1 00:00:05,080 --> 00:00:08,399 Speaker 1: Can you get convicted in a court of law based 2 00:00:08,480 --> 00:00:13,920 Speaker 1: on your brain activity? Our brain scan lie detectors going 3 00:00:14,000 --> 00:00:18,640 Speaker 1: to be accepted? And when does measuring somebody's brain count 4 00:00:18,800 --> 00:00:22,800 Speaker 1: as illegal search and seizure? When does it violate the 5 00:00:22,960 --> 00:00:28,080 Speaker 1: privacy of the mind? What constitutional issues are triggered by 6 00:00:28,080 --> 00:00:30,840 Speaker 1: all this? And what does this have to do with 7 00:00:31,080 --> 00:00:34,960 Speaker 1: whether your mouth gets dry when you lie, or the 8 00:00:35,120 --> 00:00:39,040 Speaker 1: orbits under your eyes get hot, or your voice constricts, 9 00:00:39,600 --> 00:00:42,040 Speaker 1: and what the networks in your brain are up to 10 00:00:42,760 --> 00:00:50,760 Speaker 1: when you deceive. Welcome to Inner Cosmos with me David Eagleman. 11 00:00:51,080 --> 00:00:54,200 Speaker 1: I'm a neuroscientist and author at Stanford and in these 12 00:00:54,240 --> 00:00:58,200 Speaker 1: episodes we sail deeply into our three pound universe to 13 00:00:58,320 --> 00:01:02,120 Speaker 1: understand why and how our lives look the way they do. 14 00:01:11,400 --> 00:01:16,319 Speaker 1: Today's episode is about the fascinating topic of whether we 15 00:01:16,440 --> 00:01:22,080 Speaker 1: can use technology to figure out whether a person is lying. 16 00:01:22,760 --> 00:01:26,520 Speaker 1: What is lie detection actually measuring? How good is it? 17 00:01:26,640 --> 00:01:29,679 Speaker 1: When is it accepted in courts of law? And as 18 00:01:29,680 --> 00:01:33,760 Speaker 1: we come to have better and better brain reading technology, 19 00:01:34,319 --> 00:01:38,400 Speaker 1: what is the future of lie detection? So in a 20 00:01:38,440 --> 00:01:41,240 Speaker 1: court of law, someone says I didn't do it. I 21 00:01:41,280 --> 00:01:44,240 Speaker 1: wasn't even there that night, I was in my apartment 22 00:01:44,319 --> 00:01:48,800 Speaker 1: watching TV. If there's no other information that can be gathered, 23 00:01:49,360 --> 00:01:51,600 Speaker 1: like his cell phone is off, so there's no GPS 24 00:01:51,680 --> 00:01:55,960 Speaker 1: signal and there's no eyewitnesses, there's no alimis. If there's 25 00:01:56,040 --> 00:02:00,760 Speaker 1: no other information besides his testimony, how do you know 26 00:02:00,840 --> 00:02:03,880 Speaker 1: if that's really what happened or not? Is he telling 27 00:02:03,960 --> 00:02:08,720 Speaker 1: the truth or is he lying? Now? On every crime 28 00:02:08,800 --> 00:02:12,760 Speaker 1: television show, there are lots of clever clues that are 29 00:02:12,880 --> 00:02:17,400 Speaker 1: surfaced by the industrious detective. But the issue in real 30 00:02:17,480 --> 00:02:20,200 Speaker 1: courts of law most of the time is that other 31 00:02:20,360 --> 00:02:24,960 Speaker 1: data just isn't available, so you only have the testimony 32 00:02:25,080 --> 00:02:29,040 Speaker 1: of the people who were involved. Most crime doesn't happen 33 00:02:29,480 --> 00:02:33,639 Speaker 1: in fancy office buildings with full video coverage, but instead, 34 00:02:33,840 --> 00:02:37,800 Speaker 1: more commonly happens in areas without any coverage, and so 35 00:02:38,320 --> 00:02:42,079 Speaker 1: you need to ask people probing questions about whether they 36 00:02:42,120 --> 00:02:46,600 Speaker 1: were involved and what happened, and given human nature, you 37 00:02:46,720 --> 00:02:49,520 Speaker 1: really want to know whether they are telling the truth 38 00:02:49,800 --> 00:02:55,160 Speaker 1: or not. So let's start today's episode with two legal cases. 39 00:02:55,760 --> 00:02:59,840 Speaker 1: One involves a woman named Aditi Sharma. She moved to 40 00:03:00,080 --> 00:03:03,000 Speaker 1: a town in India with her boyfriend Udit. This was 41 00:03:03,080 --> 00:03:06,600 Speaker 1: an arranged marriage they were going to be together forever. 42 00:03:07,280 --> 00:03:09,600 Speaker 1: But while she was there, she met a new man 43 00:03:09,880 --> 00:03:12,720 Speaker 1: named Provene, and they fell in love with each other 44 00:03:12,880 --> 00:03:16,079 Speaker 1: and decided that they wanted to be together rather than 45 00:03:16,240 --> 00:03:20,280 Speaker 1: her and Udit being together. So she ran away with Provene, 46 00:03:20,800 --> 00:03:25,399 Speaker 1: but Udit kept following them, so she and Pravene decided 47 00:03:25,440 --> 00:03:28,560 Speaker 1: that they would murder Udit, and they did so by 48 00:03:28,600 --> 00:03:31,520 Speaker 1: inviting him to meet them Adam McDonald's, and they gave 49 00:03:31,600 --> 00:03:35,080 Speaker 1: him a popular drink in India that was laced with 50 00:03:35,360 --> 00:03:38,520 Speaker 1: arsenic and he drank it and died. So she was 51 00:03:38,520 --> 00:03:42,120 Speaker 1: brought up on charges of murder, and in court she 52 00:03:42,320 --> 00:03:45,960 Speaker 1: denied the crime. But part of what led her murder 53 00:03:46,040 --> 00:03:50,360 Speaker 1: conviction was a lie detection test what's called an electro 54 00:03:50,480 --> 00:03:54,840 Speaker 1: encephalogram or EEG. So we'll come back to this in 55 00:03:54,880 --> 00:03:57,480 Speaker 1: a few moments now. Case number two is in the 56 00:03:57,560 --> 00:04:01,920 Speaker 1: United States, a case in Tennessee. A man named Semrau 57 00:04:02,200 --> 00:04:05,640 Speaker 1: was the CEO of two nursing home facilities, and these 58 00:04:05,640 --> 00:04:11,000 Speaker 1: facilities were accused of having the employees fraudulently fill out 59 00:04:11,080 --> 00:04:15,120 Speaker 1: Medicare and Medicaid forms, and Semrau said, I had no 60 00:04:15,280 --> 00:04:17,599 Speaker 1: knowledge of this. I did not know this was happening 61 00:04:17,640 --> 00:04:21,279 Speaker 1: in my facilities. And so in going to court, he 62 00:04:21,480 --> 00:04:25,279 Speaker 1: enlisted the help of a lie detection company that used 63 00:04:25,680 --> 00:04:30,120 Speaker 1: new brain scan lie detection. This is a new technology 64 00:04:30,120 --> 00:04:32,800 Speaker 1: that we'll talk about today. So this company did the 65 00:04:32,800 --> 00:04:37,000 Speaker 1: brain scan and the results suggested that in fact Semrau 66 00:04:37,240 --> 00:04:40,479 Speaker 1: was not lying, that he had no knowledge of the fraud. 67 00:04:40,720 --> 00:04:43,479 Speaker 1: And so he went to court and the question is 68 00:04:43,839 --> 00:04:47,719 Speaker 1: did the court accept the brain scan? The answer is 69 00:04:47,839 --> 00:04:50,279 Speaker 1: they did not. They wouldn't accept in court. They said, 70 00:04:50,640 --> 00:04:54,120 Speaker 1: you can't present that here. So should they have accepted it? 71 00:04:54,440 --> 00:04:57,279 Speaker 1: How do we decide when to accept a new technology 72 00:04:57,360 --> 00:05:01,440 Speaker 1: or not? Under what circumstances do you say this technology 73 00:05:01,560 --> 00:05:03,880 Speaker 1: is good enough that it's telling us what we need 74 00:05:03,920 --> 00:05:06,080 Speaker 1: to know. So these are all the issues that we're 75 00:05:06,080 --> 00:05:08,479 Speaker 1: going to talk about today, and we'll get into the 76 00:05:08,520 --> 00:05:11,480 Speaker 1: background of what a lie is and how you could 77 00:05:11,520 --> 00:05:14,400 Speaker 1: measure something about it and where this is all going. 78 00:05:15,279 --> 00:05:18,000 Speaker 1: So we'll start with the basic question of what is 79 00:05:18,120 --> 00:05:22,000 Speaker 1: a lie. Well, first, lying might seem like a human invention. 80 00:05:22,200 --> 00:05:26,960 Speaker 1: So it's interesting to know that we see deception throughout 81 00:05:27,000 --> 00:05:30,720 Speaker 1: the kingdom of life. For example, in the world of plants, 82 00:05:31,160 --> 00:05:34,240 Speaker 1: there's a type of orchid that develops to look just 83 00:05:34,360 --> 00:05:38,240 Speaker 1: like a female bee, and this way male bees come 84 00:05:38,279 --> 00:05:41,120 Speaker 1: and land on it, and then the plant eats them. 85 00:05:41,440 --> 00:05:45,000 Speaker 1: It's a strategy of deception to lure in the bee, 86 00:05:45,480 --> 00:05:48,320 Speaker 1: and we see this throughout the animal kingdom. For example, 87 00:05:48,839 --> 00:05:53,200 Speaker 1: fireflies blink to say, hey, I'm here, please meet with me. 88 00:05:53,880 --> 00:05:58,520 Speaker 1: And there's a predatory firefly beetle named Fotaurus, and this 89 00:05:58,680 --> 00:06:04,880 Speaker 1: meetle has evolved bioluminescence so that it flashes light signals 90 00:06:04,880 --> 00:06:09,279 Speaker 1: of other firefly species, and that attracts the males and 91 00:06:09,320 --> 00:06:12,200 Speaker 1: then they kill and eat the males. So this is 92 00:06:12,240 --> 00:06:16,880 Speaker 1: known as aggressive mimicry. They're pretending they're the other fireflies. 