groups and treat 0:04:03.680 --> 0:04:06.760 them better and listen to their opinions more strongly. 0:04:07.200 --> 0:04:10.280 Our outgroups not so much. When it comes to someone 0:04:10.600 --> 0:04:13.720 from the other side of the railroad tracks, or who 0:04:13.800 --> 0:04:17.640 happens to have grown up under a different deity, or 0:04:17.680 --> 0:04:22.039 who sings different songs or uses different language sounds to 0:04:22.080 --> 0:04:25.799 transmit information, or whatever it is, we tend to treat 0:04:25.880 --> 0:04:31.280 them with more suspicion. And in decades of neuroscience experiments, 0:04:31.320 --> 0:04:34.120 including those from my lab, we've been able to show 0:04:34.520 --> 0:04:37.880 that your brain simply does not care as much about 0:04:37.960 --> 0:04:42.680 people in your outgroup. At the extreme, your brain will 0:04:42.760 --> 0:04:45.040 view somebody in an outgroup the same way it will 0:04:45.120 --> 0:04:49.160 view an object, not like a fellow human the way 0:04:49.200 --> 0:04:51.560 it does with people in your in group, but instead 0:04:52.000 --> 0:04:56.560 as something that's not particularly worthy of empathy, something that 0:04:56.600 --> 0:04:58.279 you have to deal with in the same way that 0:04:58.320 --> 0:05:01.160 you would deal with a vice, rusts, or a cockroach 0:05:01.279 --> 0:05:03.920 or a rat. And if you're interested in more detail 0:05:04.000 --> 0:05:07.360 on this point, I dove deep into examples of this 0:05:07.600 --> 0:05:10.880 in episode twenty and one of the things I talked 0:05:10.920 --> 0:05:13.680 about was an experiment that we did in my lab 0:05:14.080 --> 0:05:18.000 that got at this in group outgroup difference, where all 0:05:18.040 --> 0:05:21.880 we needed were single word labels. So I'll just tell 0:05:21.880 --> 0:05:25.200 you the short version here of the experiment. We put 0:05:25.240 --> 0:05:29.640 you in brain imaging in fMRI, and you're looking at 0:05:29.720 --> 0:05:32.919 six hands on the screen and one of them gets 0:05:33.200 --> 0:05:37.720 stabbed by a syringe needle. Now, seeing that activates a 0:05:38.279 --> 0:05:42.600 network in your brain that we summarize as the pain matrix. 0:05:42.680 --> 0:05:46.120 Your brain is showing a big spike, and what that 0:05:46.200 --> 0:05:50.040 represents is you are feeling the other person's pain even 0:05:50.080 --> 0:05:53.279 though it's not your hand. The eras in your brain 0:05:53.279 --> 0:05:56.279 that care about pain are coming online. This is the 0:05:56.320 --> 0:06:01.640 basis of empathy, of feeling someone else else's pain. But 0:06:01.880 --> 0:06:11.040 now we label the six hands with one word labels Christian, Jewish, Muslim, Hindu, Scientologist, atheist, 0:06:11.360 --> 0:06:14.120 And now we measure people of all religions in the 0:06:14.160 --> 0:06:17.560 scanner as these different hands get stabbed. And what we 0:06:17.720 --> 0:06:21.880 found is that everyone has a larger brain response when 0:06:21.920 --> 0:06:25.240 it's their own group that's getting stabbed, and their brain 0:06:25.320 --> 0:06:28.720 shows a smaller response when it's any of the other 0:06:29.000 --> 0:06:33.479 five out groups. And this was true across religions and 0:06:33.720 --> 0:06:36.800 even for the atheist group, which cared more when the 0:06:36.839 --> 0:06:40.520 atheist hand gets stabbed than the other hands. So this 0:06:40.640 --> 0:06:43.560 is not an indictment of religion, but more broadly, it 0:06:43.680 --> 0:06:49.920 highlights your brain's exquisite skills at distinguishing in group from 0:06:49.960 --> 0:06:52.920 your outgroups and the fact that your brain simply does 0:06:53.000 --> 0:06:57.200 not spend as much energy on simulating what it is 0:06:57.440 --> 0:07:01.279 like to be someone who is in your If you 0:07:01.320 --> 0:07:04.599 want more details on that, please listen to episode twenty, 0:07:04.640 --> 0:07:07.039 which is called why does your brain care about some 0:07:07.160 --> 