Full Transcript for Supercommunicators, by Charles Duhigg

00:04:17.848 --> 00:04:20.036 And today's book is Super Communicators. 59 00:04:20.883 --> 00:04:25.447 How to Unlock a Secret Language of Connection by Charles Duhigg, published in 2024. 60 00:04:26.187 --> 00:04:27.348 Duhigg is a journalist. 61 00:04:27.388 --> 00:04:29.150 He wrote The Power of Habit. 62 00:04:29.190 --> 00:04:39.639 He won a Pulitzer Prize for his explanatory reporting, and he spent several years working through the research on why some conversations go well and most of them don't. 63 00:04:40.219 --> 00:04:47.445 The book stitches together neuroscience, psychology, and negotiation studies with a set of stories that are well told. 64 00:04:47.465 --> 00:04:48.466 And on the surface, it's 65 00:04:48.726 --> 00:04:54.050 This is a book about two people in a room, a doctor and a patient, or a husband and a wife. 66 00:04:54.110 --> 00:04:58.274 It's about everyday conversations, often at the individual level. 67 00:04:58.754 --> 00:05:05.820 And yet, it's fully applicable to our field because the thing this book explains, how two people fail to hear each other, 68 00:05:06.440 --> 00:05:17.866 while they're both paying close attention, both being reasonable, both trying, is the same failure we run at scale every day on files worth millions, if not billions of dollars. 69 00:05:18.306 --> 00:05:21.768 And it's just much harder to see that from that distance. 70 00:05:21.968 --> 00:05:25.430 I'll tell you now that there's a tension in this episode. 71 00:05:25.490 --> 00:05:27.391 I'm not going to be able to fully resolve it for you, though. 72 00:05:27.731 --> 00:05:31.593 And Duhigg's answer to almost every problem is to get closer to somebody. 73 00:05:32.493 --> 00:05:38.619 And we work in a business that operates often at a distance by design, by budget, by necessity, whatever it is. 74 00:05:38.940 --> 00:05:44.765 And I've been stewing over this for several weeks since I reread the book, and I haven't figured out how to fully adapt it at scale. 75 00:05:44.846 --> 00:05:46.207 So fair warning on that. 76 00:05:46.687 --> 00:05:48.709 We'll explore this challenge together. 77 00:05:48.729 --> 00:05:52.213 We'll do it in this episode, but I'm not going to pretend I have the answer on it. 78 00:05:52.393 --> 00:06:00.277 Now, if you listen to the Littman episode, the very first episode in this canon, you'll remember what Walter Littman's core argument is. 79 00:06:00.378 --> 00:06:06.401 And that's that people navigate the world by the pictures in their heads, not by the world itself. 80 00:06:06.581 --> 00:06:14.866 This book is what happens when two people with different pictures have to make a decision together or when they have to converse together. 81 00:06:15.606 --> 00:06:20.252 Now, Charles Duhigg wrote this book because he felt he was failing at his own job. 82 00:06:20.272 --> 00:06:22.795 So a few years ago, he was asked to manage a complex project. 83 00:06:23.416 --> 00:06:27.361 He'd never managed anyone before, but he'd worked for plenty of bosses. 84 00:06:27.502 --> 00:06:29.124 He had an MBA from Harvard Business School. 85 00:06:29.264 --> 00:06:31.447 And as he puts it, he communicated for a living. 86 00:06:31.487 --> 00:06:32.047 He's a journalist. 87 00:06:32.107 --> 00:06:32.989 How hard could it be? 88 00:06:34.302 --> 00:06:35.484 very hard, it turned out. 89 00:06:35.844 --> 00:06:48.161 He was fine at the schedules and the logistics and the project management, but what he couldn't do was connect with people he was working with, at least not to the standard and expectations that he had set for himself. 90 00:06:48.362 --> 00:06:49.283 And that challenge 91 00:06:49.403 --> 00:06:50.744 followed him home. 92 00:06:50.804 --> 00:07:03.116 He describes coming back from a bad day and complaining to his wife about a colleague and getting a perfectly sensible suggestion in return, and then getting angry because what he actually wanted was for her to just tell him he was right to be annoyed. 93 00:07:03.256 --> 00:07:07.059 His kids would try to talk to him and he'd half listen until they wandered off. 94 00:07:07.120 --> 00:07:07.520 And he'd 95 00:07:07.720 --> 00:07:11.502 He could see all of it happening sort of in slow motion and just really couldn't stop it. 96 00:07:11.602 --> 00:07:14.683 So here's a man who won a Pulitzer for explanatory reporting. 97 00:07:14.903 --> 00:07:16.844 He wrote a bestselling book about human behavior. 98 00:07:16.984 --> 00:07:20.566 His entire professional identity is built on understanding why people do what they do. 99 00:07:20.646 --> 00:07:22.647 And he couldn't hear his own family. 100 00:07:22.687 --> 00:07:24.628 He couldn't communicate well with his own family. 101 00:07:24.848 --> 00:07:26.850 And that's the problem that the book is trying to solve. 102 00:07:27.390 --> 00:07:33.654 And in many ways, Duhigg is credible precisely because he admits to finding this to be a very difficult task. 103 00:07:34.135 --> 00:07:37.737 And as a result, he takes us on what is a learning journey. 104 00:07:37.798 --> 00:07:38.718 And that's what the book is. 105 00:07:39.339 --> 00:07:41.040 Now, let me give you the scene that opens the book. 106 00:07:41.420 --> 00:07:43.762 I do think it sets the stage wonderfully. 107 00:07:44.062 --> 00:07:47.445 And it's a group of scientists, psychologists, researchers. 108 00:07:47.845 --> 00:07:51.909 They've brought in this man to interview, and his name is Felix. 109 00:07:51.969 --> 00:07:54.211 Felix is exceptionally easy. 110 00:07:54.351 --> 00:08:00.376 People love talking to him because they always come away feeling a little smarter, funnier, more interesting. 111 00:08:00.777 --> 00:08:07.543 And if you've had nothing in common with Felix, the conversation inevitably revealed the kinds of opinions, experiences, or friends you shared. 112 00:08:07.743 --> 00:08:10.865 And it felt as if he heard you, like you had some kind of bond. 113 00:08:10.885 -->

00:08:11.326 Yeah. 114 00:08:11.666 --> 00:08:13.787 That's why the scientists had sought him out. 115 00:08:14.147 --> 00:08:18.170 They were told they had to speak to him to better understand what a super communicator looks like. 116 00:08:18.610 --> 00:08:28.095 Felix had been with the Federal Bureau of Investigation for two decades, and he had joined after college and a stint in the military, and then he had spent a few years as an agent in the field. 117 00:08:28.555 --> 00:08:32.377 And that's where his superiors had first taken note of his easy way with others. 118 00:08:32.958 --> 00:08:40.642 A series of promotions followed, and eventually he landed as a senior regional administrator with a mandate to serve as an all-around negotiator. 119 00:08:41.282 --> 00:08:49.967 As Duhigg writes, Felix was the guy who could coax statements from reluctant witnesses, convince fugitives to turn themselves in, console families as they grieved. 120 00:08:50.487 --> 00:09:00.032 He once persuaded a man who barricaded himself in a room with six cobras and 19 rattlesnakes and an iguana to come out peacefully and then name his accomplices in an animal smuggling rake. 121 00:09:00.052 --> 00:09:02.673 So just really effective at his craft. 122 00:09:04.070 --> 00:09:09.233 In other words, the FBI had a crisis negotiation unit for hostage situations. 123 00:09:09.393 --> 00:09:13.776 And when things got unusually complicated, they called someone like Felix. 124 00:09:14.396 --> 00:09:17.838 And this is why the social scientists wanted to spend time with him. 125 00:09:18.078 --> 00:09:19.639 So they spent an hour with him. 126 00:09:19.919 --> 00:09:25.763 And for most of that hour, Felix asked them questions about their work, their lives, their families. 127 00:09:25.823 --> 00:09:29.385 And near the end, one of them, a psychology professor, gets a little impatient. 128 00:09:29.865 --> 00:09:35.128 She says, this has been wonderful, but I don't feel any closer to understanding what you actually do. 129 00:09:35.929 --> 00:09:37.890 Why do you think so many people recommend we talk to you? 130 00:09:38.750 --> 00:09:39.651 Fair question, he says. 131 00:09:39.911 --> 00:09:42.473 But before I answer it, you mentioned earlier that you're a single mother. 132 00:09:42.773 --> 00:09:45.975 And I imagine there's a lot of juggling to motherhood and a career. 133 00:09:46.875 --> 00:09:51.338 This might seem a bit unusual, but I'm wondering, what would you tell someone who was getting a divorce? 134 00:09:52.681 --> 00:09:53.522 She goes quiet for a beat. 135 00:09:54.163 --> 00:09:56.125 Then she says, all right, I'll play along. 136 00:09:56.385 --> 00:09:57.146 I've got plenty of advice. 137 00:09:57.686 --> 00:10:02.572 When I separated from my husband, and then he stops her gently, he says, I don't actually need an answer. 138 00:10:03.032 --> 00:10:05.515 I just want to point out that you're in a room full of professional colleagues. 139 00:10:05.935 --> 00:10:11.802 You've known me for under an hour, and you were about to talk about one of the most intimate parts of your life. 140 00:10:11.982 --> 00:10:12.703 Then he explains why. 141 00:10:13.713 --> 00:10:14.853 Not magic, not charm. 142 00:10:15.814 --> 00:10:16.654 He had listened closely. 143 00:10:16.854 --> 00:10:20.276 He had asked questions that invited people to be a bit vulnerable. 144 00:10:20.976 --> 00:10:24.618 Everybody in that room, including him, had given up something real about themselves. 145 00:10:24.658 --> 00:10:27.099 He'd encouraged them to explain how they saw the world. 146 00:10:27.319 --> 00:10:29.480 And then he proved to them that he actually heard. 147 00:10:29.860 --> 00:10:37.543 And whenever somebody said something with feeling in it, even when they hadn't noticed they were doing it, he answered by naming a feeling of his own. 148 00:10:38.773 --> 00:10:39.894 Dozens of small choices. 149 00:10:40.595 --> 00:10:41.837 None of them are particularly dramatic. 150 00:10:42.397 --> 00:10:45.581 All of that adding up to a room where a stranger felt safe. 151 00:10:45.801 --> 00:10:51.227 And he says the line that Charles Duhigg will use as his central call to action throughout this book. 152 00:10:51.948 --> 00:10:53.570 It's a set of skills, he tells them. 153 00:10:53.810 --> 00:10:55.012 There's nothing magical about it. 154 00:10:55.432 --> 00:10:57.534 Anyone can learn to be a super communicator. 155 00:10:58.354 --> 00:11:09.078 It's a point we'll see throughout the book that we should not assume that the best communicators come at it naturally and that there is a natural divide between those who are naturally gifted communicators and those who aren't. 156 00:11:09.719 --> 00:11:14.521 Charles Zuhigg will use the book to argue that this is a skill we can all learn, that it's trainable. 157 00:11:14.721 --> 00:11:24.485 And even someone as effective as Felix, who is, by the way, an incredibly shy kid, was able to become an effective super communicator by applying himself. 158 00:11:25.672 --> 00:11:31.495 George Bernard Shaw is supposed to have said that the single biggest problem with communication is the illusion that it has taken place. 159 00:11:32.055 --> 00:11:36.437 Of course, it's a clever line, but it does speak to what we covered in our Frank Luntz episode. 160 00:11:36.917 --> 00:11:40.358 It is a variation on that theme that underpins this book. 161 00:11:41.079 --> 00:11:44.360 We walk out of rooms believing something happened, that we delivered our message. 162 00:11:44.960 --> 00:11:51.123 Often, we don't realize that we had a totally different conversation than the other person or the other party. 163 00:11:52.509 --> 00:11:59.671 If you take one thing from this episode, the core idea, take this, the person receiving your message isn't one person. 164 00:12:00.691 --> 00:12:07.273 They're a stack of identities and only one of them, only one of those identities is listening at a time. 165