93 00:06:17,000 --> 00:06:18,840 Speaker 1: So one way we can think about this is that 94 00:06:18,920 --> 00:06:23,720 Speaker 1: they are lying. They're not displaying truthy signals about who 95 00:06:23,720 --> 00:06:27,559 Speaker 1: they are, but instead they are deceiving to accomplish their ends. 96 00:06:28,000 --> 00:06:31,480 Speaker 1: So deception is not simply a human thing. But let's 97 00:06:31,520 --> 00:06:34,320 Speaker 1: turn to humans because that's what we care about. So 98 00:06:34,480 --> 00:06:38,680 Speaker 1: the first question is how do we define a lie? Well, 99 00:06:39,080 --> 00:06:41,880 Speaker 1: it turns out that's not so easy. Let's consider an 100 00:06:41,920 --> 00:06:46,760 Speaker 1: example like the scientist Copernicus. He developed a mathematical model 101 00:06:46,800 --> 00:06:49,240 Speaker 1: in which the sun goes around the earth, and that 102 00:06:49,320 --> 00:06:52,400 Speaker 1: turns out to be totally wrong. But he wasn't lying, 103 00:06:52,520 --> 00:06:57,000 Speaker 1: he wasn't intending to deceive. And occasionally what you'll find 104 00:06:57,080 --> 00:07:00,599 Speaker 1: is a situation where someone tells a man this truth, 105 00:07:00,920 --> 00:07:03,000 Speaker 1: but they had no intention to deceive. I'll give you 106 00:07:03,080 --> 00:07:05,880 Speaker 1: an example. Some years ago, there was a Harvard kid 107 00:07:05,880 --> 00:07:08,800 Speaker 1: who went out and drank too much, and as he 108 00:07:08,880 --> 00:07:12,440 Speaker 1: was stumbling home, he was jumped by two locals who 109 00:07:12,480 --> 00:07:15,960 Speaker 1: beat him up badly and smashed his head into the sidewalk. 110 00:07:16,120 --> 00:07:19,720 Speaker 1: So in an effort to save himself, he pulled a 111 00:07:19,760 --> 00:07:22,840 Speaker 1: pen knife out of his pocket and he stabbed at 112 00:07:22,840 --> 00:07:25,840 Speaker 1: the assailants to get them away. So they ran away 113 00:07:26,320 --> 00:07:29,600 Speaker 1: seemingly unharmed. But it turns out his knife had gone 114 00:07:29,960 --> 00:07:33,360 Speaker 1: through the chest of one assailant and tore the lining 115 00:07:33,400 --> 00:07:37,160 Speaker 1: around the heart, causing what's called a cardiac tamponaud in 116 00:07:37,200 --> 00:07:40,560 Speaker 1: which blood fills the lining around the heart and squishes it, 117 00:07:40,840 --> 00:07:45,920 Speaker 1: and within hours this assailant died. Now, when the police 118 00:07:46,040 --> 00:07:49,520 Speaker 1: came to question this Harvard kid, he told them a 119 00:07:49,640 --> 00:07:54,160 Speaker 1: totally untrue story about what had happened. Why, While a 120 00:07:54,240 --> 00:07:58,000 Speaker 1: neurologist and researcher named Jeremy Schmahman was called in to 121 00:07:58,160 --> 00:08:01,360 Speaker 1: diagnose and testify, and he realized it was because this 122 00:08:01,480 --> 00:08:05,000 Speaker 1: young man had been badly concussed, His head had been 123 00:08:05,040 --> 00:08:09,160 Speaker 1: pounded into the sidewalk, and his brain had constructed a 124 00:08:09,280 --> 00:08:13,520 Speaker 1: false memory about the whole event. So the opposing lawyer asked, 125 00:08:13,920 --> 00:08:18,360 Speaker 1: do concussions make you lie? But Schmaman explained that a 126 00:08:18,520 --> 00:08:23,840 Speaker 1: patient with a concussion is not lying, he's confabulating. There's 127 00:08:23,840 --> 00:08:28,800 Speaker 1: a critical difference there because the patient believes it. In 128 00:08:28,840 --> 00:08:32,959 Speaker 1: other words, just because a statement is not factual doesn't 129 00:08:33,000 --> 00:08:35,880 Speaker 1: make it a lie. And in a future episode, I'm 130 00:08:35,920 --> 00:08:40,000 Speaker 1: going to dive really deeply into this issue about confabulation 131 00:08:40,240 --> 00:08:43,560 Speaker 1: because we see this all the time in many different 132 00:08:43,600 --> 00:08:47,040 Speaker 1: types of brain damage. This even happened to a Supreme 133 00:08:47,120 --> 00:08:50,959 Speaker 1: Court Justice, William Douglas, who ended up with a disorder 134 00:08:51,040 --> 00:08:54,800 Speaker 1: called a nosygnosia in which he lost the ability to 135 00:08:54,920 --> 00:08:58,319 Speaker 1: control part of his body, and yet he spoke falsely 136 00:08:58,400 --> 00:09:02,280 Speaker 1: about it. The key is he was lying, he was confabulating. 137 00:09:02,360 --> 00:09:05,880 Speaker 1: He believed that he could still move his body. So 138 00:09:06,240 --> 00:09:09,720 Speaker 1: there are many situations where a person can believe something 139 00:09:09,920 --> 00:09:14,480 Speaker 1: falsely and that is not equivalent to a lie. And 140 00:09:14,559 --> 00:09:16,560 Speaker 1: of course, this comes up all the time when people 141 00:09:16,600 --> 00:09:21,080 Speaker 1: are asked about their memory of an event, because memory 142 00:09:21,520 --> 00:09:25,440 Speaker 1: doesn't store sequences of zeros and ones like a computer does. 143 00:09:25,920 --> 00:09:28,679 Speaker 1: And I talked about this at length in the context 144 00:09:28,679 --> 00:09:32,360 Speaker 1: of eyewitness testimony in episode twenty seven. We are not 145 00:09:32,760 --> 00:09:36,040 Speaker 1: like a video recorder, and my memory of something that 146 00:09:36,080 --> 00:09:40,320 Speaker 1: went down might genuinely be very different from your memory. 147 00:09:40,480 --> 00:09:42,760 Speaker 1: But again, if we both probe our memories in the 148 00:09:42,800 --> 00:09:46,160 Speaker 1: court of law, and we have different versions, it doesn't 149 00:09:46,200 --> 00:09:51,199 Speaker 1: require that one of us is purposefully lying. But let's 150 00:09:51,240 --> 00:09:57,160 Speaker 1: turn to actual deceptive lies, where you're intending to fool somebody. 151 00:09:57,200 --> 00:10:02,080 Speaker 1: What do we know about that? Welltive lying starts in 152 00:10:02,480 --> 00:10:06,480 Speaker 1: children at about the age of five, they start to lie. 153 00:10:06,600 --> 00:10:11,520 Speaker 1: It's a cognitive development. When you're very young, you believe 154 00:10:11,559 --> 00:10:14,760 Speaker 1: that your parents know everything. But around five years into 155 00:10:14,800 --> 00:10:18,079 Speaker 1: the game, you realize that your parents are not omniscient, 156 00:10:18,160 --> 00:10:20,960 Speaker 1: they don't know everything that happened, and it strikes you 157 00:10:21,040 --> 00:10:24,920 Speaker 1: that you can inject false information. And so we train 158 00:10:25,000 --> 00:10:27,800 Speaker 1: our youth on the value of truth telling because this 159 00:10:27,880 --> 00:10:31,680 Speaker 1: is very pro social behavior. But the fact is, when 160 00:10:31,679 --> 00:10:36,080 Speaker 1: the stakes are high, humans will often continue to practice deception, 161 00:10:36,240 --> 00:10:40,840 Speaker 1: perhaps in small ways throughout their lives. Now, interestingly, we 162 00:10:40,920 --> 00:10:44,120 Speaker 1: see that other primates do the same thing. They do 163 00:10:44,200 --> 00:10:48,080 Speaker 1: what's called tactical deception. And the general story we find 164 00:10:48,120 --> 00:10:52,600 Speaker 1: across primate species is that the bigger the brain they have, 165 00:10:53,240 --> 00:10:57,240 Speaker 1: the more deception they do, because bigger brains allow us 166 00:10:57,320 --> 00:11:01,920 Speaker 1: to model what other brains are thinking, which allows us 167 00:11:01,960 --> 00:11:05,199 Speaker 1: to be better at manipulating them. But what's interesting is 168 00:11:05,240 --> 00:11:09,440 Speaker 1: that understanding what other brains are thinking is also the 169 00:11:09,520 --> 00:11:14,320 Speaker 1: development that allows species to build larger social groups with 170 00:11:14,600 --> 00:11:19,160 Speaker 1: richer collaborations. So we find this strange truth that the 171 00:11:19,200 --> 00:11:24,040 Speaker 1: same evolution that allows us to live in groups harmoniously 172 00:11:24,600 --> 00:11:28,920 Speaker 1: also allows us to deceive. Okay, so you might think, well, 173 00:11:29,000 --> 00:11:31,160 Speaker 1: we have these big brains and we're able to lie. 174 00:11:31,320 --> 00:11:34,040 Speaker 1: But you might think, how could you possibly build a 175 00:11:34,200 --> 00:11:38,640 Speaker 1: technology that can tell whether something is a truth or untruth. 