0:07:11.280 people more than others? And beyond the fact that we 0:07:11.320 --> 0:07:15.160 are naturally tribalistic, there's a third point to note, which 0:07:15.200 --> 0:07:18.400 is that every one of us feels like we know 0:07:18.720 --> 0:07:21.440 the truth, and it's not clear to us why other 0:07:21.520 --> 0:07:24.840 people don't see the truth as clearly as we do. 0:07:24.880 --> 0:07:29.800 They must be stupid or ill informed, or trolls or 0:07:30.240 --> 0:07:34.080 stubborn or who knows why they don't simply admit to 0:07:34.160 --> 0:07:38.360 what is so obviously the truth. And deep down most 0:07:38.400 --> 0:07:41.880 people have an intuition that if you could just simply 0:07:42.400 --> 0:07:46.040 shout loudly enough in all caps on social media, that 0:07:46.280 --> 0:07:49.800 everyone would agree with you, everyone would see the light. 0:07:50.480 --> 0:07:53.240 And this is true of everything, not just political positions. 0:07:53.320 --> 0:07:57.800 Take religion. There are two thousand active religions on this 0:07:57.920 --> 0:08:01.000 small planet, and many people feel that if they could 0:08:01.120 --> 0:08:04.760 just share with all the infidels why their religion is 0:08:04.840 --> 0:08:09.240 right and everyone else is wrong, that everyone would agree 0:08:09.280 --> 0:08:12.600 with them. Everyone would come to know that the religion 0:08:12.720 --> 0:08:17.120 you happened to grow up in is the correct one. Now, 0:08:17.160 --> 0:08:20.200 if there actually were one correct religion, we might expect 0:08:20.240 --> 0:08:23.920 it to spread everywhere equally, But obviously that's not the case. 0:08:24.200 --> 0:08:28.120 Religions remain clustered around where they started, and you're not 0:08:28.200 --> 0:08:32.360 going to see a blossoming of Islam in Boise, Idaho, 0:08:32.600 --> 0:08:35.560 just like you won't expect to see a blossoming of 0:08:35.600 --> 0:08:39.600 Protestantism in Mecca. But here's the interesting part. Whether we're 0:08:39.640 --> 0:08:44.920 examining one's political preferences, or religious affiliation or whatever, people 0:08:45.040 --> 0:08:50.640 are generally reluctant to change to consider the perspectives of 0:08:50.840 --> 0:08:55.520 other people. So, given the deep wiring that we have 0:08:55.760 --> 0:08:58.160 for having our in groups and for feeling we know 0:08:58.320 --> 0:09:01.319 the right answer because of our very limited internal models, 0:09:01.840 --> 0:09:05.520 this leads to the critical question about whether there would 0:09:05.520 --> 0:09:11.800 be any way for eight billion brains to find peaceful coexistence. Now, 0:09:11.840 --> 0:09:14.360 if you've been a regular listener to the podcast, you 0:09:14.440 --> 0:09:18.120 know that I'm generally highly optimistic about everything, but I 0:09:18.160 --> 0:09:21.160 would suggest the answer to the question of whether we 0:09:21.240 --> 0:09:26.120 will achieve peace is probably not. We're probably not going 0:09:26.200 --> 0:09:30.040 to find ourselves ever in some magical moment where all 0:09:30.080 --> 0:09:33.760 the newspapers say, Wow, nobody is fighting, we're taking a 0:09:33.840 --> 0:09:36.200 day off today, and we're not going to find some 0:09:36.320 --> 0:09:41.320 extraordinary presidential candidate where everyone says, yeah, that person's great. 0:09:41.400 --> 0:09:45.080 I think we can all agree presidential elections are always 0:09:45.160 --> 0:09:48.880 fairly close to fifty to fifty. Why this results from 0:09:48.920 --> 0:09:53.320 the enormous variety of internal models inside everyone's heads. There 0:09:53.440 --> 0:09:57.720 is no perfect candidate because no one can fit all 0:09:57.760 --> 0:10:02.120 the constraints at once of all all the different models. Okay, 0:10:02.160 --> 0:10:04.480 so if we take this all as our starting point 0:10:04.559 --> 0:10:08.240 for today, the question is what does this all mean 0:10:08.480 --> 0:10:12.800 for the future of conflict? Are we facing a future 0:10:13.120 --> 0:10:16.760 where wars and polarization will stay as they are or 0:10:16.840 --> 0:10:21.480 even grow worse. Well, that's certainly a reasonable fear. But 0:10:21.559 --> 0:10:24.760 a little while ago I discovered there was a movement 0:10:24.920 --> 0:10:29.080 of political scholars around the idea of how to have 0:10:29.679 --> 0:10:33.600 better conflict. Now, what does better conflict mean? We'll get 0:10:33.640 --> 0:10:36.400 into that in a second, but first I just want 0:10:36.440 --> 0:10:39.959 to say I've been moved and inspired by this movement. 