00:12:07.713 --> 00:12:12.954 Duhigg's organizing claim is that most discussions are actually one of three different conversations. 166 00:12:13.134 --> 00:12:16.575 There's a practical one, which asks, what is this really about? 167 00:12:17.295 --> 00:12:20.636 There's an emotional one, which asks, how do we feel? 168 00:12:21.416 --> 00:12:23.777 And there's a social one, which asks, who are we? 169 00:12:24.777 --> 00:12:25.658 Who are we to each other? 170 00:12:25.838 --> 00:12:27.578 And who am I in this room? 171 00:12:28.278 --> 00:12:32.860 And we move between all three constantly, often inside of a single exchange. 172 00:12:32.900 --> 00:12:35.941 So a colleague comes to you for help thinking through a problem at work. 173 00:12:36.301 --> 00:12:37.382 That's a practical discussion. 174 00:12:37.822 --> 00:12:39.663 Two minutes later, he admits he's stressed about it. 175 00:12:39.803 --> 00:12:40.643 Well, that's emotional. 176 00:12:41.063 --> 00:12:44.164 And then two minutes after that, he's worrying about how it'll look to the rest of the team. 177 00:12:44.484 --> 00:12:45.925 And that's a social conversation. 178 00:12:46.785 --> 00:12:48.546 Now, the useful part isn't the taxonomy. 179 00:12:48.606 --> 00:12:52.048 The useful part is what happens when two people are in different conversations. 180 00:12:52.489 --> 00:12:56.071 Duhigg's phrase for it is that they're speaking different cognitive languages. 181 00:12:56.631 --> 00:13:03.676 And I think that framing matters because it means the failure isn't rudeness and it isn't inattention and it isn't a lack of goodwill. 182 00:13:04.296 --> 00:13:05.377 It's a category error. 183 00:13:05.777 --> 00:13:10.160 You can be listening intently to somebody and still be in the wrong conversation with them. 184 00:13:11.119 --> 00:13:12.119 You complain about your boss. 185 00:13:12.580 --> 00:13:14.300 Your partner suggests you invite him to lunch. 186 00:13:14.841 --> 00:13:17.222 That's a practical answer to an emotional statement. 187 00:13:17.362 --> 00:13:18.882 And it produces a fight. 188 00:13:19.362 --> 00:13:21.223 Maybe not a big fight, just a bit of friction. 189 00:13:21.843 --> 00:13:25.745 And both of you walk away feeling misunderstood because you both were. 190 00:13:25.925 --> 00:13:27.326 Which brings us back to a day. 191 00:13:28.166 --> 00:13:30.867 He was in the practical conversation. 192 00:13:30.927 --> 00:13:35.508 He wanted to provide his patients with options, with evidence, probabilities, trade-offs. 193 00:13:36.228 --> 00:13:42.610 But his patients were somewhere else entirely in the emotional conversation, mostly, and maybe in the social one as well. 194 00:13:43.450 --> 00:13:44.531 How will my family take this? 195 00:13:44.731 --> 00:13:46.591 Am I ready to think about dying? 196 00:13:46.811 --> 00:13:48.952 What kind of man decides to just wait and see? 197 00:13:50.209 --> 00:13:51.410 He was speaking to the patient. 198 00:13:51.450 --> 00:13:53.452 The patient wasn't in the room. 199 00:13:53.552 --> 00:13:58.537 A husband was, a father was, a man doing math about how much time he had left on this planet. 200 00:13:58.737 --> 00:14:00.599 So here's the second formulation. 201 00:14:00.799 --> 00:14:05.223 We treat communication as a thing we already know how to do, which is 202 00:14:05.563 --> 00:14:07.885 Probably precisely why we're so bad at it. 203 00:14:08.265 --> 00:14:11.547 We assume competence because we've been doing it since we were two years old. 204 00:14:11.967 --> 00:14:20.392 And so the entire discipline of figuring out which conversation is actually occurring gets skipped every day on files where the stakes are enormous. 205 00:14:20.632 --> 00:14:31.218 So for the rest of this episode, we're going to look at how the live identity, the selected identity, how that gets set, which of those cells is the one doing the listening. 206 00:14:31.919 --> 00:14:33.720 So with that, let's explore 207 00:14:34.560 --> 00:14:35.661 Four mechanisms. 208 00:14:35.821 --> 00:14:39.926 One is which self is in the room and who put it there. 209 00:14:40.186 --> 00:14:42.148 So you are a lot of things at once. 210 00:14:42.329 --> 00:14:43.109 You might be a parent. 211 00:14:43.129 --> 00:14:43.810 You might be a daughter. 212 00:14:44.030 --> 00:14:44.831 Someone's employee. 213 00:14:44.851 --> 00:14:45.712 Someone else's boss. 214 00:14:45.892 --> 00:14:47.254 A person from a particular town. 215 00:14:47.274 --> 00:14:48.335 A person of particular faith. 216 00:14:48.395 --> 00:14:49.136 Or maybe no faith. 217 00:14:49.316 --> 00:14:50.918 Maybe your coach on Saturday mornings. 218 00:14:51.519 --> 00:14:54.202 Someone who has been sick or cared for someone who was. 219 00:14:55.428 --> 00:15:02.333 All of those are true simultaneously, but they are not all switched on at the same time. 220 00:15:02.873 --> 00:15:11.819 At any given moment, one or two of them are doing the work of interpreting what's happening to you, and the rest are sitting in the background, waiting to be activated at the right moment. 221 00:15:12.439 --> 00:15:15.701 Psychologists call the switched on ones salient. 222 00:15:16.202 --> 00:15:20.885 And the finding that matters for us is that you don't choose which ones those are. 223 00:15:21.565 --> 00:15:22.426 Your surroundings do. 224 00:15:23.146 --> 00:15:24.668 Here's an example that Duhigg uses. 225 00:15:25.148 --> 00:15:27.691 Say you own a pro-Obama t-shirt. 226 00:15:28.132 --> 00:15:32.677 You wear it to a neighborhood barbecue where every single person voted Obama. 227 00:15:33.418 --> 00:15:34.639 It means almost nothing, right? 228 00:15:34.679 --> 00:15:38.904 Because nobody notices, nobody cares, and it certainly doesn't make you feel connected to anyone. 229 00:15:39.877 --> 00:15:47.346 Now, wear that same shirt to a National Rifle Association rally and run into the other person wearing it. 230 00:15:47.526 --> 00:15:50.930

Suddenly, that shirt is the most important fact about both of you. 231 00:15:51.090 --> 00:15:54.634 And you feel a bond with a total stranger on the strength of it. 232 00:15:55.535 --> 00:16:01.844 That shirt and its associated identity carries more weight because the surrounding context changed. 233 00:16:02.465 --> 00:16:07.351 So the identity that's alive in someone isn't exclusively their own property. 234 00:16:07.532 --> 00:16:11.757 It's also property of the situation, the one that they're standing in at that time. 235 00:16:12.398 --> 00:16:15.503 So let's bring this to life with another great example from the book. 236 00:16:15.723 --> 00:16:19.789 We're about to spend real time in a city in northern Iraq called Karakosh. 237 00:16:19.809 --> 00:16:25.778 And I'll give you the full story later momentarily, but I need to pull out one detail from it now. 238 00:16:26.539 --> 00:16:27.320 Karakosh was... 239 00:16:27.700 --> 00:16:30.783 for a long time, a mixed city, Christian and Muslim. 240 00:16:30.863 --> 00:16:33.025 And it was almost completely segregated. 241 00:16:33.045 --> 00:16:37.968 There were Christian restaurants and Muslim restaurants, Christian grocery stores and Muslim grocery stores. 242 00:16:38.509 --> 00:16:45.134 And at the door of these places were bouncers and the bouncers checked your ID and the ID card listed your religion. 243 00:16:45.855 --> 00:16:46.655 Think about what that does. 244 00:16:47.036 --> 00:16:55.642 You can't even go and buy a bag of rice in that city without your religious identity being read off a card and you used to decide whether you can even enter. 245 00:16:56.643 --> 00:17:03.165 Every threshold every day is a machine for making one identity louder than all the others you carry. 246 00:17:03.365 --> 00:17:05.805 Nobody in Karakosh sat down and decided to do that. 247 00:17:05.825 --> 00:17:06.666 It just accumulated. 248 00:17:06.686 --> 00:17:07.686 It became the default. 249 00:17:08.226 --> 00:17:14.748 The environment and its social norms are always setting which identity is live, which identity has been activated. 250 00:17:15.188 --> 00:17:20.029 In every room you walk into, in every channel you communicate through, something is already turning that dial. 251 00:17:20.789 --> 00:17:26.875 The venue does it, the format does it, the seating does it, who speaks first does it, the name on the letterhead does it. 252 00:17:27.396 --> 00:17:35.383 And if they can be switched on like this, it makes me wonder how often we actually choose the identity through which we're interpreting the world in any given moment. 253 00:17:35.924 --> 00:17:37.385 What does it look like if someone... 254 00:17:38.798 --> 00:17:41.899 trials tries to dial this deliberately, right? 255 00:17:42.379 --> 00:17:46.200 Well, we do have another story to turn to answer that. 256 00:17:46.520 --> 00:17:49.380 So this time we'll go back in time to 2021. 257 00:17:50.081 --> 00:17:58.082 COVID vaccines are being rolled out and a doctor, Dr. Reema Shami is sitting with a patient who is not going to take this vaccine. 258 00:17:58.282 --> 00:17:59.983 The patient is a retired police officer. 259 00:18:00.003 --> 00:18:04.944 He's got three grandchildren and his church is the most important place in his life. 260 00:18:05.784 --> 00:18:07.426 That's why I don't need the shot, he tells her. 261 00:18:07.526 --> 00:18:08.668 God will take care of me. 262 00:18:09.028 --> 00:18:09.829 I wash my hands. 263 00:18:09.909 --> 00:18:10.690 I wear the mask. 264 00:18:10.810 --> 00:18:11.611 God will provide. 265 00:18:11.651 --> 00:18:12.432 He knows my path. 266 00:18:13.490 --> 00:18:16.651 Now, Chimmy has all the material you'd expect a doctor to have. 267 00:18:16.751 --> 00:18:17.691 She has the safety data. 268 00:18:18.231 --> 00:18:19.252 She has efficacy numbers. 269 00:18:19.932 --> 00:18:22.773 She could mention that the Pope had told people to get vaccinated. 270 00:18:23.233 --> 00:18:29.095 And she knows sitting there that every one of those would land as an attack on the man's faith. 271 00:18:29.275 --> 00:18:32.376 And that the only thing they'd achieve is that he'd stop listening. 272 00:18:33.196 --> 00:18:34.756 So she doesn't mention COVID again. 273 00:18:34.957 --> 00:18:38.938 What she says is, it's wonderful that your faith gives you so much strength. 274 00:18:39.138 --> 00:18:42.379 You clearly have a really close relationship with God. 275 00:18:43.271 --> 00:18:48.575 And then, almost as an aside, she says, I imagine your grandchildren's health is probably very important to you. 276 00:18:49.755 --> 00:18:50.036 That's it. 277 00:18:50.216 --> 00:18:52.337 She didn't argue against the identity that was doing the resisting. 278 00:18:52.357 --> 00:18:53.818 In fact, she affirmed it. 279 00:18:54.559 --> 00:18:57.100 And then she put a second identity on the table. 280 00:18:57.861 --> 00:18:59.122 One the man already had. 281 00:18:59.342 --> 00:19:00.943 One he told her about himself. 282 00:19:02.182 --> 00:19:08.003 that pointed in a different direction on the same question, and she let him do the reconciling. 283 00:19:08.623 --> 00:19:15.245 This is consistent with what we learned about deep canvassing in episode two on how minds change. 284 00:19:15.265 --> 00:19:18.305 So if you haven't read that one, obviously if you haven't listened to that one, check it in. 285 00:19:18.345 --> 00:19:19.026 It's in your feed. 286 00:19:19.206 --> 00:19:20.826 So the appointment ends. 287 00:19:21.506 --> 00:19:26.527 30 minutes later, she notices he is still in the exam room, and she pulls the nurse aside and asks why. 288 00:19:27.187 --> 00:19:30.108 Well, turns out he wanted the vaccine, the nurse tells her. 289 00:19:30.727 --> 00:19:42.184 Now, one anecdote is one anecdote, and I'd be careful about how much weight we hang on a single exam room, even if this anecdote aligns with some of my personal experiences and having similar