176 00:11:39,080 --> 00:11:44,160 Speaker 1: It all comes down to a simple principle. Lying is hard. 177 00:11:44,600 --> 00:11:48,199 Speaker 1: Why because you have to think of the true testimony 178 00:11:48,520 --> 00:11:51,200 Speaker 1: and inhibit that so you don't blurt it out, and 179 00:11:51,240 --> 00:11:54,920 Speaker 1: then you have to generate a plausible alternative, and all 180 00:11:54,960 --> 00:11:56,720 Speaker 1: the while you have to do a good job of 181 00:11:56,880 --> 00:12:01,760 Speaker 1: hiding your anxiety and guilt. You have to manage your emotions. 182 00:12:02,000 --> 00:12:05,120 Speaker 1: You have to inhibit any behaviors that would give you away. 183 00:12:05,200 --> 00:12:09,559 Speaker 1: You might try to generate honest looking behaviors. So that's 184 00:12:09,559 --> 00:12:11,280 Speaker 1: what we're going to talk about today. How do you 185 00:12:11,360 --> 00:12:14,920 Speaker 1: build a technology to detect all of that extra work 186 00:12:15,200 --> 00:12:33,560 Speaker 1: that goes into a lie. Now, what you might find 187 00:12:33,640 --> 00:12:37,880 Speaker 1: unexpected is that the idea of lie detection of using 188 00:12:37,880 --> 00:12:41,360 Speaker 1: a technology in the courtroom, this has a long history. 189 00:12:41,559 --> 00:12:44,840 Speaker 1: So take a guess when the first lie detector was 190 00:12:44,920 --> 00:12:50,080 Speaker 1: used in courts two thousand years ago in China, and 191 00:12:50,120 --> 00:12:56,240 Speaker 1: the technology was uncooked rice. The defendant would put uncooked 192 00:12:56,280 --> 00:12:59,920 Speaker 1: rice in his mouth and he was then asked questions. Now, 193 00:13:00,320 --> 00:13:04,440 Speaker 1: presumably his answers were marginally more difficult to understand, but 194 00:13:04,480 --> 00:13:07,959 Speaker 1: that wasn't the important part. The critical test was whether 195 00:13:08,000 --> 00:13:10,680 Speaker 1: he could spit the rice back out of his mouth. 196 00:13:11,080 --> 00:13:14,760 Speaker 1: If he couldn't, it was assumed that his dry mouth 197 00:13:15,320 --> 00:13:19,640 Speaker 1: served as a physiologic marker of his nervousness about lying, 198 00:13:20,040 --> 00:13:21,959 Speaker 1: And the idea was that if they were telling the truth, 199 00:13:22,200 --> 00:13:24,360 Speaker 1: they'd have plenty of saliva in their mouth and they'd 200 00:13:24,360 --> 00:13:27,280 Speaker 1: be able to spit the rice out. So this was 201 00:13:27,320 --> 00:13:31,480 Speaker 1: the first lie detector test. But Eventually, starting about a 202 00:13:31,520 --> 00:13:35,360 Speaker 1: century ago, humans leveled this up. They invented what's commonly 203 00:13:35,400 --> 00:13:39,439 Speaker 1: known as the polygraph test, which is usually what people 204 00:13:39,480 --> 00:13:42,760 Speaker 1: refer to when they talk about a lie detector. Now, 205 00:13:42,760 --> 00:13:45,520 Speaker 1: what you do is you attach some sensors to a 206 00:13:45,559 --> 00:13:50,439 Speaker 1: person's hand, and you're measuring their physiologic responses while they're 207 00:13:50,480 --> 00:13:54,000 Speaker 1: answering questions. You're measuring their heart rate, their blood pressure, 208 00:13:54,240 --> 00:13:58,680 Speaker 1: their breathing, their skin conductance. And the underlying principle is 209 00:13:58,679 --> 00:14:03,600 Speaker 1: that when a person lives, their body reveals changes because 210 00:14:03,600 --> 00:14:08,200 Speaker 1: of the stress and anxiety associated with deception, and so 211 00:14:08,320 --> 00:14:12,960 Speaker 1: the examiner analyzes the recordings to look for patterns or 212 00:14:13,000 --> 00:14:17,480 Speaker 1: deviations that indicate lying. So to make this clear, let's 213 00:14:17,559 --> 00:14:21,080 Speaker 1: zoom in on the issue of skin conductance. So you've 214 00:14:21,120 --> 00:14:24,000 Speaker 1: got all these little sweat glands in your skin, and 215 00:14:24,040 --> 00:14:27,760 Speaker 1: when you have a sudden stress response, those sweat glands 216 00:14:27,800 --> 00:14:30,600 Speaker 1: open more, which would cause you to sweat more. But 217 00:14:30,640 --> 00:14:33,440 Speaker 1: it's not sweat that's being measured here. What's being measured 218 00:14:34,200 --> 00:14:37,720 Speaker 1: is that there's a slight electrical current between point A 219 00:14:37,880 --> 00:14:42,120 Speaker 1: and point B across your skin, and the conductance between 220 00:14:42,200 --> 00:14:46,800 Speaker 1: A and B changes when those sweat glands open, so 221 00:14:46,960 --> 00:14:51,560 Speaker 1: the electrical conductance across your skin, or inversely the resistance 222 00:14:52,120 --> 00:14:55,520 Speaker 1: that changes. And so the measure here is just looking 223 00:14:55,560 --> 00:14:59,240 Speaker 1: at how well electricity can pass from here to there 224 00:14:59,280 --> 00:15:03,000 Speaker 1: across your skin. And when you are stressed, the electricity 225 00:15:03,040 --> 00:15:05,800 Speaker 1: passes better. So you get a blip on the graph 226 00:15:06,160 --> 00:15:09,840 Speaker 1: when you have a sudden moment of stress. Now here's 227 00:15:09,880 --> 00:15:13,960 Speaker 1: the critical point. It's not measuring a lie, it's just 228 00:15:14,000 --> 00:15:18,200 Speaker 1: measuring the stress response that's associated with lying. You're looking 229 00:15:18,240 --> 00:15:22,800 Speaker 1: for that stress response as a correlate of deception. So 230 00:15:23,040 --> 00:15:25,640 Speaker 1: you ask me, were you at this location on the 231 00:15:25,720 --> 00:15:28,880 Speaker 1: night of March first, and I say no, I wasn't. 232 00:15:29,160 --> 00:15:32,400 Speaker 1: Now if I actually was there, the lie is stressful, 233 00:15:32,480 --> 00:15:35,200 Speaker 1: even at a subconscious level, and that can be read 234 00:15:35,240 --> 00:15:38,840 Speaker 1: out because I have an automatic under the hood stress 235 00:15:38,920 --> 00:15:43,760 Speaker 1: response that registers. The sweat glands on my skin open slightly, 236 00:15:44,000 --> 00:15:48,560 Speaker 1: which changes the electrical conductance across my skin, Maybe my 237 00:15:48,680 --> 00:15:51,480 Speaker 1: breathing changes a bit, my heart rate changes slightly, and 238 00:15:51,520 --> 00:15:55,880 Speaker 1: as a result, this deception is brought to light. Now, 239 00:15:56,560 --> 00:15:59,280 Speaker 1: this is the common way to use the polygraph test, 240 00:15:59,320 --> 00:16:01,920 Speaker 1: but there's also second way that it's used, and that's 241 00:16:02,040 --> 00:16:06,200 Speaker 1: known as the guilty knowledge technique. It turns out that 242 00:16:06,280 --> 00:16:10,200 Speaker 1: if I show you something that's familiar to you versus unfamiliar, 243 00:16:10,680 --> 00:16:15,760 Speaker 1: your brain will have a different unconscious response. So imagine 244 00:16:15,800 --> 00:16:18,760 Speaker 1: I show you something like a strange little three D 245 00:16:19,000 --> 00:16:22,160 Speaker 1: statue of a Pokemon. So if it's the first time 246 00:16:22,200 --> 00:16:25,560 Speaker 1: you've ever seen that statue, your brain has a different 247 00:16:25,600 --> 00:16:28,920 Speaker 1: response to it than if it's something you've seen before. 