0:10:40.000 --> 0:10:43.080 It feels like it's the opposite of what we see 0:10:43.120 --> 0:10:46.959 online every day, where people pick aside and scream and 0:10:47.080 --> 0:10:50.280 yell for it or fight and die for it at 0:10:50.280 --> 0:10:57.760 the cost of not meaningfully considering the complexity of human situations. 0:10:58.600 --> 0:11:01.640 Just take what's happening in the middle least. Typically, if 0:11:01.679 --> 0:11:05.880 someone points out that the history there is extraordinarily multi 0:11:06.000 --> 0:11:10.079 layered and complex, people will often say, no, this is 0:11:10.120 --> 0:11:13.040 actually quite simple, and then they will give their version 0:11:13.120 --> 0:11:16.040 of the history. But the extraordinary part is that people 0:11:16.080 --> 0:11:19.199 on both sides will deliver that same line. They will say, 0:11:19.200 --> 0:11:22.920 it's actually very simple, And to me, that's a barometer 0:11:23.040 --> 0:11:26.520 that tells me the situation is not so simple. But instead, 0:11:26.760 --> 0:11:30.520 like most human conflict, it's full of different paths that 0:11:30.559 --> 0:11:33.880 you can take through the history to come to a 0:11:34.040 --> 0:11:38.079 variety of conclusions. And we witness this sort of complexity 0:11:38.440 --> 0:11:43.160 whenever you have millions of people with all personalities involved 0:11:43.640 --> 0:11:48.040 in conflict on all sides. You have sinners and saints, 0:11:48.120 --> 0:11:52.440 and you have psychopaths and peace seekers, and no one 0:11:52.520 --> 0:11:57.400 can hope to derive a meaningful solution without doing the 0:11:57.480 --> 0:12:01.040 hard work of digging in and trying to understand what 0:12:01.240 --> 0:12:06.439 happens between groups of millions of humans, rather than imagining 0:12:06.480 --> 0:12:09.800 there's a good guy side and a bad guy side. So, 0:12:09.960 --> 0:12:12.880 as I said, I was very moved to discover that 0:12:12.960 --> 0:12:17.720 some people really take on this sort of political intellectual 0:12:17.800 --> 0:12:21.200 humility and try to figure out what do we do 0:12:21.760 --> 0:12:23.840 with humans? What do we do about the fact that 0:12:23.840 --> 0:12:26.840 they're never going to agree, There's always going to be conflict. 0:12:27.080 --> 0:12:32.000 How do we make conflict more productive rather than simply 0:12:32.080 --> 0:12:36.240 being about bloodshed? So to explore this, I called up 0:12:36.280 --> 0:12:40.680 my colleague Jonathan Stray. Jonathan is a former journalist who 0:12:40.720 --> 0:12:43.800 is now a senior scientist at the Berkeley Center for 0:12:43.920 --> 0:12:48.240 Human Compatible AI, where he works on the algorithms that 0:12:48.400 --> 0:12:52.960 feed up content to online and how their operation affects 0:12:53.000 --> 0:12:56.839 things like polarization. And he writes a terrific newsletter called 0:12:57.160 --> 0:13:05.760 The Better Conflict Bulletin. So here's John Jonathan. So you 0:13:05.840 --> 0:13:09.400 are a conflict researcher. Tell us about what that means. 