conversations with people of faith during COVID. 290 00:19:42.224 --> 00:19:46.751 But there's a study underneath it that tells us why this approach works. 291 00:19:47.146 --> 00:19:50.727 Researchers took a group of women and gave them a mathematics test. 292 00:19:51.287 --> 00:19:56.908 Before the test, they primed them with the stereotype, the one that says women underperform at mathematics. 293 00:19:57.388 --> 00:20:01.589 Predictably, the primed women scored worse than the men. 294 00:20:01.769 --> 00:20:02.949 Then they ran two variations. 295 00:20:03.289 --> 00:20:11.171 One group was asked before the test to sketch out how they saw themselves, to write me in the middle of a page and draw a few lines from it. 296 00:20:11.711 --> 00:20:11.891 Right. 297 00:20:12.031 --> 00:20:13.131 So I'm a sister. 298 00:20:13.211 --> 00:20:13.932 I'm an employee. 299 00:20:13.952 --> 00:20:15.172 I'm a student, a volunteer. 300 00:20:16.083 --> 00:20:16.763 four or five things. 301 00:20:17.603 --> 00:20:23.885 The second group got the same instruction with one difference, write down as much as you possibly can. 302 00:20:24.566 --> 00:20:27.387 Every club, every hobby, every role, every part of your life. 303 00:20:28.087 --> 00:20:30.708 And those pages came back fully covered. 304 00:20:31.368 --> 00:20:36.509 Good friend, employee of the month, tutor, animal lover, chess club president, older sister, younger sister, on and on and on and on. 305 00:20:36.689 --> 00:20:38.030 Then everybody took the test. 306 00:20:38.790 --> 00:20:43.392 And the woman who'd drawn the short version scored just as badly as the woman who'd drawn nothing at all. 307 00:20:44.152 --> 00:20:44.912 Barely in effect. 308 00:20:45.433 --> 00:20:50.495 But the woman who'd drawn the long version scored the same as the men. 309 00:20:50.555 --> 00:20:52.675 The gap had closed completely. 310 00:20:53.536 --> 00:20:58.918 And I do find this genuinely helpful because the mechanism isn't reminding someone that they're complicated. 311 00:20:59.907 --> 00:21:04.328 Four roles on a page is a reminder and four roles does nothing. 312 00:21:04.809 --> 00:21:12.191 The mechanism is crowding, filling the room with so many live identities that no single one can dominate how you read what's happening to you. 313 00:21:12.891 --> 00:21:16.272 Which means the half measure of this, I think, would fail. 314 00:21:16.292 --> 00:21:21.614 And the half measure is probably what we're inclined to reach for as a default in our line of work. 315 00:21:22.902 --> 00:21:24.083 Okay, let's move on to the third mechanism. 316 00:21:24.703 --> 00:21:30.849 Everything so far has been one person working on one other person, at least the examples that we've walked through. 317 00:21:31.009 --> 00:21:37.154 Now, let's go into a room where nobody has any authority over anybody and a decision has to be unanimous. 318 00:21:37.434 --> 00:21:39.976 We're going to go to 1985 in Milwaukee. 319 00:21:40.577 -->

00:21:43.139 A man named Leroy Reed goes to trial. 320 00:21:43.739 --> 00:21:45.400 Reed had a prior felony conviction. 321 00:21:46.121 --> 00:21:49.065 He also has significant intellectual disabilities. 322 00:21:49.485 --> 00:21:55.572 He could barely read, and he'd seen an advertisement in a magazine for a mail order private detective course. 323 00:21:56.133 --> 00:22:02.381 And he sent away his $20, and back came an envelope with a tin badge and a set of instructions. 324 00:22:03.081 --> 00:22:04.563 Exercise regularly, it said. 325 00:22:05.144 --> 00:22:05.725 Buy a pistol. 326 00:22:06.756 --> 00:22:07.336 So he did it. 327 00:22:07.596 --> 00:22:08.597 He jogged in the mornings. 328 00:22:08.917 --> 00:22:15.000 And a week later, he took the bus to a sporting goods store and body 22, took it home, put it in his closet, in its box. 329 00:22:15.280 --> 00:22:17.901 As far as anyone could tell, he never touched it again. 330 00:22:19.321 --> 00:22:24.684 He got caught because he was loitering around the courthouse, hoping somebody would hire him to solve a crime. 331 00:22:24.964 --> 00:22:27.385 And a police officer asked him for identification. 332 00:22:27.565 --> 00:22:31.407 And the only thing in his pocket with his name on it was the receipt from the gun shop. 333 00:22:32.500 --> 00:22:35.302 Under Wisconsin law, this was open and shut. 334 00:22:35.823 --> 00:22:37.124 Felon in possession. 335 00:22:37.704 --> 00:22:40.186 So the judge told the jury they had three questions to answer. 336 00:22:40.687 --> 00:22:41.567 Was Reed a felon? 337 00:22:42.248 --> 00:22:43.369 Had he acquired a firearm? 338 00:22:43.809 --> 00:22:44.970 Did he know he'd acquired it? 339 00:22:45.531 --> 00:22:47.712 Yes, yes, and yes are the facts, of course. 340 00:22:48.313 --> 00:22:51.476 The judge also told them not to be swayed by sympathy. 341 00:22:52.433 --> 00:22:57.036 And the research says that juries tend to convict. 342 00:22:57.676 --> 00:23:01.578 Juries with a defendant who has a record almost always get there in the end. 343 00:23:01.658 --> 00:23:02.399 Guilty as charged. 344 00:23:02.419 --> 00:23:06.001 And it only takes one immovable person to hang the whole thing. 345 00:23:07.161 --> 00:23:11.444 And yet, this jury came back not guilty in two and a half hours. 346 00:23:12.602 --> 00:23:16.465 The story on how we get there is fascinating. 347 00:23:16.545 --> 00:23:18.107 So this is how it occurs, right? 348 00:23:18.147 --> 00:23:28.396 So the newly chosen foreman says he knows some juries like to vote right away, but maybe they could avoid committing to positions and instead go around the room and each give their general impressions. 349 00:23:29.076 --> 00:23:30.017 He didn't argue a position. 350 00:23:30.057 --> 00:23:31.658 He just set a rule about how they talk. 351 00:23:33.260 --> 00:23:40.006 And the reason it matters is that once somebody commits to a verdict from the start, it becomes very expensive for them to say anything else. 352 00:23:41.118 --> 00:23:41.938 So it only half works. 353 00:23:42.318 --> 00:23:52.682 Carl, a firefighter, states straight away that Reed is guilty and that the extenuating circumstances belong to the judge to deal with that sentencing, not for them, the jury. 354 00:23:52.882 --> 00:23:54.203 Two others agree with him. 355 00:23:54.403 --> 00:23:58.184 A teacher named Lorraine says he's technically guilty, but wants his reading difficulty weighed. 356 00:23:59.024 --> 00:24:02.606 A juror named Henry says technically the man is guilty, guilty as sin. 357 00:24:03.226 --> 00:24:06.127 But I want to quit Leroy because I don't think he was fully aware of the rules. 358 00:24:07.780 --> 00:24:16.142 After the round, when the four man tallies the votes, there are three for conviction, two leaning towards acquittal and seven genuinely undecided. 359 00:24:16.342 --> 00:24:19.423 And then there's this juror named John Boley. 360 00:24:19.443 --> 00:24:21.344 He's a professor of contemporary literature. 361 00:24:22.764 --> 00:24:23.965 He's wearing a suit. 362 00:24:24.265 --> 00:24:25.905 A lot of others are in jeans. 363 00:24:25.945 --> 00:24:26.886 So he does stand out. 364 00:24:27.386 --> 00:24:33.608 And one of them even said afterwards that when Boley started talking about Kafka and trials, his reaction was, what are you talking about, man? 365 00:24:33.728 --> 00:24:34.748 What planet are you from? 366 00:24:35.428 --> 00:24:38.529 Anyway, the person who's about to move this room, though, is this guy. 367 00:24:39.430 --> 00:24:45.732 He's the person who the room finds most alien and most remembering after the fact. 368 00:24:45.772 --> 00:24:49.313 And his opening move is to decline to have an opinion. 369 00:24:50.894 --> 00:24:54.035 I'm not really at all sure what I feel on this case, he says. 370 00:24:54.095 --> 00:24:57.917 I want to listen to other people and figure this out together as we go. 371 00:24:58.905 --> 00:25:00.086 And then he starts asking questions. 372 00:25:00.426 --> 00:25:01.667 They sound like small talk at first. 373 00:25:01.948 --> 00:25:03.049 What do you think of the handguns? 374 00:25:03.149 --> 00:25:03.969 Do you own a handgun? 375 00:25:04.049 --> 00:25:05.871 Can we talk about what possession actually means? 376 00:25:06.291 --> 00:25:07.092 What is justice? 377 00:25:07.733 --> 00:25:08.833 But he's not making conversation. 378 00:25:08.854 --> 00:25:15.039 He's building a list in his head of what each person in that room actually wants out of the next two hours. 379 00:25:15.099 --> 00:25:16.620 And by the end of it, he has one. 380 00:25:16.680 --> 00:25:18.682 He has this ledger in his own head. 381 00:25:19.162 --> 00:25:20.183 Henry wants guidance. 382 00:25:20.523 --> 00:25:21.744 Barbara wants compassion. 383 00:25:21.865 --> 00:25:23.186 Carl wants to go buy the book. 384 00:25:24.083 --> 00:25:27.747 Which is what Duhigg argues is the first part of a good conversation. 385