248 00:16:29,720 --> 00:16:33,560 Speaker 1: So the police or the private investigator can leverage the 249 00:16:33,680 --> 00:16:37,360 Speaker 1: guilty knowledge technique because let's say they know that this 250 00:16:38,000 --> 00:16:41,280 Speaker 1: unique little statue of the Pokemon was at the scene 251 00:16:41,280 --> 00:16:43,720 Speaker 1: of the crime. It was on the desk of the 252 00:16:43,760 --> 00:16:46,520 Speaker 1: man who got murdered. So they sit you down in 253 00:16:46,560 --> 00:16:49,720 Speaker 1: the chair and they put the polygraph electrodes on your skin, 254 00:16:50,240 --> 00:16:52,960 Speaker 1: and then they show you a can of coke, and 255 00:16:52,960 --> 00:16:54,920 Speaker 1: then they show you a computer printer, and then they 256 00:16:54,920 --> 00:16:57,480 Speaker 1: show you a large spoon, and then they show you 257 00:16:57,880 --> 00:17:02,240 Speaker 1: the Pokemon statue. Woh big response. Your brain has a 258 00:17:02,280 --> 00:17:05,880 Speaker 1: different response because you were there, and only somebody who 259 00:17:05,880 --> 00:17:09,520 Speaker 1: had seen that statue before and therefore would have been 260 00:17:09,520 --> 00:17:13,600 Speaker 1: at the scene of the crime, would have that physiologic response. 261 00:17:14,040 --> 00:17:19,560 Speaker 1: You're looking for a familiarity response for something that only 262 00:17:19,600 --> 00:17:22,680 Speaker 1: a person who was there would know. That's the guilty 263 00:17:22,800 --> 00:17:26,960 Speaker 1: knowledge technique. It's also known as the concealed information test. 264 00:17:27,560 --> 00:17:29,800 Speaker 1: And keep in mind that your brain picks up all 265 00:17:29,920 --> 00:17:31,960 Speaker 1: kinds of things that you're not even aware of. So 266 00:17:32,000 --> 00:17:36,560 Speaker 1: it turns out that these familiarity responses can be big, 267 00:17:36,680 --> 00:17:39,679 Speaker 1: whether or not you know that you have seen that 268 00:17:39,720 --> 00:17:43,240 Speaker 1: thing before. So the polygraph test is used in both 269 00:17:43,280 --> 00:17:48,080 Speaker 1: these ways, judging stress or judging familiarity. So let's go 270 00:17:48,160 --> 00:17:51,840 Speaker 1: back to nineteen thirty two when a couple got married 271 00:17:52,000 --> 00:17:56,920 Speaker 1: in the Northwestern Crime Scene Laboratory under the polygraph test. 272 00:17:57,160 --> 00:17:59,920 Speaker 1: This couple got married and the newspapers at the time 273 00:18:00,119 --> 00:18:03,560 Speaker 1: reported that her heart almost stopped beating when she said 274 00:18:03,600 --> 00:18:06,440 Speaker 1: I do, and he twitched a little when he said 275 00:18:06,480 --> 00:18:08,800 Speaker 1: I do. And the papers reported that they thought this 276 00:18:08,880 --> 00:18:12,000 Speaker 1: all looked fine and good, but of course people didn't 277 00:18:12,160 --> 00:18:14,760 Speaker 1: typically get married while hooked up to this test. This 278 00:18:14,960 --> 00:18:17,879 Speaker 1: was a publicity stunt because the groom was one of 279 00:18:17,880 --> 00:18:21,040 Speaker 1: the founders of the polygraph test, and he had launched 280 00:18:21,240 --> 00:18:25,000 Speaker 1: a polygraph testing company and he was jazzed to get 281 00:18:25,040 --> 00:18:30,040 Speaker 1: this pr But let's dive into its usage. So it 282 00:18:30,119 --> 00:18:33,440 Speaker 1: was first used in a criminal case in nineteen twenty 283 00:18:33,480 --> 00:18:37,960 Speaker 1: three in Berkeley, California, after a retired police chief was murdered, 284 00:18:38,440 --> 00:18:41,640 Speaker 1: and then it started gaining traction in law enforcement all 285 00:18:41,680 --> 00:18:44,040 Speaker 1: through the twenties and thirties, and then by the time 286 00:18:44,080 --> 00:18:46,600 Speaker 1: of the Cold War, there was a lot of concern 287 00:18:46,680 --> 00:18:50,840 Speaker 1: about possible espionage, so a lot of agencies started to 288 00:18:50,960 --> 00:18:54,800 Speaker 1: use the polygraph test for security screenings, and by the 289 00:18:54,880 --> 00:18:59,560 Speaker 1: nineteen sixties, essentially all federal law enforcement agencies were leveraging 290 00:18:59,600 --> 00:19:03,640 Speaker 1: this test. But then by the nineteen seventies some academics 291 00:19:03,680 --> 00:19:07,800 Speaker 1: started raising questions about the accuracy of the polygraph test, 292 00:19:07,840 --> 00:19:11,000 Speaker 1: because it's quite good, but it's not perfect, and so 293 00:19:11,280 --> 00:19:15,000 Speaker 1: battles over its admissibility, and courts have raged since at 294 00:19:15,080 --> 00:19:17,520 Speaker 1: least that time. So I'll give you a sense of 295 00:19:17,560 --> 00:19:19,960 Speaker 1: how these battles can go. In nineteen ninety two, there 296 00:19:20,040 --> 00:19:23,360 Speaker 1: was a US Air Force man named Edward Scheffer who 297 00:19:23,560 --> 00:19:27,000 Speaker 1: applied to work at the Office of Special Investigations, and 298 00:19:27,080 --> 00:19:30,879 Speaker 1: they said, cool, but you have to take a urinalysis 299 00:19:31,080 --> 00:19:34,080 Speaker 1: and a polygraph test to make sure you're not on drugs. 300 00:19:34,400 --> 00:19:36,960 Speaker 1: So he said, great, and he did that. What happened 301 00:19:37,040 --> 00:19:42,720 Speaker 1: is that the uranalysis test came back positive for methamphetamine use. 302 00:19:43,240 --> 00:19:46,840 Speaker 1: But his polygraph test in which he said I have 303 00:19:46,960 --> 00:19:49,480 Speaker 1: not done any drugs since I joined the Air Force, 304 00:19:49,760 --> 00:19:53,199 Speaker 1: that came back and showed no evidence of deception, in 305 00:19:53,240 --> 00:19:57,080 Speaker 1: other words, showing he was not lying. Now, based just 306 00:19:57,160 --> 00:20:01,560 Speaker 1: on the uranalysis, he was tried by Marshal for using 307 00:20:01,640 --> 00:20:04,520 Speaker 1: drugs and at the court martial he said, yeah, I know, 308 00:20:04,600 --> 00:20:07,639 Speaker 1: my year analysis came back positive, but I would like 309 00:20:07,720 --> 00:20:10,720 Speaker 1: to submit my polygraph test as part of my defense 310 00:20:10,800 --> 00:20:14,320 Speaker 1: because that indicates I am not lying. And the court 311 00:20:14,359 --> 00:20:17,920 Speaker 1: martial judge said sorry, but as a rule, the Air 312 00:20:17,960 --> 00:20:22,520 Speaker 1: Force never accepts polygraph tests in the courtroom. And Scheffer said, 313 00:20:23,080 --> 00:20:26,720 Speaker 1: this violates my sixth Amendment rights, which say that I 314 00:20:26,760 --> 00:20:29,879 Speaker 1: should be able to submit evidence in my defense, and 315 00:20:29,880 --> 00:20:32,880 Speaker 1: you're not letting me do this. But the military court 316 00:20:32,880 --> 00:20:36,679 Speaker 1: wouldn't accept the polygraph and so Scheffer was convicted and 317 00:20:36,720 --> 00:20:39,680 Speaker 1: dishonorably discharged. And so he went to the Air Force 318 00:20:39,720 --> 00:20:42,160 Speaker 1: Court of Appeals and they said, yeah, no, we don't 319 00:20:42,160 --> 00:20:44,800 Speaker 1: have to take the polygraph test, even if it would 320 00:20:44,880 --> 00:20:47,880 Speaker 1: exculpate you. So then he went to the Armed Forces 321 00:20:47,920 --> 00:20:50,920 Speaker 1: Court of Appeals and they said, you know you're right, 322 00:20:51,000 --> 00:20:55,199 Speaker 1: this does seem to violate your sixth Amendment right to 323 00:20:55,400 --> 00:20:58,359 Speaker 1: present a defense. So this went all the way up 324 00:20:58,359 --> 00:21:01,679 Speaker 1: to the Supreme Court, and they had to struggle with 325 00:21:01,760 --> 00:21:05,840 Speaker 1: this because the polygraph test is controversial in many ways, 326 00:21:06,400 --> 00:21:08,560 Speaker 1: and so what they decided is that they were going 327 00:21:08,640 --> 00:21:11,639 Speaker 1: to side with the military on this. The military courts, 328 00:21:11,680 --> 00:21:15,200 Speaker 1: they said, do not have to take polygraph evidence if 329 00:21:15,240 --> 00:21:18,560 Speaker 1: they don't want to. They said, we need to preserve 330 00:21:19,200 --> 00:21:26,199 Speaker 1: the jury's core function of making credibility determinations in criminal trials. 331 00:21:26,240 --> 00:21:29,040 Speaker 1: In other words, they were saying, it should be about 332 00:21:29,080 --> 00:21:32,800 Speaker 1: the person in the jury box making a decision, not 333 00:21:32,960 --> 00:21:38,280 Speaker 1: about some machine or technology that says this is the truth. 