0:13:09.920 --> 0:13:15.040 Well, I try to study what causes people to disagree 0:13:15.040 --> 0:13:19.120 and then how that disagreement transforms into polarization and eventually 0:13:19.200 --> 0:13:23.520 violence if unchecked, And how to build computer controlled media 0:13:23.559 --> 0:13:25.760 systems like social media that will prevent that. 0:13:26.280 --> 0:13:30.120 I know that you think about conflict in terms of 0:13:30.240 --> 0:13:33.160 good and bad conflict. So what does that mean? 0:13:33.800 --> 0:13:37.240 Well, we don't want no conflict in society. Conflict is 0:13:37.360 --> 0:13:39.960 so first of all, it's inevitable. We all disagree, we 0:13:40.000 --> 0:13:45.520 want different things unavoidably. But sometimes when we disagree, it 0:13:45.600 --> 0:13:47.520 comes to something productive, you know, we talk about it, 0:13:47.559 --> 0:13:50.520 we figure something out, and sometimes it doesn't right, Sometimes 0:13:50.520 --> 0:13:54.040 it escalates to some sort of zero sum game or 0:13:54.080 --> 0:13:58.679 even violence. The idea is, when we think about conflict, 0:13:58.679 --> 0:14:01.000 we're not trying to prevent people from fighting. We're trying 0:14:01.040 --> 0:14:03.080 to have them fight in some sort of productive way. 0:14:03.800 --> 0:14:05.480 So what does that look like? What does that mean? 0:14:06.000 --> 0:14:09.199 Well, the simplest example might be violent versus nonviolent, right, 0:14:09.280 --> 0:14:13.080 so we can all understand the difference between you know, 0:14:13.520 --> 0:14:15.920 people trying to get their way by using force or 0:14:15.960 --> 0:14:20.080 getting their way by discussion, yes, but also non violent 0:14:20.160 --> 0:14:24.520 conflict tactics, you know, protests. You know, there's famous historical 0:14:24.600 --> 0:14:29.040 figures like doctor Martin, Luther King or Gandhi who really 0:14:29.080 --> 0:14:33.880 developed these techniques and showed that you could achieve you know, 0:14:34.000 --> 0:14:38.200 large political victories without using physical force. So that's that's 0:14:38.240 --> 0:14:41.200 maybe the most obvious way but there's a bunch of 0:14:41.200 --> 0:14:42.800 other stuff when you start thinking about this. 0:14:43.040 --> 0:14:45.480 So maybe you think about this at a personal level. 0:14:45.520 --> 0:14:47.520 You know, we've all had the experience of an argument 0:14:47.560 --> 0:14:50.320 where you know, we said some ugly things that can't 0:14:50.360 --> 0:14:52.560 be unsaid, and you feel worse at the end of it. 0:14:52.920 --> 0:14:54.280 And then I think most of us have had the 0:14:54.360 --> 0:14:56.600 experience of an argument where you know, there were some 0:14:56.680 --> 0:15:00.200 hard truths, but it was honest, it was carrying, it 0:15:00.240 --> 0:15:03.200 was empathetic, and even if we couldn't solve the problem 0:15:03.280 --> 0:15:05.480 at the end of it, we still have a positive 0:15:05.520 --> 0:15:06.400 regard for each other. 0:15:06.600 --> 0:15:08.520 And so that would be an example of good versus 0:15:08.520 --> 0:15:09.200 bad conflict. 0:15:09.720 --> 0:15:11.880 So one of the things that has always intrigued me 0:15:12.000 --> 0:15:14.880 is that we all have this illusion that we know 0:15:15.040 --> 0:15:17.640 the truth, that we have a complete story of what's 0:15:17.720 --> 0:15:21.600 going on, and therefore other people that disagree with us 0:15:21.680 --> 0:15:26.240 seem like they're trolls, or they're uninformed, or they're disingenuous. 0:15:26.720 --> 0:15:30.440 And so the question is, how do we address that 0:15:30.720 --> 0:15:33.440 illusion that we have. How do we get ourselves a 0:15:33.520 --> 0:15:36.400 little bit outside our own fish bowls to see the 0:15:36.440 --> 0:15:37.040 other side. 0:15:37.560 --> 0:15:40.200 Yeah, no, that's a great question. So there's a bunch 0:15:40.200 --> 0:15:42.000 of ways of answering that. So there's I was at 0:15:42.040 --> 0:15:45.320 a conference yesterday on the topic of intellectual humility, the 0:15:45.400 --> 0:15:48.600 idea that actually we don't know the answer, there's always 0:15:48.600 --> 0:15:51.600 something we're missing. And this is important not just as 0:15:51.720 --> 0:15:55.400 an epistemological practice, that is, to be more correct in 0:15:55.440 --> 0:15:59.680 what we believe, but as a relational