00:25:28.548 --> 00:25:32.952 Figuring out what everybody needs from the discussion before you try to give it to them. 386 00:25:34.033 --> 00:25:36.956 Meanwhile, the room is sliding towards conviction. 387 00:25:37.137 --> 00:25:40.560 The foreman declares that he wants to vote guilty. 388 00:25:40.660 --> 00:25:41.842 A fence sitter joins him. 389 00:25:41.942 --> 00:25:43.023 Carl reinforces it. 390 00:25:43.583 --> 00:25:46.366 Reid didn't hurt anyone this time, but what about next time? 391 00:25:47.360 --> 00:25:50.161 What if he bought that gun and killed some innocent bystander? 392 00:25:50.601 --> 00:25:50.801 Right. 393 00:25:50.861 --> 00:25:52.101 Voices are starting to rise. 394 00:25:52.822 --> 00:25:55.622 And Bolit changes his tactic a little bit. 395 00:25:55.642 --> 00:26:00.104 He's finished asking questions and now he's going to change how the room decides. 396 00:26:00.644 --> 00:26:07.806 He says in a deliberately light tone as the conversation is starting to heat up that he'd noticed something about the gun. 397 00:26:08.486 --> 00:26:10.367 If you look at it closely, he says it looks like a toy. 398 00:26:11.299 --> 00:26:13.040 And then he makes himself the idiot of his own story. 399 00:26:13.120 --> 00:26:17.362 I'd bet if I bought a gun and it came with a holster, the first thing I'd do is stick it right here. 400 00:26:17.582 --> 00:26:19.983 And he gestures to his belt and walk around Milwaukee. 401 00:26:20.003 --> 00:26:24.506 And every time I go past an underpass, I wouldn't have to worry about what steps out from behind the lamppost. 402 00:26:24.726 --> 00:26:25.646 I'm 10 feet tall. 403 00:26:25.706 --> 00:26:26.667 I'm packing a rod. 404 00:26:27.634 --> 00:26:31.116 The rest of the jury has no idea what he's talking about, but hang tight. 405 00:26:31.736 --> 00:26:36.098 Reed, he says, handled that gun almost like it was a sacramental thing. 406 00:26:36.639 --> 00:26:37.239 He locked it up. 407 00:26:37.459 --> 00:26:38.380 He put it in the closet. 408 00:26:38.500 --> 00:26:39.200 He shut the door. 409 00:26:39.220 --> 00:26:40.601 He didn't put it in a holster. 410 00:26:40.621 --> 00:26:42.222 He didn't wear it on his hip. 411 00:26:42.422 --> 00:26:45.923 And another juror, someone who had been drifting towards guilty, picks it straight up. 412 00:26:46.644 --> 00:26:47.804 He didn't take it out of the box. 413 00:26:49.025 --> 00:26:52.627 And another one goes further, says we can't even say he even knew how to use a gun. 414 00:26:53.807 --> 00:26:59.151 So Duhigg observes that there's no evidence at the trial about whether Leroy Reed knew how to fire a weapon. 415 00:26:59.631 --> 00:27:03.694 The jury invented a man who might not understand that a gun needs bullets. 416 00:27:03.915 --> 00:27:07.037 And they did this in about 90 seconds. 417 00:27:07.677 --> 00:27:09.259 But look at the move Bully made. 418 00:27:09.279 --> 00:27:10.780 He didn't argue for innocence. 419 00:27:10.980 --> 00:27:16.264 He puts his own behavior beside Reed's behavior and let the contrast do the work. 420 00:27:16.644 --> 00:27:23.249 And because he cast himself as a fool in comparison, nobody in that room had to lose an argument in order to accept it. 421 00:27:24.449 --> 00:27:26.751 They take a formal vote after about an hour. 422 00:27:27.071 --> 00:27:29.793 We're now at nine voting to acquit, three to convict. 423 00:27:30.253 --> 00:27:30.994 So things are changing. 424 00:27:31.880 --> 00:27:33.341 The foreman flips, right? 425 00:27:33.521 --> 00:27:36.964 While he had voted guilty, he admits that he was listening to the others. 426 00:27:37.024 --> 00:27:45.150 He found himself thinking about a time he got pulled over for speeding and told the officer it wasn't right to ticket him because he hadn't put anybody at risk. 427 00:27:46.011 --> 00:27:50.674 And sitting in that jury room, it occurred to him that Leroy Reed was in the same position. 428 00:27:50.855 --> 00:27:55.758 He says, I can see a reason for somewhat of a doubt, however minor it might be. 429 00:27:56.699 --> 00:27:57.480 That's not evidence. 430 00:27:57.580 --> 00:28:01.702 That's a man asking himself, what somebody like me does in a situation like this? 431 00:28:02.763 --> 00:28:05.664 Duhigg's name for this is the logic of similarities. 432 00:28:06.145 --> 00:28:09.507 And it's one of the two ways people make joint decisions. 433 00:28:10.107 --> 00:28:19.833 The other is just a plain arithmetic of costs and benefits, which matters because now there's one holdout and the logic of similarities is not going to touch him. 434 00:28:19.993 --> 00:28:21.994 We're talking about Carl, of course, right? 435 00:28:22.054 --> 00:28:23.114 Carl is a firefighter. 436 00:28:23.154 --> 00:28:23.595 He's immune. 437 00:28:24.713 --> 00:28:28.437 We're going way too deep into his psychological thoughts, he says. 438 00:28:28.477 --> 00:28:29.918 We're guessing at what he was thinking. 439 00:28:29.938 --> 00:28:30.939 We're guessing at what he knew. 440 00:28:30.979 --> 00:28:33.142 We're guessing at what he didn't know. 441 00:28:34.287 --> 00:28:36.228 Now, here's what makes Carl interesting. 442 00:28:36.348 --> 00:28:42.451 Every other person in that room has been telling stories about themselves all afternoon, making themselves vulnerable, asides, memories, little revelations. 443 00:28:42.991 --> 00:28:45.152 Carl has told none, not one. 444 00:28:45.612 --> 00:28:57.918 His son later described him as an ideal fireman, a follow the checklist, respected chain of command kind of man who had trained himself to reason through costs and benefits because in an emergency, that is the thinking that keeps people alive. 445 00:28:58.779 --> 00:29:02.320 So a juror asked him a question, not an argument, a question. 446 00:29:02.740 --> 00:29:08.322 It appears to me that your decision that this man is guilty is very important and complete in your mind. 447 00:29:08.982 --> 00:29:10.683

Share more of that with us, if you will. 448 00:29:12.283 --> 00:29:13.584 Carl deflects at first. 449 00:29:13.864 --> 00:29:14.224 I can't. 450 00:29:14.244 --> 00:29:19.065 He says, I don't have the education and the training to put myself in your class as far as understanding the human mind. 451 00:29:20.006 --> 00:29:24.107 Then somebody asks whether he thinks there's ever a case where an exception could be made. 452 00:29:25.115 --> 00:29:27.757 And Carl finally says what he's actually been protecting this whole time. 453 00:29:28.378 --> 00:29:29.579 Reed's no threat, he says. 454 00:29:29.679 --> 00:29:31.900 I don't think he had any ill intent. 455 00:29:32.000 --> 00:29:33.642 I don't feel he's a threat to society. 456 00:29:34.062 --> 00:29:36.984 But if juries stop enforcing laws, that's anarchy. 457 00:29:37.385 --> 00:29:40.827 He'd made an exception if it bought something for public safety. 458 00:29:40.847 --> 00:29:42.649 He just can't see what it buys here. 459 00:29:42.669 --> 00:29:43.649 Right? 460 00:29:43.690 --> 00:29:44.110 There it is. 461 00:29:44.410 --> 00:29:48.013 90 minutes of argument and nobody had asked him what he cared about until then. 462 00:29:48.393 --> 00:29:50.155 What he cared about was public safety. 463 00:29:50.935 --> 00:29:54.082 And Bolli doesn't try to make Carl into a different kind of thinker. 464 00:29:54.182 --> 00:29:55.504 He switches languages. 465 00:29:56.065 --> 00:30:01.376 He starts by conceding, I think this is a good law, and I don't want to say anything that suggests I don't take it seriously. 466 00:30:02.355 --> 00:30:05.196 But then you go at it with a pure cost benefit arithmetic. 467 00:30:05.816 --> 00:30:06.537 It's finals weeks. 468 00:30:07.217 --> 00:30:09.738 He's got a mountain of work and his students have been mugged. 469 00:30:09.778 --> 00:30:11.339 They've been beaten and raped. 470 00:30:11.599 --> 00:30:16.481 One of them a week ago, he says he wants to do his civic duty and instead he's spending it here. 471 00:30:16.781 --> 00:30:24.824 In spite of this room and these very serious people and their lovely dog and pony show and legal rigmarole, he says, I'm sitting here thinking this is Mickey Mouse. 472 00:30:24.904 --> 00:30:26.765 They could be putting rapists and murders away. 473 00:30:26.805 --> 00:30:28.406 He says, you give me Leroy. 474 00:30:29.405 --> 00:30:30.385 And then the actual argument. 475 00:30:31.106 --> 00:30:44.110 If we acquit, we send the district attorney's offices a message about where to spend their attention, which means acquitting Leeway Reid serves public safety, which is obviously the kind of thing Carl said he wanted. 476 00:30:45.290 --> 00:30:47.651 Carl says definitely he shouldn't be here. 477 00:30:49.213 --> 00:30:50.715 He's still not all the way there, though. 478 00:30:50.755 --> 00:30:53.198 So Bolli makes one last move, and look at what he does. 479 00:30:53.599 --> 00:31:02.932 He says, I have an enormous amount of respect for your sense of the importance of law, your sense of the importance of getting it right, your dedication to the integrity of the judicial process. 480 00:31:04.786 --> 00:31:06.728 Two and a half hours in, they vote. 481 00:31:07.288 --> 00:31:10.371 And of course, as I said earlier, not guilty, all 12. 482 00:31:10.891 --> 00:31:11.452 So what happened? 483 00:31:11.492 --> 00:31:15.735 Well, changing your mind has a cost and it's paid by our ego. 484 00:31:15.795 --> 00:31:19.198 It also has a benefit, the self-respect of having done the right thing. 485 00:31:19.879 --> 00:31:23.281 Bully made the second one bigger than the first for Carl. 486 00:31:24.347 --> 00:31:25.508 And you know what it takes to do that? 487 00:31:26.450 --> 00:31:26.810 Empathy. 488 00:31:27.531 --> 00:31:27.771 All right. 489 00:31:27.791 --> 00:31:34.720 Now, back to Northern Iraq, because this is the case that changed how I'm thinking about all of this. 490 00:31:34.780 --> 00:31:36.021 So this is our fourth mechanism. 491 00:31:36.282 --> 00:31:38.985 It's what happens when nobody has a conversation at all. 492 00:31:39.045 --> 00:31:40.006 So Karakosh. 493 00:31:40.938 --> 00:31:45.883 remember is the segregated city I described earlier, the one with the bouncers reading religion off ID cars. 494 00:31:45.923 --> 00:31:50.548 And by 2018, it had been through something far worse than segregation. 495 00:31:50.728 --> 00:31:52.370 ISIS had taken the city. 496 00:31:52.470 --> 00:31:54.892 Hundreds of the Christian population were killed. 497 00:31:55.493 --> 00:31:57.575 Around 50,000 were driven out of their homes. 498 00:31:57.775 --> 00:31:58.956 Churches were ransacked. 499 00:31:59.077 --> 00:32:00.398 Christian businesses burned. 500 00:32:00.518 --> 00:32:01.659 Christian women assaulted. 501 00:32:02.770 --> 00:32:06.333 ISIS pulled out in 2016 and the refugees started coming home. 502 00:32:06.493 --> 00:32:13.640 And when they pulled those returning Christians, a large majority said they believed their Muslim neighbors had betrayed them. 503 00:32:14.821 --> 00:32:15.622 This is the setting. 504 00:32:15.642 --> 00:32:16.322 It's pretty grim. 505 00:32:17.303 --> 00:32:22.408 Then in the spring of 2018, flyers went up in the Christian neighborhoods advertising a new soccer league. 506 00:32:22.948 --> 00:32:25.831 The meeting was held in a church that was still half burnt out. 507 00:32:26.883 --> 00:32:27.484 The offer was good. 508 00:32:28.104 --> 00:32:36.893 A free tournament, professional referees at every match, new nets, new balls, trophies for the winners, and a jersey for every player with his own name printed on the back. 509 00:32:38.108 --> 00:32:39.810 But there was one condition. 510 00:32:40.450 --> 00:32:42.452 Only teams that already existed could enter. 511 00:32:42.552 --> 00:32:44.934 So these were real teams with real histories. 512 00:32:44.974 --> 00:32:47.416 They're not groups assembled by some researcher. 513