334 00:21:38,920 --> 00:21:42,560 Speaker 1: Now did their decision have something to do with the 335 00:21:42,880 --> 00:21:47,240 Speaker 1: accuracy of the polygraph? Maybe? It turns out the polygraph 336 00:21:47,280 --> 00:21:51,400 Speaker 1: test is actually quite good. In tests by the American 337 00:21:51,440 --> 00:21:55,600 Speaker 1: Polygraph Association, they find it's about eighty seven percent accurate 338 00:21:55,840 --> 00:21:59,719 Speaker 1: in determining whether someone is telling the truth or telling 339 00:21:59,760 --> 00:22:03,159 Speaker 1: a although some scientists claim it's lower, maybe as low 340 00:22:03,200 --> 00:22:06,320 Speaker 1: as seventy five percent. The thing to note is that's 341 00:22:06,359 --> 00:22:10,439 Speaker 1: not bad, but some critics argue that you just can't 342 00:22:10,480 --> 00:22:14,199 Speaker 1: accept and errorrate that high if somebody is guilt or 343 00:22:14,280 --> 00:22:18,480 Speaker 1: innocence depends on this, and some courts argue that it's 344 00:22:18,480 --> 00:22:21,879 Speaker 1: not just a matter of whether the accuracy is good enough. 345 00:22:22,359 --> 00:22:26,000 Speaker 1: Part of the concern expressed by courts is whether jurors 346 00:22:26,560 --> 00:22:30,840 Speaker 1: understand probabilities well enough. In other words, if a defendant 347 00:22:31,280 --> 00:22:35,440 Speaker 1: fails the lie detector test, will a juror think, Okay, well, 348 00:22:35,520 --> 00:22:38,560 Speaker 1: it's not one hundred percent accuracy. There's still a thirteen 349 00:22:38,600 --> 00:22:41,359 Speaker 1: to twenty five percent chance that the machine got it wrong, 350 00:22:41,640 --> 00:22:44,440 Speaker 1: and so here's how I'm going to work those probabilities 351 00:22:44,480 --> 00:22:49,359 Speaker 1: into my decision. Many courts assumed, probably correctly, that the 352 00:22:49,440 --> 00:22:53,639 Speaker 1: technology will sway people more than it should. They will 353 00:22:53,680 --> 00:22:58,679 Speaker 1: believe whatever the machine says. And beyond the issue of 354 00:22:58,800 --> 00:23:02,080 Speaker 1: jury interpretation, there's another problem as well, which is that 355 00:23:02,440 --> 00:23:07,320 Speaker 1: some people are pathological liars. Some people, for example, those 356 00:23:07,359 --> 00:23:12,639 Speaker 1: with psychopathy, have real deficits in empathy, and they're not 357 00:23:12,840 --> 00:23:15,960 Speaker 1: good predictors of future punishment, and they lack what we 358 00:23:16,040 --> 00:23:20,240 Speaker 1: might poetically call a conscience, and so for them, lying 359 00:23:20,680 --> 00:23:24,400 Speaker 1: just doesn't trigger a stress response. Many of the studies 360 00:23:24,440 --> 00:23:28,720 Speaker 1: on lie detection are done with undergraduates or otherwise normal people, 361 00:23:29,000 --> 00:23:32,760 Speaker 1: but the people we really care about most are psychopaths, 362 00:23:32,840 --> 00:23:49,840 Speaker 1: and the technology doesn't always work so well with them. 363 00:23:52,720 --> 00:23:57,080 Speaker 1: So as it stands now, the polygraph test is generally 364 00:23:57,200 --> 00:23:59,960 Speaker 1: not used at the federal level in the United States 365 00:24:00,600 --> 00:24:04,199 Speaker 1: unless the prosecution and defense both agree to allow it, 366 00:24:04,680 --> 00:24:08,320 Speaker 1: and at the state level it varies widely. Essentially one 367 00:24:08,480 --> 00:24:12,760 Speaker 1: half of the states allow polygraph results while the other 368 00:24:12,840 --> 00:24:17,800 Speaker 1: half generally prohibits them. But the polygraph test, which has 369 00:24:17,840 --> 00:24:21,040 Speaker 1: been around for over a century now, is just one 370 00:24:21,160 --> 00:24:25,760 Speaker 1: method that people use, and as science cranks forward, there 371 00:24:25,760 --> 00:24:29,360 Speaker 1: have been many newer proposals. Actually a lot of these 372 00:24:29,400 --> 00:24:33,040 Speaker 1: ideas sprung up after nine to eleven when people wondered 373 00:24:33,240 --> 00:24:36,600 Speaker 1: how can you do lie detection on big groups of 374 00:24:36,640 --> 00:24:41,320 Speaker 1: people quickly. So one of these techniques is thermal imaging. 375 00:24:41,359 --> 00:24:44,679 Speaker 1: There's some evidence that when you're lying, you get hotter 376 00:24:44,880 --> 00:24:47,399 Speaker 1: under the orbits of your eyes. So the idea is 377 00:24:47,400 --> 00:24:51,560 Speaker 1: to use cameras at airports to do infrared imaging as 378 00:24:51,600 --> 00:24:55,520 Speaker 1: people are passing through and answering questions. And people have 379 00:24:55,600 --> 00:24:59,320 Speaker 1: also been looking at the subtle detection of stress in 380 00:24:59,640 --> 00:25:03,680 Speaker 1: the voi, the way that your voice gets constricted when 381 00:25:03,720 --> 00:25:08,040 Speaker 1: you have stress overlying about something. But things get a 382 00:25:08,080 --> 00:25:12,480 Speaker 1: little more interesting and spicy when we start looking directly 383 00:25:12,640 --> 00:25:16,280 Speaker 1: at signals of the brain, and one of these technologies 384 00:25:16,680 --> 00:25:20,159 Speaker 1: involves sticking electrodes on the outside of the scalp, and 385 00:25:20,200 --> 00:25:24,119 Speaker 1: this is known as electron cepholography or EEG. And the 386 00:25:24,320 --> 00:25:27,960 Speaker 1: idea here is to use things like the guilty knowledge 387 00:25:28,080 --> 00:25:32,400 Speaker 1: test to see if there's a big brain response when 388 00:25:32,480 --> 00:25:36,280 Speaker 1: the person sees something they are familiar with. So let's 389 00:25:36,320 --> 00:25:38,960 Speaker 1: go back to that story that I began with about 390 00:25:39,040 --> 00:25:43,960 Speaker 1: Aditi Sharma and her boyfriend Pravine, who together killed her fiance. 391 00:25:45,080 --> 00:25:47,960 Speaker 1: In Aditi's court trial in two thousand and eight, she 392 00:25:48,119 --> 00:25:50,840 Speaker 1: was tested with a thirty two electrode system on her 393 00:25:50,880 --> 00:25:55,399 Speaker 1: head called the Brain Electrical Oscillations Signature system. And the 394 00:25:55,440 --> 00:25:59,119 Speaker 1: way it worked is this, a forensic scientist had a 395 00:25:59,200 --> 00:26:03,520 Speaker 1: dit listen to a voice talk about the alleged murder 396 00:26:04,040 --> 00:26:08,320 Speaker 1: and when she heard specific incriminating bits of the story 397 00:26:08,359 --> 00:26:12,200 Speaker 1: about how the murder was committed. For example, he was 398 00:26:12,280 --> 00:26:15,000 Speaker 1: killed with a special kind of drink in McDonald's that 399 00:26:15,080 --> 00:26:17,919 Speaker 1: was poisoned with arsenic When she heard those bits of 400 00:26:17,920 --> 00:26:22,840 Speaker 1: the story, woosh, her brain had larger activity, which was 401 00:26:22,880 --> 00:26:26,879 Speaker 1: interpreted as the activation of memories that she had of 402 00:26:27,000 --> 00:26:31,399 Speaker 1: concealed knowledge, so this was presented as evidence of her 403 00:26:31,480 --> 00:26:34,800 Speaker 1: guilt in this case. Now, I think all legal scholars 404 00:26:34,840 --> 00:26:37,280 Speaker 1: agree that this was not the only evidence on which 405 00:26:37,320 --> 00:26:40,800 Speaker 1: her guilty verdict was based. There were other pieces of 406 00:26:40,800 --> 00:26:44,680 Speaker 1: evidence that pointed to her guilt. But nonetheless, this kindled 407 00:26:44,680 --> 00:26:48,760 Speaker 1: an ongoing controversy about when and whether it makes sense 408 00:26:48,840 --> 00:26:52,800 Speaker 1: to accept lie detector technology in a court of law, 409 00:26:53,280 --> 00:26:56,520 Speaker 1: because this was one of the first cases, possibly the 410 00:26:56,600 --> 00:27:00,760 Speaker 1: very first, where someone was convicted in part on brain 411 00:27:01,080 --> 00:27:05,520 Speaker 1: based lie detection. But this really was just the beginning, 412 00:27:06,119 --> 00:27:09,280 Speaker 1: because starting over two decades ago, people got interested in 413 00:27:09,280 --> 00:27:13,040 Speaker 1: the idea of whether you could detect a lie in 414 00:27:13,119 --> 00:27:16,560 Speaker 1: a brain scanner, and in two thousand and six at 415 00:27:16,640 --> 00:27:19,920 Speaker 1: least two companies launched with this goal. There was no 416 00:27:20,080 --> 00:27:24,720 Speaker 1: Lie MRI in San Diego and Cephos Corporation in Massachusetts, 417 00:27:24,760 --> 00:27:30,120 Speaker 1: and their mission was using brain scanning fMRI or functional 418 00:27:30,160 --> 00:27:34,520 Speaker 