practice, meaning it 0:15:59.800 --> 0:16:02.040 changed is how we relate and how we can relate 0:16:02.040 --> 0:16:04.880 to other people. And I think part of this is 0:16:05.680 --> 0:16:08.520 we always have a partial picture, and we tend to, 0:16:09.080 --> 0:16:13.000 in conflict situations, believe that the other person actually holds 0:16:13.040 --> 0:16:16.560 more extreme views than they do. So there's really good 0:16:16.600 --> 0:16:18.600 evidence of that. This is the case in the American 0:16:18.600 --> 0:16:21.520 Culture War, where if you ask one side what the 0:16:21.520 --> 0:16:24.320 other side thinks, they will imagine the other side is 0:16:24.360 --> 0:16:28.920 actually more extreme than they are. So conflict in particular 0:16:29.520 --> 0:16:31.480 is prone to these misperceptions of the other. 0:16:32.320 --> 0:16:35.400 So give us an example of how we know that 0:16:35.400 --> 0:16:37.880 that each side thinks the other side is more extreme. 0:16:38.400 --> 0:16:39.720 Yeah. Yeah, there's tons of work on this. 0:16:39.760 --> 0:16:41.840 So there's a group called the Perception Gap which has 0:16:41.880 --> 0:16:45.040 done a bunch of research from this. And so, for example, 0:16:45.120 --> 0:16:52.480 if you ask Republicans how many Democrats would support completely 0:16:52.480 --> 0:16:56.200 open borders, they'll say something like seventy five percent. If 0:16:56.200 --> 0:16:58.400 you ask Democrats how many of them would support completely 0:16:58.480 --> 0:17:03.160 open borders, it's something like, so they're twenty or twenty 0:17:03.160 --> 0:17:06.560 five percentage points off. And this pattern is consistent across 0:17:06.560 --> 0:17:09.479 a range of issues, and it's bidirectional as well. So 0:17:09.720 --> 0:17:11.880 you know, if you ask Republicans or if you ask 0:17:11.920 --> 0:17:16.159 democrats something like, you know, how many Republicans would say that, 0:17:17.200 --> 0:17:20.000 you know, everyone should have access to as many guns 0:17:20.040 --> 0:17:22.840 as they want, Democrats will guess that number a lot 0:17:22.880 --> 0:17:24.760 higher than Republicans will say. 0:17:24.880 --> 0:17:28.320 So, speaking of Democrats and Republicans, here's the question. Are 0:17:28.359 --> 0:17:33.239 we more polarized currently in America? Is that just an 0:17:33.280 --> 0:17:35.480 impression or is that backed up by statistics? 0:17:36.119 --> 0:17:39.280 No, Unfortunately, it's real, and you can measure it a 0:17:39.280 --> 0:17:41.920 lot of different ways. You can look at where people 0:17:41.960 --> 0:17:44.639 are on political issues, you can look at how people 0:17:44.680 --> 0:17:48.400 feel about each other. You can look at congressional voting patterns, 0:17:48.440 --> 0:17:51.720 whether you know members of Congress will cross the aisle 0:17:51.800 --> 0:17:54.680 to vote on each other's builts, and all of these 0:17:54.880 --> 0:17:57.760 measures show the same basic pattern, which is that polarization 0:17:57.840 --> 0:18:01.120 has been increasing since the the mid seventies or maybe 0:18:01.119 --> 0:18:04.760 the early eighties and is now at historically high levels. 0:18:05.160 --> 0:18:07.399 And it's not just America, it's several other countries in 0:18:07.440 --> 0:18:08.560 the world as well. 0:18:08.640 --> 0:18:11.359 Right now, I know you and I have independently made 0:18:11.440 --> 0:18:15.200 arguments about why this is not entirely about social media, 0:18:15.240 --> 0:18:17.560 and that, of course seems like a good thing to 0:18:17.600 --> 0:18:19.840 point to is the fact that this started in the seventies. 0:18:19.880 --> 0:18:24.240 But tell us your sense of the involvement and the 0:18:24.280 --> 0:18:27.000 non involvement of social media in this polarization. 0:18:27.760 --> 0:18:29.920 Yeah, well, there's definitely some involvement. 