00:32:48.157 --> 00:32:50.980 Local teams normally ran nine players. 514 00:32:51.200 --> 00:32:54.303 And this league required now 12 players, three extra players. 515 00:32:55.023 --> 00:32:59.427 And half the teams could pick their own three extra players who would all be Christian. 516 00:32:59.668 --> 00:33:04.031 The other half would have three Muslim players assigned to them by the organizers. 517 00:33:05.043 --> 00:33:12.647 This was designed by a woman named Salma Musa, then a PhD student at Stanford, testing an old idea called the contact hypothesis. 518 00:33:13.248 --> 00:33:20.712 That if you bring people with clashing identities together under the right conditions, you can wear down, you can break down old hatreds. 519 00:33:21.512 --> 00:33:23.193 At the information meeting, a lot of people walked out. 520 00:33:23.533 --> 00:33:25.614 They told us this would ruin their teams, Musa said. 521 00:33:25.855 --> 00:33:28.416 They were going to cause another war. 522 00:33:28.976 --> 00:33:31.578 But the referees were professional and the trophies were large. 523 00:33:31.678 --> 00:33:33.019 A few teams signed up anyway. 524 00:33:33.599 --> 00:33:36.401 Then the jerseys arrived and suddenly everybody wanted in. 525 00:33:36.721 --> 00:33:40.083 So 42 teams ultimately participated in this. 526 00:33:41.243 --> 00:33:44.685 Early practices were exactly as bad as you'd expect. 527 00:33:44.845 --> 00:33:48.348 Christian players refused to introduce themselves to their new Muslim teammates. 528 00:33:48.708 --> 00:33:51.810 They sat as far apart on the sidelines as the bench allowed. 529 00:33:51.930 --> 00:33:53.411 The Muslim players tried to fit in. 530 00:33:53.431 --> 00:33:55.932 The Christian players were openly hostile. 531 00:33:56.513 --> 00:34:02.697 And Musa had put in place one rule that would do some very serious heavy lifting throughout the project. 532 00:34:03.837 --> 00:34:06.499 Every player had to get equal playing time. 533 00:34:07.379 --> 00:34:07.680 That's it. 534 00:34:08.682 --> 00:34:11.162 Cooperation wasn't encouraged or facilitated or workshopped. 535 00:34:11.563 --> 00:34:13.643 It was made structurally unavoidable. 536 00:34:14.143 --> 00:34:21.845 If you want to win and the man you refuse to speak to is going to be on the pitch for a third of the match, regardless, you're going to have to work something out. 537 00:34:22.045 --> 00:34:25.765 And then the players started solving problems that nobody had asked them to solve. 538 00:34:25.905 --> 00:34:31.566 So some teams, you know, had been playing in Syriac, the language Christians in that region speak and which the Muslim players didn't understand. 539 00:34:32.207 --> 00:34:34.047 It was wrecking their communications on the pitch. 540 00:34:34.707 --> 00:34:37.067 So two coaches made a rule on their teams. 541 00:34:37.648 --> 00:34:38.588 Everyone must speak Arabic. 542 00:34:39.048 --> 00:34:40.248 And those two teams started winning. 543 00:34:40.728 --> 00:34:41.948 And the other coaches copied them. 544 00:34:43.008 --> 00:34:49.230 A week or so later, a group of Christian players complained that their Muslim teammates kept turning up late and eating into practice time. 545 00:34:49.630 --> 00:34:55.571 The Muslims explained that they were coming from the other side of the city through multiple checkpoints on slow buses. 546 00:34:56.131 --> 00:35:00.311 So the Christian players put money together to pay for taxes to get them there faster. 547 00:35:00.531 --> 00:35:01.752 Nobody designed either of those. 548 00:35:01.972 --> 00:35:03.072 There are consequences. 549 00:35:04.012 --> 00:35:07.433 Within months, Musa was struggling to tell the two groups apart. 550 00:35:07.753 --> 00:35:08.553 They sat together. 551 00:35:08.793 --> 00:35:10.053 They celebrated goals together. 552 00:35:10.453 --> 00:35:12.634 One team elected a Muslim captain. 553 00:35:13.214 --> 00:35:18.795 And the all-Christian team started complaining that they were at a competitive disadvantage for not having any Muslims. 554 00:35:18.995 --> 00:35:21.295 But look at the ensuing research data. 555 00:35:21.355 --> 00:35:31.517 So compared to players on all-Christian teams, players on the mixed teams were 13 points more likely to say they wouldn't mind a mixed team next season. 556 00:35:32.337 --> 00:35:38.881 26 points more likely to vote a Muslim player from a different team a sportsmanship prize. 557 00:35:39.442 --> 00:35:44.965 And 49 points more likely to still be training with Muslims six months after the whole thing had ended. 558 00:35:45.866 --> 00:35:49.288 This is six months later, after the trophies were handed out and the league was over. 559 00:35:50.311 --> 00:35:51.352 It's not all roses, though. 560 00:35:51.732 --> 00:35:52.934 Prejudice did not evaporate. 561 00:35:52.994 --> 00:35:58.559 Christian players admitted they still felt uncertain about Muslims generally, the ones who weren't their teammates. 562 00:35:58.739 --> 00:36:03.804 The effect attached to the people they played beside, and it generalized only partway. 563 00:36:04.365 --> 00:36:05.766 But Musa is clear about why it works. 564 00:36:05.786 --> 00:36:09.890 She says, maybe some people will say, well, that's because sports breaks down barriers. 565 00:36:09.950 --> 00:36:11.191 But it's not just that. 566 00:36:11.331 --> 00:36:13.393 It's how we structured everything that made the difference. 567 00:36:14.266 --> 00:36:15.486 She names three decisions. 568 00:36:16.107 --> 00:36:17.387 The first was rules. 569 00:36:17.787 --> 00:36:21.248 Teams naturally handed people jobs that had nothing to do with religion. 570 00:36:21.809 --> 00:36:23.329 This one's a goalkeeper. 571 00:36:23.669 --> 00:36:25.170 This one leads the stretches at halftime. 572 00:36:25.190 --> 00:36:26.430 This one brings the sports drinks. 573 00:36:26.850 --> 00:36:28.811 This one's captain and gives a speech before kickoff. 574 00:36:29.471 -->

00:36:34.853 And as Musa puts it, those identities became more important than religion because they were connected to winning. 575 00:36:35.613 --> 00:36:39.034 That's like the mathematics test that we saw earlier. 576 00:36:39.515 --> 00:36:42.796 Crowd the room with identities and no single one can dominate. 577 00:36:43.716 --> 00:36:45.377 The second was equality. 578 00:36:46.018 --> 00:36:47.399 Karakosh had a hierarchy. 579 00:36:47.439 --> 00:36:50.361 Christians had historically been wealthier, better educated. 580 00:36:51.101 --> 00:36:55.144 And as the refugees returned, the old order was reasserting itself. 581 00:36:55.944 --> 00:36:59.907 Equal playing time deleted that hierarchy inside the touchlines. 582 00:36:59.927 --> 00:37:05.751 And on the field, Musa says, because everyone had to play the same amount, all the players were the same. 583 00:37:06.292 --> 00:37:07.673 There weren't power differentials. 584 00:37:08.651 --> 00:37:13.595 And the third was that the new group was built out of an identity people already had. 585 00:37:13.816 --> 00:37:16.818 Nobody was asked to be less Christian or less Muslim. 586 00:37:17.159 --> 00:37:21.503 They were asked to be teammates, which they already were by the second week. 587 00:37:22.671 --> 00:37:31.276 And now I want to put this next to something from earlier in the book, because Duhigg puts these two stories sort of three, several chapters apart and never really sets them side by side. 588 00:37:31.796 --> 00:37:34.637 And I think the comparison is potentially the most useful thing here. 589 00:37:35.398 --> 00:37:38.980 In an earlier chapter, there's an experiment where gun control activists. 590 00:37:39.907 --> 00:37:49.594 And gun rights advocates were flown to Washington, trained in specific communication techniques over a weekend, and then moved on to a private Facebook group to keep talking. 591 00:37:49.774 --> 00:37:55.318 We're talking about 150 people, 15,000 posts in four weeks. 592 00:37:55.718 --> 00:37:59.621 Most of the group hadn't been at the training and hadn't met anybody in person. 593 00:38:00.281 --> 00:38:01.162 And can you guess what happened? 594 00:38:01.202 --> 00:38:02.643 Well, the whole thing completely fell apart. 595 00:38:03.084 --> 00:38:05.085 People called each other idiots, Nazis and fascists. 596 00:38:05.406 --> 00:38:13.153 The moderators had been trained as models of curiosity and careful listening, and they tried to teach the techniques directly and online. 597 00:38:13.333 --> 00:38:15.214 But those approaches fell short. 598 00:38:15.795 --> 00:38:25.844 So we have 150 motivated volunteers, professionally facilitated, explicitly taught how to listen, and they struggled to apply this at scale with 150 people online. 599 00:38:27.555 --> 00:38:35.370 Compare that to the 42 teams in Karakosh, in a city recovering from an attempted extermination, given no communication training at all. 600 00:38:36.132 --> 00:38:37.454 And it held for six

months. 601 00:38:38.558 --> 00:38:41.340 The difference, I think, isn't the people. 602 00:38:41.620 --> 00:38:43.421 It's not the technique necessarily. 603 00:38:43.501 --> 00:38:45.362 In Washington, they taught conversation. 604 00:38:45.862 --> 00:38:52.245 In Karakos, they just built conditions in which conversation was the only way to get something everybody wanted. 605 00:38:52.786 --> 00:38:56.408 My takeaway on this is that you don't scale the conversation. 606 00:38:56.808 --> 00:39:00.430 What you scale are the conditions that produce it. 607 00:39:01.987 --> 00:39:03.288 Okay, that was a lot of mechanisms. 608 00:39:03.868 --> 00:39:05.450 Let's move on to mental models. 609 00:39:06.070 --> 00:39:08.592 This time I have four models to explore with you. 610 00:39:08.632 --> 00:39:11.654 The first one is the salience dial. 611 00:39:13.076 --> 00:39:22.283 In every room, on every channel, something is deciding which of a person's many identities is the one doing the listening, is the one processing the information. 612 00:39:23.418 --> 00:39:24.558 It is never neutral. 613 00:39:25.099 --> 00:39:26.239 It is never switched off. 614 00:39:26.539 --> 00:39:28.940 So in Karakosh, it was a bouncer with an ID card. 615 00:39:29.520 --> 00:39:33.762 In Moose's League, it was a jersey with your name on the back and a rule about playing time. 616 00:39:34.762 --> 00:39:41.105 In a consultation session, it's the venue, the seating, the agenda, who chairs, whose logo is on the wall. 617 00:39:41.805 --> 00:39:44.106 The dial is always set to something. 618 00:39:44.866 --> 00:39:47.487 Most of the time, nobody chooses that setting. 619 00:39:47.527 --> 00:39:52.089 And the setting is doing, I think, a lot more work than the message itself. 620 00:39:53.009 --> 00:39:54.270 Okay, so that's the salient style. 621 00:39:54.890 --> 00:39:57.751 The other mental model is the near universal identity. 622 00:39:58.791 --> 00:40:03.973 Every audience contains an identity that almost all of them hold and that nobody is currently speaking to. 623 00:40:04.153 --> 00:40:12.637 The instinct in our business when we learn that identity shapes reception is to slice even finer, more segments, more personas, more tailored variants. 624 00:40:12.657 --> 00:40:19.800 But people carry dozens of identities at once and picking the one you think is live could easily be a bad guess. 625 00:40:20.803 --> 00:40:25.916 The identities that do the most work tend to be the broad ones sitting in plain view. 626 00:40:27.115 --> 00:40:31.616 parent, neighbor, patient, someone who has waited for something. 627 00:40:32.116 --> 00:40:34.597 And Shami didn't find an exotic segment. 628 00:40:34.617 --> 00:40:40.578 You'll recall Shami is our doctor during COVID who was able to relate with her patient on the question of faith. 629 00:40:41.198 --> 00:40:46.860 What she found was a grandfather, which was sitting right there in the first two minutes of the conversation. 630 00:40:47.040 --> 00:40:54.781 The near universal identity is usually available, usually unoccupied and usually ignored because it doesn't feel clever enough. 631 00:40:54.841 --> 00:40:56.462 I think sometimes we overthink these things. 632 00:40:57.202 --> 00:40:59.663 Okay, that is the near universal identity. 633 00:41:00.324 --> 00:41:03.445 The next mental model is the silent ledger. 634 00:41:04.266 --> 00:41:10.569 In any conversation with more than two people in it, everybody wants something from that discussion that they have not said out loud. 635 00:41:12.010 --> 00:41:12.910 Some people wanted guidance. 636 00:41:12.930 --> 00:41:14.271 When we go back to that jury, right? 637 00:41:14.291 --> 00:41:15.111 Henry wanted guidance. 638 00:41:15.191 --> 00:41:16.212 Barbara wanted compassion. 639 00:41:16.332 --> 00:41:17.633 Carl wanted to go buy the book. 640 00:41:18.093 --> 00:41:19.534 None of them really announced it. 641 00:41:19.974 --> 00:41:23.155 All of them acted on it though for two and a half hours. 642 00:41:23.195 --> 00:41:25.837 And it took a guy like Boley to extract that out of them. 643 00:41:26.903 --> 00:41:30.005 The ledger exists whether or not anyone's actually keeping it. 644 00:41:30.386 --> 00:41:32.888 Boldy's advantage in that jury, the deliberation... 645 00:41:34.425 --> 00:41:36.827 over the others was not necessarily intelligence. 646 00:41:37.067 --> 00:41:38.868 And it certainly wasn't status, right? 647 00:41:38.888 --> 00:41:40.149 Because he was seen as the odd one. 648 00:41:40.289 --> 00:41:41.810 He was seen as an alien in that room. 649 00:41:42.230 --> 00:41:46.773 It was that he wrote the ledger down in his head while everyone else was arguing the case. 650 00:41:46.813 --> 00:41:48.314 And that gave him something to work from. 651 00:41:48.614 --> 00:41:56.379 And it gave him the connective glue or the tissue that allowed him to stitch together these disparate individuals around one shared view. 652 00:41:56.399 --> 00:41:58.600 Okay, that's the silent ledger. 653 00:41:58.621 --> 00:42:02.123 The fourth mental model I have here is the directive penalty. 654 00:42:03.275 --> 00:42:11.062 So Duhigg talks about a researcher in the book that had groups watch film clips separately, scan their brains, had them discuss the clips and scan them again. 655 00:42:11.082 --> 00:42:14.305 And the idea was to see how is that shared moment? 656 00:42:14.626 --> 00:42:16.267 Where is their alignment in that shared moment? 657 00:42:16.648 --> 00:42:21.832 And are the conversations around that illustrative of where people get that alignment? 658 00:42:22.653 --> 00:42:25.237 Talking pulled their brains into alignment. 659 00:42:25.458 --> 00:42:28.222 And that was expected and expected finding, right? 660 00:42:28.242 --> 00:42:30.446 You talk about things together, you are going to get some type of alignment. 661 00:42:30.846 --> 00:42:33.591 What wasn't expected was which groups aligned most. 662