1: magnetic resonance imaging for the purpose of lie detection. Now, 419 00:27:34,720 --> 00:27:37,679 Speaker 1: how do these companies come about, Well, they were based 420 00:27:37,680 --> 00:27:41,720 Speaker 1: on the discovery that when you are lying, you can 421 00:27:41,840 --> 00:27:46,720 Speaker 1: image in the scanner more activity in frontal brain areas 422 00:27:46,840 --> 00:27:50,000 Speaker 1: just behind the forehead, more activity than when you are 423 00:27:50,040 --> 00:27:53,359 Speaker 1: telling the truth. In other words, while it's easy to 424 00:27:53,920 --> 00:27:59,080 Speaker 1: blurt out the truth, lying requires more effort. You are 425 00:27:59,160 --> 00:28:01,560 Speaker 1: inhibiting telling the truth. You don't want to blurt out 426 00:28:01,600 --> 00:28:05,560 Speaker 1: the truth, and you have to generate falsified data. Remember 427 00:28:05,600 --> 00:28:07,879 Speaker 1: I mentioned earlier that it's easier to tell the truth, 428 00:28:07,920 --> 00:28:13,120 Speaker 1: and we see that reflected in brain activity. So back 429 00:28:13,119 --> 00:28:16,359 Speaker 1: in two thousand and one, a researcher named Sean Spence 430 00:28:16,400 --> 00:28:21,440 Speaker 1: and his colleagues interviewed participants on what activities they had 431 00:28:21,480 --> 00:28:25,000 Speaker 1: done the previous day. Then in the scanner fMRI, these 432 00:28:25,000 --> 00:28:28,920 Speaker 1: people were presented with various activities and were asked whether 433 00:28:28,960 --> 00:28:30,600 Speaker 1: they had done them, and they had two buttons, a 434 00:28:30,720 --> 00:28:34,480 Speaker 1: yes and a no. Now, they were also given either 435 00:28:34,600 --> 00:28:37,920 Speaker 1: a red light or a green light to tell them 436 00:28:38,200 --> 00:28:42,400 Speaker 1: whether they should lie or not on this round. So 437 00:28:42,440 --> 00:28:44,680 Speaker 1: the first thing the researchers noted is that it always 438 00:28:44,720 --> 00:28:47,400 Speaker 1: takes longer to lie, about a fifth of a second longer. 439 00:28:47,680 --> 00:28:50,040 Speaker 1: So you went to the beach yesterday, and you're asked 440 00:28:50,080 --> 00:28:51,720 Speaker 1: if you went to the beach, But you see that 441 00:28:51,760 --> 00:28:55,160 Speaker 1: there's a red light and so you answer no. Now, 442 00:28:55,200 --> 00:28:57,760 Speaker 1: the first thing the researchers noted is that it always 443 00:28:57,840 --> 00:29:01,280 Speaker 1: takes longer to lie, about a fifth of a second longer, 444 00:29:01,680 --> 00:29:04,080 Speaker 1: And this is consistent with the idea that you are 445 00:29:04,480 --> 00:29:07,320 Speaker 1: taking an answer you already know and you're squelching that. 446 00:29:07,840 --> 00:29:11,920 Speaker 1: So they demonstrated that this squelching of the truth response 447 00:29:12,520 --> 00:29:16,600 Speaker 1: activates an area called the ventralateral prefrontal cortex. This is 448 00:29:16,640 --> 00:29:20,480 Speaker 1: a region that becomes active when you need to suppress 449 00:29:20,600 --> 00:29:23,520 Speaker 1: a behavior that you would normally just go with, in 450 00:29:23,560 --> 00:29:27,080 Speaker 1: this case telling the truth, and other brain areas are 451 00:29:27,080 --> 00:29:30,560 Speaker 1: involved too. So there was a study Bilangualmen in colleagues 452 00:29:30,600 --> 00:29:34,520 Speaker 1: the next year where participants went into the scanner and 453 00:29:34,600 --> 00:29:37,040 Speaker 1: they saw a playing card on the screen and they 454 00:29:37,040 --> 00:29:40,080 Speaker 1: were asked to remember it like Ace of Spades. Then 455 00:29:40,120 --> 00:29:42,920 Speaker 1: they were shown a series of cards and they were asked, 456 00:29:43,640 --> 00:29:46,240 Speaker 1: was that your card or not? They were instructed to 457 00:29:46,440 --> 00:29:50,880 Speaker 1: lie sometimes, and here's the key. When they lied, the 458 00:29:50,960 --> 00:29:54,360 Speaker 1: researcher saw a lot of activity in an area called 459 00:29:54,360 --> 00:29:58,320 Speaker 1: the anterior singular cortex, which is a region that becomes 460 00:29:58,440 --> 00:30:03,239 Speaker 1: active when there's con between other brain areas. So in 461 00:30:03,240 --> 00:30:05,800 Speaker 1: both of those studies, people were just saying yes or no. 462 00:30:06,600 --> 00:30:09,880 Speaker 1: But in the next study by Spence and colleagues, they 463 00:30:09,920 --> 00:30:13,600 Speaker 1: wanted to see what happens when people get more imaginative, 464 00:30:13,600 --> 00:30:16,000 Speaker 1: when they go beyond yes or no to make up 465 00:30:16,040 --> 00:30:19,440 Speaker 1: a new story. So imagine I ask you where were 466 00:30:19,480 --> 00:30:22,560 Speaker 1: you on the afternoon of March first, and you know 467 00:30:22,680 --> 00:30:25,280 Speaker 1: the answer to that, But imagine that you lie to 468 00:30:25,320 --> 00:30:27,960 Speaker 1: me and you say, oh, I left my cell phone 469 00:30:27,960 --> 00:30:30,040 Speaker 1: in the car and I was just hiking by myself 470 00:30:30,120 --> 00:30:33,400 Speaker 1: up the mountain. Now, in order to do that lie, 471 00:30:33,520 --> 00:30:37,040 Speaker 1: you have to think of the true response and suppress that, 472 00:30:37,480 --> 00:30:40,760 Speaker 1: and then you have to cook something up. So when 473 00:30:40,800 --> 00:30:44,320 Speaker 1: you are thinking of the truth, but you actively suppress that, 474 00:30:44,320 --> 00:30:48,280 Speaker 1: that activates the ventralateral prefrontal cortex, as we just saw. 475 00:30:48,840 --> 00:30:51,840 Speaker 1: But then you need to generate the lie, and this 476 00:30:51,920 --> 00:30:56,560 Speaker 1: activates the dorsal lateral prefrontal cortex, which is an area 477 00:30:56,600 --> 00:31:00,200 Speaker 1: that cranks up when a person generates something new, like 478 00:31:00,240 --> 00:31:03,440 Speaker 1: a new response that they hadn't done before. And we 479 00:31:03,480 --> 00:31:06,720 Speaker 1: see other areas too, like the anterior singular cortex, which 480 00:31:06,760 --> 00:31:11,640 Speaker 1: cares about conflict, and the venturemedial prefrontal cortex, which cranks 481 00:31:11,720 --> 00:31:14,720 Speaker 1: up when you're trying to regulate your emotions. So there 482 00:31:14,720 --> 00:31:21,280 Speaker 1: are a whole constellation of brain areas that indicate deception. Now, interestingly, 483 00:31:21,360 --> 00:31:24,720 Speaker 1: in all these studies about lying, you never see areas 484 00:31:24,760 --> 00:31:28,000 Speaker 1: where there's more activity when you tell the truth. There's 485 00:31:28,080 --> 00:31:31,959 Speaker 1: only more activity when you're lying. So the idea that 486 00:31:32,000 --> 00:31:34,920 Speaker 1: it's easier to tell the truth holds as a rule 487 00:31:34,960 --> 00:31:38,080 Speaker 1: in life and as a rule in the brain. Now, 488 00:31:38,120 --> 00:31:43,200 Speaker 1: that's how you can detect something about deception using an fMRI, 489 00:31:43,960 --> 00:31:48,280 Speaker 1: and you can also use the guilty knowledge technique here. 490 00:31:48,720 --> 00:31:51,960 Speaker 1: So one research paper said, we're going to try a 491 00:31:51,960 --> 00:31:56,520 Speaker 1: guilty knowledge technique. We're going to rennew certain dates on 492 00:31:56,560 --> 00:31:59,000 Speaker 1: the calendar, and one of those dates is going to 493 00:31:59,040 --> 00:32:02,160 Speaker 1: be your birth birthday, and of course all the participants 494 00:32:02,200 --> 00:32:06,040 Speaker 1: have different birthdays, so it's a nicely controlled study. So 495 00:32:06,280 --> 00:32:10,040 Speaker 1: the murder weapon they're looking for is your birthday. So 496 00:32:10,120 --> 00:32:15,320 Speaker 1: they say March twelfth, November twenty third, August fifth, and 497 00:32:15,520 --> 00:32:19,440 Speaker 1: at some point they say your birthday and several eras 498 00:32:19,440 --> 00:32:21,760 Speaker 1: in your brain light up. You're hearing something that is 499 00:32:21,920 --> 00:32:25,160 Speaker 1: familiar to you. And generally this is the same thing 500 00:32:25,160 --> 00:32:28,880 Speaker 1: that happens when you see that Pokemon statue or when 501 00:32:28,920 --> 00:32:32,640 Speaker 1: you see the murder weapon that you used. Now, it 502 00:32:32,680 --> 00:32:35,600 Speaker 