0:18:30.119 --> 0:18:34.000 There are a number of ways that social media, in 0:18:34.040 --> 0:18:38.160 particular social media algorithms, that is to say, the systems 0:18:38.160 --> 0:18:40.120 that decide what we see. 0:18:40.040 --> 0:18:42.639 And in what order, can have some involvement. 0:18:43.160 --> 0:18:45.240 So probably a lot of your listeners have heard of 0:18:45.240 --> 0:18:47.520 the idea of the filter bubble, the idea that, you know, 0:18:47.600 --> 0:18:51.560 we're each trapped in this sort of algorithmically produced reality 0:18:51.560 --> 0:18:54.639 where we don't see the other side. The evidence tends 0:18:54.680 --> 0:18:57.560 to be against that. We actually do see quite a 0:18:57.600 --> 0:18:59.720 lot of cross cutting content, you know, even if it's 0:18:59.760 --> 0:19:02.920 you know, rage clicking on an article that has a 0:19:02.920 --> 0:19:06.199 different viewpoint, but there's a bunch of other things going on. So, 0:19:06.480 --> 0:19:12.000 for example, most social media and other systems which select 0:19:12.000 --> 0:19:15.360 content for us, like news recommenders and so forth, rank 0:19:15.480 --> 0:19:18.560 things based on basically the number of clicks they get engagement, 0:19:19.320 --> 0:19:22.040 And we really do pay more attention to things that 0:19:22.080 --> 0:19:24.520 are valuable to us, but also we pay more attention 0:19:24.600 --> 0:19:31.320 to things that are threatening, offensive, fearful. So this approach 0:19:31.359 --> 0:19:34.160 of ranking things by how much attention we give them 0:19:34.400 --> 0:19:54.840 also tends to amplify more extreme or scarier items. 0:19:57.800 --> 0:20:00.159 Now, I did an episode a few episodes ag Go 0:20:00.240 --> 0:20:05.000 where I mentioned for the Iroquois Native Americans up in 0:20:05.240 --> 0:20:09.479 essentially Wisconsin and Canada, some hundreds of years ago, they 0:20:09.480 --> 0:20:13.160 were having bloody battles between these different tribes all the time, 0:20:13.680 --> 0:20:16.840 and they had a new leader come in who came 0:20:16.880 --> 0:20:19.000 to be known as the Great Peacemaker because one of 0:20:19.000 --> 0:20:22.680 the things he did is assigned everybody in the tribes 0:20:23.080 --> 0:20:25.879 to different clans. So you might be a member of 0:20:25.880 --> 0:20:28.920 the Turtle clan and I'm a member of the Heron clan, 0:20:28.960 --> 0:20:31.199 even though we're members of the same tribe. And it 0:20:31.280 --> 0:20:35.000 turns out these allegiances were cross cutting, so that each 0:20:35.080 --> 0:20:37.760 person had their allegiance to their tribe and also to 0:20:37.800 --> 0:20:41.200 their clan. And these weren't equivalent, and that ended up 0:20:41.240 --> 0:20:45.000 bringing peace because now it wasn't so simple to have 0:20:45.080 --> 0:20:48.360 a clear in group and out group because they were mixed. 0:20:48.760 --> 0:20:51.720 And I know that you in your work with social 0:20:51.760 --> 0:20:55.680 media have been looking at something similar to this, which 0:20:55.720 --> 0:21:00.600 is this issue of how an algorithm should rank posts. 0:21:00.680 --> 0:21:01.600 So tell us about that. 0:21:02.640 --> 0:21:05.720 Yeah, so that is a great example. I've heard of 0:21:05.760 --> 0:21:09.439 this Ariquais example as well. And what is happening is 0:21:09.480 --> 0:21:13.480 conflict is tied up with identity. We fight basically against 0:21:13.560 --> 0:21:17.200 people who are different than us. And in the last 0:21:17.200 --> 0:21:20.480 few decades in this country, identity has become more and 0:21:20.600 --> 0:21:24.560 more collapsed into a single dimension of you know, left 0:21:24.640 --> 0:21:28.080 versus right, red versus blue, Republicans versus Democrats. 0:21:28.200 --> 0:21:29.640 It didn't actually used to be this way. 0:21:30.080 --> 0:21:34.080 People didn't identify with their political parties in the same 0:21:34.080 --> 0:21:38.880 way a generation ago. And also there was much more 0:21:39.000 --> 0:21:44.000 mixed identification at the national versus local level. So you know, 0:21:44.080 --> 0:21:46.360 you would have some town where, you know, you had 0:21:46.840 --> 0:21:50.080 a Republican mayor, but you had, you know, progressive politics 0:21:50.080 --> 0:21:52.560 because that's what the people there believed. 