00:42:34.352 --> 00:42:37.136 His hypothesis was that a strong leader would help. 663 00:42:37.963 --> 00:42:39.444 The data, though, said the reverse. 664 00:42:39.624 --> 00:42:44.706 Groups with a dominant leader showed the least neural synchrony of any in a study. 665 00:42:45.186 --> 00:42:55.009 His example is a participant who was talkative and direct, who assigned roles and kept everybody on task, and whose team ended up thinking less alike than any other. 666 00:42:56.210 --> 00:43:06.514 Being the person who runs the room and being the person who makes the room work are two different jobs, and I think they may interfere with each other. 667 00:43:07.528 --> 00:43:09.409 Big, big, big mental model for me. 668 00:43:09.829 --> 00:43:12.550 So the directive penalty, I'm thinking about how to apply this in my day to day. 669 00:43:13.331 --> 00:43:13.871 OK, good. 670 00:43:14.231 --> 00:43:15.471 Let's move on to misunderstandings. 671 00:43:15.732 --> 00:43:20.254 There are four ways people could walk away from this book with something it really doesn't say. 672 00:43:20.474 --> 00:43:23.735 The first is that matching means mirroring. 673 00:43:23.755 --> 00:43:24.715 And this one's big. 674 00:43:25.056 --> 00:43:29.558 And Duhigg heads it off twice, at least twice, maybe more in two different chapters. 675 00:43:30.598 --> 00:43:37.022 The matching principle says that effective communication means recognizing what kind of conversation is happening and then aligning with it. 676 00:43:37.783 --> 00:43:46.648 And the moment people hear that, a lot of them reach for a shortcut of like body language and mirrored posture and repeating back the last three words somebody said to you. 677 00:43:46.848 --> 00:43:47.949 That's not quite it. 678 00:43:48.329 --> 00:43:50.111 And the book and Duhigg is pretty blunt about it. 679 00:43:51.171 --> 00:43:53.873 Simply mirroring somebody doesn't prove anything. 680 00:43:54.233 --> 00:43:56.595 If you laugh loudly and I just merely smile. 681 00:43:57.235 --> 00:43:58.576 It doesn't feel like I want to bond with you. 682 00:43:58.877 --> 00:44:00.318 It feels like I'm humoring you, right? 683 00:44:00.879 --> 00:44:03.501 So you go back to Bolli and Carl in that deliberation. 684 00:44:04.122 --> 00:44:05.463 Bolli did not mirror Carl. 685 00:44:05.603 --> 00:44:08.486 He didn't adopt his manner or his vocabulary or his posture. 686 00:44:08.586 --> 00:44:12.350 What he adopted was Carl's reasoning, cost against benefits. 687 00:44:12.890 --> 00:44:14.872 And then he expanded the options available inside. 688 00:44:16.027 --> 00:44:22.274 That's a much harder thing to do than mirroring, and it requires you to have worked out how the other person decides things. 689 00:44:23.075 --> 00:44:26.219 The related error is thinking that matching means giving in. 690 00:44:26.900 --> 00:44:27.260 It doesn't. 691 00:44:27.280 --> 00:44:31.705 Duhigg is explicit that surrendering to somebody else's preoccupations

doesn't work either. 692 00:44:32.266 --> 00:44:35.430 Bolli never conceded that Reid was guilty. 693 00:44:36.868 --> 00:44:41.790 Okay, the other miscommunication is that we should always match the other person. 694 00:44:42.631 --> 00:44:48.153 In the practical guides at the back half of the book, Duhigg closes each section with a short how-to chapter. 695 00:44:48.173 --> 00:44:51.114 And there's a line that quietly qualifies the book's own title. 696 00:44:51.294 --> 00:44:52.135 Sometimes he writes... 697 00:44:53.102 --> 00:44:54.623 we might not want to match emotions. 698 00:44:55.163 --> 00:44:58.566 If someone is angry and we become angry, it may drive us apart. 699 00:44:59.126 --> 00:45:02.128 But if we acknowledge their mood and energy, you seem upset. 700 00:45:02.268 --> 00:45:02.828 What's wrong? 701 00:45:03.489 --> 00:45:04.449 We can start to align. 702 00:45:04.509 --> 00:45:07.051 So the instruction isn't to join whatever the other person is doing. 703 00:45:07.091 --> 00:45:16.357 It's to read the mood and the energy and then decide whether to join in or whether to name it or acknowledge it, which is exactly what Bully did with that jury. 704 00:45:17.037 --> 00:45:17.338 When Bully, 705 00:45:17.998 --> 00:45:18.759 The jury got hot. 706 00:45:18.920 --> 00:45:21.864 He came in deliberately light against a rising argument. 707 00:45:21.964 --> 00:45:26.652 The other misunderstanding is that three conversations in here are just merely an idea. 708 00:45:27.393 --> 00:45:31.920 You'll recall there are three types of conversations that we typically have, at least according to Duhigg. 709 00:45:33.659 --> 00:45:36.000 The thing here is I think they're a diagnostic instrument. 710 00:45:36.340 --> 00:45:37.280 They tell you where you are. 711 00:45:37.600 --> 00:45:38.840 They're not necessarily an argument. 712 00:45:38.920 --> 00:45:49.923 The argument which Stu Higgs states once, and I don't know that he ever names it, is that demonstrating you heard somebody is what does the work, not hearing them and not it's actually showing it is what matters. 713 00:45:50.763 --> 00:45:54.844 It's why he replaces perspective taking with what he calls perspective getting. 714 00:45:55.524 --> 00:45:58.844 The textbook advice is to put yourself in someone else's shoes. 715 00:45:58.864 --> 00:46:01.025 In fact, I said this earlier, I talked about the importance of empathy. 716 00:46:02.090 --> 00:46:04.746 His view is that this is often impossible. 717 00:46:05.732 --> 00:46:11.674 And the best available substitute is to ask people to describe their inner lives and then listen. 718 00:46:12.334 --> 00:46:22.157 It's why the technique curates most highly repeating back what you heard in your own words and asking whether you got it right exists purely to make the listening visible. 719 00:46:22.797 --> 00:46:25.558 And that's why there's a whole section about laughter. 720 00:46:26.159 --> 00:46:32.461 A researcher recorded 1200 instances of people laughing in ordinary life and found that under 20% of it. 721 00:46:33.821 --> 00:46:36.864 was any response to something resembling a joke. 722 00:46:37.065 --> 00:46:39.767 We mostly laugh to show somebody we're with them. 723 00:46:40.368 --> 00:46:43.792 And it works as a signal precisely because it's hard to fake. 724 00:46:44.312 --> 00:46:50.819 People can distinguish friends laughing together from strangers trying to, literally from a one-second clip. 725 00:46:51.460 --> 00:46:54.503 So proof is the currency, not intent. 726 00:46:56.137 --> 00:47:01.661 The other one is you probably you may have come across this because it certainly got some buzz several years ago. 727 00:47:01.721 --> 00:47:03.102 It's the 36 questions. 728 00:47:03.302 --> 00:47:12.009 You know, in the 1990s, there were two psychologists, Elaine and Arthur Aaron, and they set out to find out whether you could reliably turn strangers into friends in a laboratory. 729 00:47:12.549 --> 00:47:19.214 They tested a lot of things that didn't work, shared interests, small talk, doing puzzles together, telling people they'd be carefully matched. 730 00:47:19.334 --> 00:47:21.296 None of it really produced closeness. 731 00:47:22.096 --> 00:47:30.240 One thing did, a sequence of 36 questions that escalate steadily in intimacy, pass back and forth between the participants. 732 00:47:31.341 --> 00:47:37.184 They start easy and they end somewhere like, of all the people in your family, whose death would you find most disturbing? 733 00:47:37.704 --> 00:47:41.506 And by the last few participants are often like crying because they've really bonded. 734 00:47:41.766 --> 00:47:47.229 The press later christened these, the 36 questions that lead to love. 735 00:47:47.749 --> 00:47:49.290 You can find them online if you Google it. 736 00:47:49.930 --> 00:47:52.691 They also only work when both people take turns. 737 00:47:53.412 --> 00:47:59.194 The researchers ran a version of the study where one person answered all 36 questions while the other listened, then swapped. 738 00:47:59.234 --> 00:48:03.136 And volunteers said it was awkward and boring and nobody felt close to anybody. 739 00:48:03.316 --> 00:48:04.617 But here's a misunderstanding. 740 00:48:04.637 --> 00:48:07.078 People might treat that list as a portable tool. 741 00:48:07.798 --> 00:48:11.701 And Duhigg does say in the book that that's not the case. 742 00:48:12.221 --> 00:48:16.104 Ask a stranger question 18 randomly out of thin air. 743 00:48:16.144 --> 00:48:18.866 You know, what's your most terrible memory at a networking event? 744 00:48:18.926 --> 00:48:20.967 And you're probably going to spend the evening alone. 745 00:48:21.427 --> 00:48:26.431 In the real world, Duhigg writes, the 36 questions are of little or of real help. 746 00:48:26.611 --> 00:48:29.053 What transfers is the underlying finding. 747 00:48:29.753 --> 00:48:36.038 Questions about beliefs, values, and meaningful experiences reliably