1: turns out that in this study, the investigators reported that 503 00:32:35,680 --> 00:32:38,360 Speaker 1: this worked one hundred percent of the time. In other words, 504 00:32:38,440 --> 00:32:40,800 Speaker 1: all the time they could tell when someone is hearing 505 00:32:41,200 --> 00:32:46,520 Speaker 1: their birthday versus hearing another date. But it turns out 506 00:32:46,840 --> 00:32:49,640 Speaker 1: that you can easily fake the test. You can fool 507 00:32:49,680 --> 00:32:53,720 Speaker 1: this entirely by saying in advance, Okay, I'm going to 508 00:32:53,800 --> 00:32:58,040 Speaker 1: associate another date with a small movement of my finger, 509 00:32:58,120 --> 00:33:00,440 Speaker 1: So you decide that before you go into the test, 510 00:33:00,480 --> 00:33:04,480 Speaker 1: and then when you hear August fifth, you were waiting 511 00:33:04,520 --> 00:33:06,959 Speaker 1: to hear that so you could move your finger. So 512 00:33:07,160 --> 00:33:09,800 Speaker 1: when you hear it, your brain has a big response 513 00:33:09,880 --> 00:33:13,040 Speaker 1: because you were waiting for that. What that means is 514 00:33:13,040 --> 00:33:16,600 Speaker 1: that you're having the same murder weapon response to some 515 00:33:16,800 --> 00:33:20,040 Speaker 1: arbitrary date or any number of arbitrary dates, because you 516 00:33:20,120 --> 00:33:22,719 Speaker 1: have pre planned that you were going to do that, 517 00:33:22,760 --> 00:33:26,320 Speaker 1: you were waiting to hear that target, and as a result, 518 00:33:26,360 --> 00:33:29,600 Speaker 1: everything else gets swamped out because you're having the same 519 00:33:29,680 --> 00:33:32,440 Speaker 1: response to some fake date as you are to the 520 00:33:32,480 --> 00:33:35,320 Speaker 1: real date, and you can't tell a difference between those 521 00:33:35,360 --> 00:33:38,640 Speaker 1: brain states. So it turns out that this is not 522 00:33:38,840 --> 00:33:41,800 Speaker 1: ready for the courts, because you can completely fool this 523 00:33:42,240 --> 00:33:45,720 Speaker 1: with a very simple countermeasure, and even with the other 524 00:33:45,880 --> 00:33:50,080 Speaker 1: fMRI technique that just involves looking for the correlates of deception. 525 00:33:50,760 --> 00:33:54,720 Speaker 1: It only works if you stay absolutely still in the scanner. 526 00:33:55,040 --> 00:33:57,440 Speaker 1: And even if someone were to fix your head into 527 00:33:57,480 --> 00:33:59,680 Speaker 1: place so that you couldn't move at all, you could, 528 00:33:59,720 --> 00:34:02,440 Speaker 1: of course do all kinds of things to think random 529 00:34:02,520 --> 00:34:06,160 Speaker 1: thoughts and move your fingers and almost certainly swamp out 530 00:34:06,520 --> 00:34:11,359 Speaker 1: any meaningful signals. So let's recap where we are. As 531 00:34:11,520 --> 00:34:14,840 Speaker 1: animals developed bigger brains, they were able to make better 532 00:34:14,960 --> 00:34:18,839 Speaker 1: models of one another, and that allowed them to cooperate, 533 00:34:18,880 --> 00:34:22,600 Speaker 1: but it also allowed them to deceive. And so throughout 534 00:34:22,680 --> 00:34:25,680 Speaker 1: human history people have been very interested in the question 535 00:34:25,719 --> 00:34:29,160 Speaker 1: of whether you can detect if a person is lying. 536 00:34:29,760 --> 00:34:32,799 Speaker 1: And this is tough because someone can be telling the 537 00:34:32,920 --> 00:34:36,480 Speaker 1: truth but they're factually incorrect. But if they believe their 538 00:34:36,520 --> 00:34:38,720 Speaker 1: story is true, then you would say, look, she passed 539 00:34:38,719 --> 00:34:42,680 Speaker 1: the polygraph test, she passed the brain scan lie detection test, 540 00:34:42,960 --> 00:34:46,359 Speaker 1: but it still doesn't tell you what actually happened. All 541 00:34:46,400 --> 00:34:49,120 Speaker 1: it's telling you is that she believes that would happened. 542 00:34:49,640 --> 00:34:53,400 Speaker 1: But sometimes people really do set out to deceive, and 543 00:34:53,640 --> 00:34:56,880 Speaker 1: that's what people want to build technologies for. And we 544 00:34:57,000 --> 00:35:02,560 Speaker 1: find this search for detecting deception all over. Sometimes this 545 00:35:02,640 --> 00:35:06,320 Speaker 1: is in arbitration situations where it's one person's word against 546 00:35:06,360 --> 00:35:08,320 Speaker 1: another and you want to know who's telling the truth. 547 00:35:08,600 --> 00:35:11,920 Speaker 1: And you find this in criminal cases where attorneys want 548 00:35:11,960 --> 00:35:14,719 Speaker 1: to know if someone has in fact broken the law. 549 00:35:15,000 --> 00:35:18,319 Speaker 1: And you find this in national security where governments want 550 00:35:18,400 --> 00:35:21,480 Speaker 1: to know when some spy that they've caught is telling 551 00:35:21,520 --> 00:35:25,800 Speaker 1: the truth or not. They want to have a meaningful interrogation. Now, 552 00:35:26,160 --> 00:35:28,680 Speaker 1: as a reminder of what we talked about, there's no 553 00:35:28,760 --> 00:35:32,120 Speaker 1: way to measure a lie directly. All we ever measure 554 00:35:32,719 --> 00:35:36,879 Speaker 1: is the physiology that's associated with the lie, whether that's 555 00:35:37,040 --> 00:35:40,399 Speaker 1: a stress response like we see in the polygraph test, 556 00:35:40,719 --> 00:35:43,719 Speaker 1: or whether it's about inhibiting the truth and cooking up 557 00:35:43,719 --> 00:35:47,239 Speaker 1: a fake story like we see in brain imaging. But 558 00:35:47,320 --> 00:35:50,279 Speaker 1: the question is will it work? Will we ever get 559 00:35:50,320 --> 00:35:53,520 Speaker 1: to a point where we can know for sure when 560 00:35:53,640 --> 00:35:57,520 Speaker 1: somebody is lying. Essentially, the debate about this falls into 561 00:35:57,560 --> 00:36:02,400 Speaker 1: two camps. The first is that this technology simply is 562 00:36:02,640 --> 00:36:06,600 Speaker 1: never going to work perfectly in the real world because 563 00:36:06,640 --> 00:36:09,520 Speaker 1: it's always facing a mountain of questions. What if the 564 00:36:09,920 --> 00:36:12,520 Speaker 1: person that you're measuring is on some sort of medication, 565 00:36:12,680 --> 00:36:16,399 Speaker 1: What if he's had a stroke. Does the technology work 566 00:36:16,440 --> 00:36:19,759 Speaker 1: for people who are outside the narrow age range that's 567 00:36:19,800 --> 00:36:24,000 Speaker 1: been studied so far. What about psychopaths and compulsive liars. 568 00:36:24,320 --> 00:36:27,319 Speaker 1: What if a person is totally misremembered something but they're 569 00:36:27,320 --> 00:36:30,719 Speaker 1: not actually lying, and on on. So the concern that 570 00:36:30,760 --> 00:36:35,360 Speaker 1: many scholars express is that these technologies, both new and old, 571 00:36:35,560 --> 00:36:39,200 Speaker 1: will appear in courts before all the wrinkles have been 572 00:36:39,239 --> 00:36:42,000 Speaker 1: ironed out. And this is a concern for many reasons. 573 00:36:42,000 --> 00:36:45,480 Speaker 1: But when it comes to fMRI, one of the concerns 574 00:36:45,800 --> 00:36:49,960 Speaker 1: is that images of the brain are very compelling to jurors. 575 00:36:50,120 --> 00:36:53,239 Speaker 1: It's different if you just say, hey, on this polygraph test, 576 00:36:53,280 --> 00:36:56,880 Speaker 1: there was a blip, versus you show a beautiful false 577 00:36:56,920 --> 00:37:00,200 Speaker 1: color image of the brain, and then a juror thinks, wow, wow, 578 00:37:00,200 --> 00:37:04,160 Speaker 1: that looks like real science, even if the accuracy probability 579 00:37:04,200 --> 00:37:07,960 Speaker 1: is the same. So that's the first concern is that 580 00:37:08,040 --> 00:37:10,200 Speaker 1: this technology is not actually going to work in the 581 00:37:10,239 --> 00:37:14,480 Speaker 1: real world. The second concern is that it will work, 582 00:37:15,200 --> 00:37:17,400 Speaker 1: and this is a concern because most people hold the 583 00:37:17,440 --> 00:37:22,080 Speaker 1: intuition that the inside of your head should be inviolable. 