0:21:52.160 --> 0:21:52.760 In and wanted. 0:21:53.760 --> 0:21:58.360 Politics has become very nationalized. There's less local news, there's 0:21:58.480 --> 0:21:59.760 less split ticket voting. 0:22:00.520 --> 0:22:02.440 Every issue takes place. 0:22:02.200 --> 0:22:05.359 At this huge scale where there's only two ends in spectrum. 0:22:06.280 --> 0:22:08.960 So how you would have to reverse this is you 0:22:09.000 --> 0:22:13.440 would have to have some sort of sense of not 0:22:13.480 --> 0:22:17.479 necessarily local, but community level values. 0:22:17.560 --> 0:22:18.360 Right, it would have to be. 0:22:18.280 --> 0:22:23.200 Okay for you know, specific groups online or perhaps discussion 0:22:23.400 --> 0:22:27.399 to have opinions that just don't quite fall along this 0:22:27.560 --> 0:22:28.600 left right axis. 0:22:29.160 --> 0:22:32.920 And what's the reason that we have seen more sorting 0:22:33.240 --> 0:22:36.439 into these groups. What's the reason that particular issues go 0:22:36.600 --> 0:22:39.280 together that don't necessarily belong together. 0:22:40.000 --> 0:22:44.359 Well, again, it's an identity thing. It's one of the 0:22:44.400 --> 0:22:47.560 things that makes all of these issues sort of one 0:22:48.040 --> 0:22:51.960 big blob where you only have two choices is merely 0:22:52.080 --> 0:22:55.920 exposure to people who are far away from us, either 0:22:56.000 --> 0:22:59.720 physically or social. So you know, maybe you only talk 0:22:59.760 --> 0:23:01.800 to your neighbors in your town, or you only read 0:23:01.800 --> 0:23:03.800 the local newspaper, and so you could have this sort 0:23:03.840 --> 0:23:07.399 of quixotic local politics in a way that was healthy. 0:23:07.720 --> 0:23:10.840 But now we are exposed to people all across the 0:23:10.920 --> 0:23:13.840 country and indeed the world on a daily basis. So 0:23:13.880 --> 0:23:17.439 now we're comparing our point of view to someone you 0:23:17.440 --> 0:23:20.400 know that will never meet, who's far away from us. 0:23:20.960 --> 0:23:25.160 And there's simulations that show you don't need any antagonism 0:23:25.359 --> 0:23:27.640 for people to sort themselves into two big tribes. 0:23:27.680 --> 0:23:30.000 All you need is when you talk. 0:23:29.840 --> 0:23:34.320 To someone and they're you know, near enough like you, 0:23:34.320 --> 0:23:36.720 you get a little bit closer to them, right, which 0:23:36.760 --> 0:23:38.720 is a natural human thing. When two people talk, they 0:23:38.720 --> 0:23:40.480 tend to you know, converge a little on their on 0:23:40.520 --> 0:23:46.040 their views and identity. But because that those interactions are 0:23:46.080 --> 0:23:49.520 now happening all across the country, we're coalescing into sort 0:23:49.560 --> 0:23:50.639 of two big groups. 0:23:51.080 --> 0:23:52.520 It's it's kind of a paradox, right. 0:23:52.560 --> 0:23:54.440 We want to be able to talk to people who 0:23:54.480 --> 0:23:56.439 are very far away from us, you know, that's the 0:23:56.480 --> 0:23:58.440 promise of these online networks, and. 0:23:58.440 --> 0:24:00.880 Yet it seems to make our con clicks global. 0:24:01.520 --> 0:24:04.400 And there are network science models on this right that 0:24:04.600 --> 0:24:07.840 demonstrate how groups end up dividing like this. And what 0:24:07.960 --> 0:24:12.520 is the evidence that these network science models actually cash 0:24:12.560 --> 0:24:13.840 out in real life? 0:24:14.520 --> 0:24:17.159 Yeah, So these models developed actually from models built in 0:24:17.200 --> 0:24:22.119 the nineteen sixties to study physical racial segregation. And what 0:24:22.160 --> 0:24:26.240 they showed is that you don't need racism in the 0:24:26.320 --> 0:24:29.119 sense of not wanting to live near people who are 0:24:29.160 --> 0:24:31.919