produce emotional answers. 748 00:48:36.058 --> 00:48:36.558 Okay. 749 00:48:36.738 --> 00:48:39.641 And questions about facts and logistics reliably don't. 750 00:48:40.482 --> 00:48:42.084 Where do you live is a fact. 751 00:48:43.465 --> 00:48:44.947 What do you like about your neighborhood? 752 00:48:46.008 --> 00:48:46.788 That's a value. 753 00:48:47.770 --> 00:48:49.031 It speaks to an experience. 754 00:48:49.191 --> 00:48:52.094 Same information, just completely different conversation. 755 00:48:52.715 --> 00:48:56.459 And there's a second misreading sitting next to this one, and it's about reciprocity. 756 00:48:57.418 --> 00:49:00.982 The book says reciprocity is a thing that makes all of this work. 757 00:49:01.403 --> 00:49:04.467 But reciprocity does not mean trading disclosures. 758 00:49:04.787 --> 00:49:10.174 If someone tells you that they've had a cancer diagnosis, answering with your own aches and pains is... 759 00:49:11.444 --> 00:49:14.426 is technically reciprocity, but it's not really reciprocity. 760 00:49:14.927 --> 00:49:17.289 It's turning the spotlight back on yourself. 761 00:49:17.869 --> 00:49:26.476 The reciprocal move is to work out what they need, comfort, empathy, advice, or someone to be straight with, whatever it is, and give them that. 762 00:49:27.057 --> 00:49:28.258 That's what the reciprocity is. 763 00:49:29.562 --> 00:49:31.123 Okay, let's move on to modern applications. 764 00:49:31.143 --> 00:49:35.887 So we have on our hands a book that focuses on interpersonal, individual communications. 765 00:49:36.087 --> 00:49:39.269 How does it translate to our work and how does it fit in the canon? 766 00:49:39.309 --> 00:49:40.971 I've got a few thoughts on this. 767 00:49:41.031 --> 00:49:42.412 We'll start with inversion. 768 00:49:43.793 --> 00:49:47.335 Here's what I think is the single most common failure in our business. 769 00:49:47.696 --> 00:49:53.460 And today, the gentleman we met at the very beginning, the urologist, gives us a solid illustration of it. 770 00:49:54.543 --> 00:50:02.986 It is enormously easier to decide what we want to communicate, then build a message architecture around what we need to convey than it is to actually do the reverse. 771 00:50:03.867 --> 00:50:04.967 That's the normal sequence, right? 772 00:50:05.007 --> 00:50:06.087 The client has an objective. 773 00:50:06.107 --> 00:50:07.868 We turn the objective into a narrative. 774 00:50:07.988 --> 00:50:09.549 The narrative becomes pillars. 775 00:50:09.589 --> 00:50:10.909 The pillars become proof points. 776 00:50:10.969 --> 00:50:12.170 The proof points get tested. 777 00:50:12.690 --> 00:50:17.972 And somewhere near the end, an audience gets consulted about whether they liked it or not. 778 00:50:18.712 --> 00:50:21.273 Aday ran that sequence perfectly, right? 779 00:50:21.533 --> 00:50:22.494 He had his objective. 780 00:50:23.545 --> 00:50:26.168 This man, this patient in front of me should not have surgery. 781 00:50:26.849 --> 00:50:27.529 He had the evidence. 782 00:50:27.810 --> 00:50:28.510 He built the case. 783 00:50:28.671 --> 00:50:29.732 He highlighted the sentence. 784 00:50:29.852 --> 00:50:31.133 It's in yellow, bright yellow. 785 00:50:31.434 --> 00:50:32.495 He kept it short and clear. 786 00:50:33.236 --> 00:50:41.325 And 40% of the time, across the whole profession, the patient walks out and books the needless surgery. 787 00:50:42.411 --> 00:50:44.774 The reverse sequence starts somewhere else entirely. 788 00:50:44.995 --> 00:50:54.988 It starts with what the receiver is already carrying, which conversation they're in, what they're protecting, which of their many selves is going to be doing the listening. 789 00:50:55.709 --> 00:50:58.373 And the messages get built to reach that. 790 00:50:59.194 --> 00:51:00.075 Like, I'm going to be honest with you. 791 00:51:00.115 --> 00:51:07.063 I said earlier, I'm still working out what the reverse looks like when it's done properly and how you spot the wrong version in a draft. 792 00:51:07.263 --> 00:51:08.745 I really don't have a clean test for it yet. 793 00:51:08.905 --> 00:51:13.750 What I do know is that almost everything in this book is an argument that the sequence is backwards. 794 00:51:14.411 --> 00:51:16.273 We tend to start at the top and work our way down. 795 00:51:16.293 --> 00:51:19.036 We've got to find a way to start at the bottom and work our way out. 796 00:51:19.786 --> 00:51:22.388 Another modern application is about empathy. 797 00:51:22.949 --> 00:51:24.650 In the fact that it's a stage, it's not an input. 798 00:51:25.391 --> 00:51:29.214 Empathy is one of the most important factors in whether we communicate well. 799 00:51:29.354 --> 00:51:31.315 It is also the hardest thing to do at scale. 800 00:51:31.355 --> 00:51:34.858 And it's a thing we spend the least time on when we're building an approach. 801 00:51:36.199 --> 00:51:37.701 And we'd all say we do it right. 802 00:51:37.781 --> 00:51:38.681 We do focus groups. 803 00:51:38.902 --> 00:51:39.902 We do in-depth interviews. 804 00:51:39.983 --> 00:51:43.085 We build audience archetypes and personas with names and photographs. 805 00:51:43.265 --> 00:51:45.047 And I think that's superficial. 806 00:51:45.367 --> 00:51:47.629 And I have a hard time articulating why. 807 00:51:49.340 --> 00:51:54.721 The best explanation I found is sitting in this book, and it's Ipley's distinction. 808 00:51:54.781 --> 00:52:00.222 Perspective taking is constructing a model of another person and then reasoning from inside it. 809 00:52:00.802 --> 00:52:05.243 An audience archetype is exactly that, a model built by us, reasoned from by us. 810 00:52:06.163 --> 00:52:10.704 And Ipley's whole point is that this doesn't work because you can't actually get inside someone. 811 00:52:11.464 --> 00:52:14.504 Perspective getting is asking and then listening to what comes back. 812 00:52:15.105 --> 00:52:17.085 Now, a focus group looks like

perspective getting. 813 00:52:18.160 --> 00:52:22.144 But what reaches a strategy isn't necessarily the room, right? 814 00:52:22.265 --> 00:52:30.133 It's the synthesis written by a moderator whose job was to keep the conversation on a guide read by people who weren't really in the room, aren't participating in that conversation. 815 00:52:30.994 --> 00:52:32.997 I don't think that fully accounts for my discomfort, though. 816 00:52:33.137 --> 00:52:35.419 It's part of it, but it's not all of it. 817 00:52:36.060 --> 00:52:38.902 The practical implication is one I'm confident about, though. 818 00:52:39.202 --> 00:52:45.466 Empathy needs its own stage in the process, and it needs to sit at the very start before the brief is typed out. 819 00:52:45.766 --> 00:52:49.889 It's not a research input that arrives once the objective is set. 820 00:52:50.732 --> 00:52:58.577 A stage with time and budget attached to it where the people who will write the messages are in the room and where the answers are given. 821 00:52:59.218 --> 00:53:08.204 If you think you or your organization do this, well, I'd love to hear from you because it's something I haven't yet fully been able to come up with a solution on. 822 00:53:08.384 --> 00:53:10.325 Okay, let's move on to our next modern application. 823 00:53:10.425 --> 00:53:11.126 It's the screen. 824 00:53:12.264 --> 00:53:21.111 Here's what I say to clients who want to know why face-to-face, face-to-face approaches can't just be ported over to digital platforms. 825 00:53:21.712 --> 00:53:27.917 The screen is only a few millimeters thin, but it is a massively thick barrier that distorts. 826 00:53:27.957 --> 00:53:32.660 The message never makes it through cleanly and never the way you intended it to be received. 827 00:53:33.481 --> 00:53:36.903 Everything in this book runs on a signal that a screen strips out. 828 00:53:37.544 --> 00:53:41.907 Tone, energy, the half-second pause, whether someone leaned in. 829 00:53:42.167 --> 00:53:46.970 The book's own advice for online conversation is also admittedly pretty thin by comparison. 830 00:53:47.750 --> 00:53:58.978 You know, findings show that you must be conspicuously polite, cut the sarcasm because the irony that you hear in your own head doesn't translate necessarily to the person on the receiving end. 831 00:53:59.858 --> 00:54:01.379 Hedge more than feels natural. 832 00:54:01.399 --> 00:54:06.881 You know, in fact, in one study, adding thanks and pleas to online arguments with nothing else changed. 833 00:54:07.042 --> 00:54:08.462 It measurably reduced the temperature. 834 00:54:08.622 --> 00:54:11.604 That's all real, but it feels small. 835 00:54:11.944 --> 00:54:16.146 And it isn't an answer to what happened on that Facebook group, right? 836 00:54:16.186 --> 00:54:20.348 Recall as soon as a group of trained facilitators who themselves... 837 00:54:20.988 --> 00:54:26.873 sitting on opposing views, were able to have reasonable and good conversations in

person. 838 00:54:26.893 --> 00:54:33.658 And once they were tasked to go out and apply what they learned and extend that onto that Facebook group, things fell by the wayside. 839 00:54:33.838 --> 00:54:39.182 The screen, digital platforms, online environment, which is where we spend so much of our time now, 840 00:54:40.490 --> 00:54:41.170 are problematic. 841 00:54:41.390 --> 00:54:46.912 And I think that makes it difficult for us to scale some of what we learn in this book. 842 00:54:47.492 --> 00:54:55.454 But that Facebook group's failure and the Karakosh success sitting right next to it do give me something I don't think I had before. 843 00:54:55.634 --> 00:55:00.655 The mistake in Washington with that Facebook group on gun rights was thinking that the intervention was a conversation. 844 00:55:01.275 --> 00:55:03.956 Teach people to listen, then put them somewhere and hope. 845 00:55:04.796 --> 00:55:07.537 Musa, with her experiment, never taught anybody anything. 846 00:55:07.657 --> 00:55:08.117 She just built 847 00:55:09.059 --> 00:55:10.721 a series of conditions and let the rest happen. 848 00:55:11.742 --> 00:55:13.003 People were interdependent. 849 00:55:13.183 --> 00:55:15.005 You cannot win without the person you're avoiding. 850 00:55:15.265 --> 00:55:24.054 Everyone was equal inside the boundary, enforced by rule rather than by good intentions, and everyone had a role worth holding that had nothing to do with the thing dividing them. 851 00:55:24.235 --> 00:55:27.678 And there was something at stake that people actually wanted. 852 00:55:27.738 --> 00:55:28.639 So those were the conditions. 853 00:55:29.937 --> 00:55:31.237 You build coalitions for a building. 854 00:55:31.277 --> 00:55:32.478 I built coalitions for a building. 855 00:55:32.938 --> 00:55:39.339 You already know, like we know how to get people in their room full of organizations that sometimes don't much like each other. 856 00:55:39.359 --> 00:55:40.119 All right. 857 00:55:40.139 --> 00:55:43.140 Particularly if you have to do a sector-wide, industry-wide coalition. 858 00:55:43.800 --> 00:55:48.841 But let me put two things from Caracos that I found genuinely awkward. 859 00:55:49.061 --> 00:55:50.701 And you can decide whether they hold. 860 00:55:51.061 --> 00:55:52.021 The first is about influence. 861 00:55:52.662 --> 00:55:56.402 In most of the coalitions I've worked on, influence tracks contributions. 862 00:55:56.502 --> 00:55:59.743 So the members bring in the most money, the most members, or the most political weight. 863 00:56:00.304 --> 00:56:01.386 have more say in the room. 864 00:56:01.967 --> 00:56:02.527 That's not a flaw. 865 00:56:02.547 --> 00:56:05.913 It's just substantially how you get the money and the members in the first place. 866 00:56:06.955 --> 00:56:08.197 Moosa's rule ran the other way. 867 00:56:08.457 --> 00:56:09.219 Equal playing time. 868 00:56:10.315 --> 00:56:12.496 enforced regardless of who was the better player. 869