584 00:37:22,480 --> 00:37:25,839 Speaker 1: Your mind should be a sanctuary where no one can 585 00:37:26,040 --> 00:37:29,600 Speaker 1: enter so given that even if the technology does work 586 00:37:29,680 --> 00:37:33,719 Speaker 1: perfectly someday, there are several reasons why it might never 587 00:37:33,800 --> 00:37:36,600 Speaker 1: get used in the legal system. I mentioned the Seffer 588 00:37:36,680 --> 00:37:38,920 Speaker 1: case in nineteen ninety eight when the Supreme Court was 589 00:37:38,920 --> 00:37:42,200 Speaker 1: deciding on whether a lower court has to accept the 590 00:37:42,239 --> 00:37:45,440 Speaker 1: polygraph test, and at that time four of the justices 591 00:37:45,680 --> 00:37:50,080 Speaker 1: suggested that even if polygraph tests were to work flawlessly 592 00:37:50,160 --> 00:37:54,880 Speaker 1: with one hundred percent reliability, they still shouldn't be admitted. 593 00:37:55,320 --> 00:37:59,080 Speaker 1: Why because the justice is said that would take away 594 00:37:59,520 --> 00:38:03,440 Speaker 1: the job of the jury as the determiners of truth. 595 00:38:03,920 --> 00:38:06,239 Speaker 1: Now that's not law, it's only a legal opinion at 596 00:38:06,239 --> 00:38:09,160 Speaker 1: this stage, but it could be a pattern for how 597 00:38:09,480 --> 00:38:13,080 Speaker 1: fancy brain imaging technologies might also get ruled on in 598 00:38:13,120 --> 00:38:17,319 Speaker 1: the future. And again, unlike a polygraph as it stands now, 599 00:38:17,440 --> 00:38:22,240 Speaker 1: the fMRI technology won't work unless you are a willing subject, 600 00:38:22,280 --> 00:38:25,120 Speaker 1: willing to keep your head very still for the images 601 00:38:25,160 --> 00:38:28,440 Speaker 1: to come out right. And in this sense, the current 602 00:38:28,640 --> 00:38:33,400 Speaker 1: brain imaging technology may become more useful as a truth 603 00:38:33,560 --> 00:38:37,960 Speaker 1: confirmer by people who feel they've been falsely accused, rather 604 00:38:38,080 --> 00:38:41,800 Speaker 1: than a lie detector. So let's return to this issue 605 00:38:41,880 --> 00:38:45,759 Speaker 1: about privacy and whether lie detection as it gets better 606 00:38:45,880 --> 00:38:50,520 Speaker 1: over the coming centuries, constitutes a violation. One of the 607 00:38:50,560 --> 00:38:53,200 Speaker 1: deepest questions from the point of view of the legal 608 00:38:53,200 --> 00:38:56,400 Speaker 1: system in the Constitution is that a criminal defendant always 609 00:38:56,480 --> 00:39:01,840 Speaker 1: retains the right not to testify and not to incriminate himself. 610 00:39:02,239 --> 00:39:06,600 Speaker 1: So can a brain scan lie detection test be forced 611 00:39:06,920 --> 00:39:09,080 Speaker 1: or does it need to be agreed to? If you 612 00:39:09,120 --> 00:39:12,239 Speaker 1: were to refuse to use a new brain scan technology, 613 00:39:12,480 --> 00:39:16,120 Speaker 1: are you protected against self incrimination which is the fifth 614 00:39:16,120 --> 00:39:19,440 Speaker 1: Amendment of the Constitution. There's a lot of legal precedent 615 00:39:19,440 --> 00:39:21,440 Speaker 1: to all these questions, but of course we don't know 616 00:39:21,520 --> 00:39:25,000 Speaker 1: exactly how this will evolve in the coming decades. Currently, 617 00:39:25,680 --> 00:39:29,920 Speaker 1: blood tests and DNA tests can be forced. Many counties 618 00:39:29,960 --> 00:39:34,600 Speaker 1: have a no refusal policy. If you're driving through and 619 00:39:34,640 --> 00:39:36,799 Speaker 1: they suspect you of being on drugs, they can draw 620 00:39:36,840 --> 00:39:40,959 Speaker 1: your blood. In other circumstances, law enforcement gets a court 621 00:39:41,080 --> 00:39:43,759 Speaker 1: order for your DNA, which is usually taken with a 622 00:39:43,840 --> 00:39:46,879 Speaker 1: cheek swab. Now, these tests can only be done if 623 00:39:46,880 --> 00:39:50,120 Speaker 1: they are very relevant to the case, if they're critical 624 00:39:50,160 --> 00:39:55,000 Speaker 1: to the outcome. So will brain scan lie detection tests 625 00:39:55,040 --> 00:39:58,719 Speaker 1: ever reach the point where they can be forced, Well, 626 00:39:58,760 --> 00:40:01,960 Speaker 1: it's difficult to say, but as a piece of legal 627 00:40:02,080 --> 00:40:05,240 Speaker 1: precedent that might be applicable. Let's go back to nineteen 628 00:40:05,320 --> 00:40:09,240 Speaker 1: fifty two. There was a man in California named Rochin, 629 00:40:09,719 --> 00:40:12,960 Speaker 1: and the police raided his home on suspicion that he 630 00:40:13,040 --> 00:40:16,480 Speaker 1: had narcotics. And when they busted in, he went running 631 00:40:16,560 --> 00:40:18,800 Speaker 1: up the stairs and they chased him up the stairs 632 00:40:19,120 --> 00:40:21,719 Speaker 1: and he got to his bedroom and there were some 633 00:40:21,880 --> 00:40:24,640 Speaker 1: pills on the dresser and they said what is that 634 00:40:25,040 --> 00:40:28,840 Speaker 1: and he popped the pills into his mouth and swallowed them. 635 00:40:29,120 --> 00:40:31,480 Speaker 1: So this made the police mad because he had just 636 00:40:31,520 --> 00:40:34,319 Speaker 1: gotten rid of the evidence. So they wrestled him to 637 00:40:34,400 --> 00:40:36,680 Speaker 1: try to get the pills out, and when that didn't work, 638 00:40:36,920 --> 00:40:39,319 Speaker 1: they handcuffed him and they took him down to the 639 00:40:39,320 --> 00:40:42,879 Speaker 1: emergency room and had his stomach pumped to get the 640 00:40:42,920 --> 00:40:47,120 Speaker 1: evidence back out. And this case spun all the way 641 00:40:47,120 --> 00:40:50,920 Speaker 1: to the Supreme Court, and the court concluded, you can't 642 00:40:50,960 --> 00:40:53,960 Speaker 1: do that. You can't pump somebody's stomach to get the 643 00:40:54,040 --> 00:40:58,200 Speaker 1: evidence out of them. Why because they said that is 644 00:40:58,520 --> 00:41:02,719 Speaker 1: unreasonable search seizure, and they said it shocks the conscience. 645 00:41:03,400 --> 00:41:06,080 Speaker 1: So given this legal background, it's not clear that you 646 00:41:06,080 --> 00:41:09,720 Speaker 1: could be forced into giving a brain scan and whether 647 00:41:09,760 --> 00:41:13,719 Speaker 1: that would count as a form of unreasonable search and seizure. 648 00:41:13,960 --> 00:41:16,360 Speaker 1: Would it be like pumping out the contents of the 649 00:41:16,440 --> 00:41:20,720 Speaker 1: stomach if you looked at the contents of your mental life. 650 00:41:20,800 --> 00:41:23,120 Speaker 1: And what if you're terrified of a brain scanner because 651 00:41:23,120 --> 00:41:25,640 Speaker 1: you have claustrophobia or you don't like loud noise. Could 652 00:41:25,680 --> 00:41:28,279 Speaker 1: you say that you're being put under undue pressure and 653 00:41:28,320 --> 00:41:31,799 Speaker 1: it's a form of torture. So these are all open questions. 654 00:41:32,320 --> 00:41:35,920 Speaker 1: Can a legal system in the future get a mental 655 00:41:36,360 --> 00:41:41,480 Speaker 1: search warrant? Does that seem reasonable under some circumstances? Or 656 00:41:42,239 --> 00:41:45,680 Speaker 1: is there something special about this inner sanctum which has 657 00:41:45,719 --> 00:41:51,200 Speaker 1: been absolutely private for all of history? Can we and 658 00:41:51,320 --> 00:41:55,880 Speaker 1: should we maintain that expectation. How all of this is 659 00:41:55,920 --> 00:41:58,480 Speaker 1: going to shake out at the intersection of neuroscience and 660 00:41:58,520 --> 00:42:01,520 Speaker 1: the legal system is It's difficult to say at the moment, 661 00:42:02,040 --> 00:42:06,719 Speaker 1: but generally, as our technology moves faster and faster and 662 00:42:06,840 --> 00:42:12,759 Speaker 1: surpasses our stone age brains, we find ourselves entering a 663 00:42:13,080 --> 00:42:20,960 Speaker 1: very strange new world, and that is no lie. Go 664 00:42:21,040 --> 00:42:24,560 Speaker 1: to Eagleman dot com slash podcast for more information and 665 00:42:24,640 --> 00:42:27,799 Speaker 1: to find further reading. Always feel free to send me 666 00:42:27,840 --> 00:42:32,080 Speaker 1: an email at podcasts at eagleman dot com with any questions, 667 00:42:32,280 --> 00:42:35,719 Speaker 1: and check out and subscribe to Inner Cosmos on YouTube 668 00:42:35,840 --> 00:42:39,480 Speaker 1: for videos of each episode and to leave comments. Until 669 00:42:39,520 --> 00:42:43,560 Speaker 1: next time. I'm David Eagleman and this is Inner Cosmos