00:56:12.916 --> 00:56:17.058 In her own reading is that deleting the power differential is a large part of why the thing worked. 870 00:56:17.619 --> 00:56:20.480 Karakosh had a steep hierarchy outside the touchline. 871 00:56:21.060 --> 00:56:26.703 Christians historically wealthier and better educated, and the old order reasserting itself as the refugees came home. 872 00:56:27.544 --> 00:56:31.046 Inside the touchline, for 90 minutes, it didn't exist. 873 00:56:31.866 --> 00:56:35.168 So I'm not going to tell you that influence shouldn't be equal in condition. 874 00:56:36.188 --> 00:56:44.380 What it does, though, is what, if anything, is genuinely equal in yours and whether it's equal because of a rule or because everyone is currently being gracious. 875 00:56:45.001 --> 00:56:45.802 Moose's was a rule. 876 00:56:46.022 --> 00:56:50.008 Graciousness tends not to survive the first real divergence of interest. 877 00:56:50.368 --> 00:56:52.291 The second is about what she was measuring. 878 00:56:53.032 --> 00:56:57.456 The headline result out of that league isn't who won the tournament, obviously. 879 00:56:57.657 --> 00:57:07.106 It's that players from the mixed teams were 49 points, more likely to still be in training alongside Muslims six months after the thing had completed. 880 00:57:07.946 --> 00:57:14.172 The trophy was the incentive to get started, but the relationship ended up, the relationships ended up being the outcome. 881 00:57:15.073 --> 00:57:20.115 Most of the coalitions I've been a part of sometimes run the other way around, right? 882 00:57:20.135 --> 00:57:21.236 The output is the outcome. 883 00:57:21.276 --> 00:57:31.140 The submissions, the campaign, the decision, and the relationships between members are instrumental to getting there, which means they're resourced until the file closes and then they aren't. 884 00:57:32.136 --> 00:57:34.477 And look, I'm not sure that's wrong, right? 885 00:57:34.497 --> 00:57:37.179 Sometimes you run a campaign and the campaign comes to its natural conclusion. 886 00:57:37.379 --> 00:57:47.606 In my case, clients pay for outcomes, files do close, and nobody's, you know, going to fund a standing relationship on the grounds that it might be useful in four years, although I argue often that they should. 887 00:57:48.106 --> 00:57:53.009 But a coalition built that way is spending relationship capital rather than accumulating it. 888 00:57:53.489 --> 00:57:58.993 Karakosh is a reminder that the second thing can be produced deliberately by design in conditions... 889 00:57:59.813 --> 00:58:05.237 a great deal worse than any file any of us is ever going to work. 890 00:58:05.337 --> 00:58:08.259 Final modern application I have is is the holdout. 891 00:58:08.299 --> 00:58:11.261 So every file has a Carl recall. 892 00:58:11.341 --> 00:58:17.305 Carl is our juror, the firefighter who cared about public safety, very sort of law and order ask in his thinking. 893 00:58:18.246 --> 00:58:21.448

The regulator who won't move is probably the analog, right? 894 00:58:21.468 --> 00:58:29.253 The committee member who blocks consensus, the coalition partner who agrees with the objective and objects just to every route that you're proposing to take it on. 895 00:58:29.433 --> 00:58:31.595 And I do have two observations from that jury room. 896 00:58:31.756 --> 00:58:36.501 So first, Carl told no stories about himself all afternoon while everybody else told several. 897 00:58:36.621 --> 00:58:38.062 And that was information. 898 00:58:38.463 --> 00:58:39.344 It was not obstruction. 899 00:58:39.364 --> 00:58:47.292 It's not that he was sitting out the conversation and not participating, but it did tell the room he was reachable in other ways, through logic, not through experience. 900 00:58:47.573 --> 00:58:48.253 And when someone, 901 00:58:49.074 --> 00:58:50.595 won't do the personal register. 902 00:58:51.035 --> 00:58:52.697 They're telling you which channel to use. 903 00:58:52.877 --> 00:58:54.838 And that's the thing Duhigg is all about in this book, right? 904 00:58:54.898 --> 00:58:56.239 Figure out what conversation you're having. 905 00:58:56.439 --> 00:58:58.941 Second, nobody asked him what he wanted for 90 minutes. 906 00:58:59.581 --> 00:59:00.462 They argued at him. 907 00:59:00.702 --> 00:59:04.084 And the question that finally moved things wasn't particularly clever. 908 00:59:04.184 --> 00:59:06.266 It was, your decision seems very complete in your mind. 909 00:59:06.346 --> 00:59:07.667 Share more of that with us. 910 00:59:08.707 --> 00:59:13.030 What came back was public safety, which nobody had processed up to that point. 911 00:59:13.110 --> 00:59:16.653 And which turned out to be the thing that could be argued with. 912 00:59:17.996 --> 00:59:20.378 And we shouldn't gloss over Baldy's move, right? 913 00:59:20.418 --> 00:59:26.862 He gave Carl a route to change his mind that didn't require admitting he'd been wrong this whole time. 914 00:59:27.362 --> 00:59:30.004 An explicit statement of respect for his rigor. 915 00:59:30.885 --> 00:59:33.706 Changing your position in front of people has a cost. 916 00:59:33.766 --> 00:59:35.327 As I said earlier, it's paid in your ego. 917 00:59:35.347 --> 00:59:42.973 And if you don't put something on the other side of that ledger, in our case, maybe a stakeholder who has publicly committed will hold a position they no longer believe. 918 00:59:43.773 --> 00:59:48.918 So it gets back into how we can shape our audiences, maybe get people to change minds. 919 00:59:50.919 --> 00:59:57.205 Now, before we get into the mastery lesson, I think there's an asymmetry to how this book will affect your work. 920 00:59:57.806 --> 01:00:02.049 If you internalize this book personally, I'd argue that changes is almost immediate. 921 01:00:02.109 --> 01:00:07.835 You will notice within days that the way you engage with your spouse, with your kids, with your colleagues will shift. 922 01:00:08.535

--> 01:00:09.955 I think that's what Duhigg intended. 923 01:00:10.455 --> 01:00:20.277 He wrote that book to fix his own life, and it will work in your own fast if you choose to be attentive to what you hear and understand and learn in this book. 924 01:00:20.477 --> 01:00:23.278 Professionally, though, it's not that immediate. 925 01:00:23.338 --> 01:00:28.939 There's no Monday morning where you deploy that matching principle on a regulatory file and watch it change your file overnight. 926 01:00:29.599 --> 01:00:31.880 What it does instead is, I think, slower. 927 01:00:32.660 --> 01:00:48.815 I think maybe more permanent, it surfaces a need to be constantly thinking about what kind of conversation you're going to have with your audience and how you're going to get to a point where that conversation is as close to one-to-one as it can possibly be so that you can actually engage with them meaningfully. 928 01:00:49.396 --> 01:00:51.538 Because that's the thing challenging people 929 01:00:51.718 --> 01:00:53.379 Large scale communicators right now. 930 01:00:53.700 --> 01:00:54.300 It's not reach. 931 01:00:54.340 --> 01:00:57.742 We have more reach than any generation of practitioners has ever had. 932 01:00:58.343 --> 01:00:59.784 Meaningful engagement is the problem. 933 01:00:59.904 --> 01:01:10.292 And it's most acute exactly where the stakes are highest on regulatory and political issues, where the decisions are made by people who are watching whether the public actually cares. 934 01:01:11.132 --> 01:01:12.733 So the standing question changes then. 935 01:01:12.793 --> 01:01:14.995 It stops being, what do we need them to hear? 936 01:01:15.816 --> 01:01:18.237 It becomes, how close to one-to-one can I get this? 937 01:01:18.458 --> 01:01:20.039 And what would that take? 938 01:01:21.346 --> 01:01:27.692 It's not a technique, it's an orientation and you'll be applying it to work that is structurally pulling in the opposite direction. 939 01:01:27.792 --> 01:01:31.916 And one to one doesn't scale in any way a budget line will recognize. 940 01:01:32.717 --> 01:01:34.899 Which brings me to a sobering lesson. 941 01:01:35.860 --> 01:01:37.001 You are mostly going to fail at this. 942 01:01:37.802 --> 01:01:41.325 Meaningful engagement at scale is the unsolved problem of our field. 943 01:01:41.566 --> 01:01:45.389 I'm sure I'm sure there are many unsolved problems, but it is one unsolved problem of our field. 944 01:01:46.392 --> 01:01:48.214 Duhigg is not going to, doesn't solve it for us. 945 01:01:48.234 --> 01:01:49.175 He doesn't even attempt it. 946 01:01:49.535 --> 01:01:51.337 It's not really what his book is about. 947 01:01:51.538 --> 01:01:56.523 And, you know, as I said earlier and a couple of times throughout this episode, I haven't solved it either. 948 01:01:56.623 --> 01:02:00.688 And I'm not going to stand here and pretend that reading one book is going to change that. 949 01:02:00.888 --> 01:02:03.751 But what does change are two things. 950 01:02:04.672 --> 01:02:13.423 You stop mistaking reach for engagement, which is a harder discipline than it sounds when reach is the number in a report, when it's a KPI sitting in a dashboard. 951 01:02:13.604 --> 01:02:17.989 And you start treating every increment of closeness as something worth paying for. 952 01:02:18.865 --> 01:02:26.266 And there's a version of closer to one to one that survives contact with a budget, which is the thing I didn't have before I read this book. 953 01:02:26.946 --> 01:02:29.727 It sounds like it means having more conversations. 954 01:02:30.547 --> 01:02:32.707 Musa didn't have a single conversation. 955 01:02:33.788 --> 01:02:40.209 She built the conditions in which one to one happened by itself across 42 teams in a city recovering from a massacre. 956 01:02:40.269 --> 01:02:42.509 And it held for six months after she went home. 957 01:02:43.749 --> 01:02:46.390 Getting closer to one to one might mean... 958 01:02:47.465 --> 01:02:49.227 designing the room rather than filling it. 959 01:02:50.568 --> 01:02:51.449 All right, let's close this out. 960 01:02:51.469 --> 01:02:54.653 I'm going to take you back to Sloan Catering Cancer Center. 961 01:02:55.033 --> 01:03:01.880 A day changed how he ran those appointments, how he had those conversations with his prostate cancer patients. 962 01:03:02.521 --> 01:03:06.906 Within six months, the number of his patients choosing surgery fell by 30%. 963 01:03:09.128 --> 01:03:10.309 He didn't get better at explaining. 964 01:03:10.489 --> 01:03:11.530 The data didn't improve. 965 01:03:11.870 --> 01:03:19.096 He never found a clearer way to describe what active surveillance is or a better chart or a more persuasive study to highlight in yellow. 966 01:03:19.836 --> 01:03:22.218 He just got better at asking. 967 01:03:23.039 --> 01:03:30.424 He started asking his patients the same basic questions in different ways over and over until somebody said something that revealed what actually mattered to them. 968 01:03:31.205 --> 01:03:33.006 And what came back was never medicine. 969 01:03:33.927 --> 01:03:35.748 Some of them were frightened and wanted reassurance. 970 01:03:36.028 --> 01:03:38.248 Some wanted to feel like they were in control of something. 971 01:03:38.429 --> 01:03:41.910 Some needed to hear how other men had made the same decision. 972 01:03:42.330 --> 01:03:44.711 And some of them just wanted the doctor to tell them what to do. 973 01:03:46.311 --> 01:03:51.373 He is now teaching other surgeons how to do this, on opioids, on breast cancer, on end-of-life decisions. 974 01:03:51.833 --> 01:03:54.474 And for years, he'd had the right answer. 975 01:03:55.234 --> 01:03:57.515 What he was missing was the right question. 976 01:03:58.624 --> 01:04:01.585 And that's it for this episode of Masters in Public Affairs. 977 01:04:01.865 --> 01:04:04.226 Let me know what you thought of this episode, what you thought of this book. 978 01:04:04.266 --> 01:04:06.807 The book is Super Communicators by Charles Duhigg. 979 01:04:07.167 --> 01:04:10.968 If you've not had a chance yet, please do subscribe to the show wherever you get your podcasts. 980 01:04:11.328 --> 01:04:13.529 We're on YouTube, we're on Spotify, we're on Apple. 981 01:04:14.069 --> 01:04:18.510 And don't be afraid to give the show a five-star review to help with discovery. 982 01:04:18.850 --> 01:04:20.111 And our website is now live. 983 01:04:20.131 --> 01:04:23.392 You can point your browser to mastersinpublicaffairs.com. 984 01:04:24.172 --> 01:04:32.793 I now have an accompanying short essay with every single episode, a slightly different take on the episode, lessons learned, and other thoughts as well. 985 01:04:33.235 --> 01:04:35.159 And so do join the newsletter if you haven't. 986 01:04:35.841 --> 01:04:38.347 Again, mastersinpublicaffairs.com. 987 01:04:38.728 --> 01:04:39.871 Okay, guys, we'll talk again.