WEBVTT 00:00.660 --> 00:04.430 PROFESSOR: So what I would like you to do is to answer 00:04.426 --> 00:06.576 the following question. 00:06.580 --> 00:10.040 "Do you have..?" oh, now wait a second -- this 00:10.035 --> 00:11.875 says polling closed. 00:11.880 --> 00:14.180 Let's see whether we can get polling to open. 00:14.181 --> 00:16.091 Do you...?" -No, that didn't work. 00:16.090 --> 00:18.250 Oh, this is the most desperately horrible thing 00:18.250 --> 00:19.550 that I've ever had happen. 00:19.550 --> 00:20.360 OK. 00:20.360 --> 00:21.990 Current slide. 00:21.990 --> 00:23.240 I've had horrible things happen that are worse than 00:23.235 --> 00:25.115 this, I'm sorry. 00:25.118 --> 00:27.588 I don't want to belittle all the really horrible things 00:27.586 --> 00:28.496 that have ever happened to me. 00:28.504 --> 00:29.074 [laughter] 00:29.070 --> 00:29.780 OK. 00:29.780 --> 00:31.850 So let's try again. 00:31.850 --> 00:33.500 Click practicalities blah blah. 00:33.500 --> 00:36.060 OK. "Do you have a clicker yet? 00:36.060 --> 00:39.460 Polling is open." OK. 00:39.460 --> 00:41.510 And there's ten seconds left. 00:41.512 --> 00:43.122 You'll see this little timeline. 00:43.123 --> 00:46.773 Eight, seven, six, five, four-- ninety two of you, so 00:46.765 --> 00:48.525 many of you have clickers. 00:48.530 --> 00:48.740 OK. 00:48.740 --> 00:51.810 Let's see what we learned. 00:51.810 --> 00:57.310 It looks like 97% of you have clickers. 00:57.313 --> 00:59.363 [laughter] 00:59.360 --> 01:00.800 Now! 01:00.800 --> 01:07.310 I actually began with this exercise to make a point about 01:07.310 --> 01:10.630 psychological research. 01:10.625 --> 01:19.055 We just made a classic and dangerous mistake, a mistake 01:19.060 --> 01:20.940 known as sampling bias. 01:23.600 --> 01:28.970 We used a measure which doesn't give us accurate 01:28.970 --> 01:31.390 information about what we wanted to find out. 01:31.390 --> 01:35.010 We wanted to find out, what percentage of people in the 01:35.007 --> 01:37.477 room had clickers. 01:37.480 --> 01:40.480 And what we found out instead was, what percentage of people 01:40.482 --> 01:43.612 in the room with clickers had clickers. 01:43.607 --> 01:45.307 [laughter] 01:45.305 --> 01:46.555 I don't know who you guys are. 01:49.200 --> 01:53.830 This error is an incredibly dangerous one, and it could 01:53.831 --> 01:55.001 have persisted. 01:55.004 --> 01:58.034 Suppose I then ask you a question whether you're from 01:58.030 --> 02:04.070 the class of 2014, '13, '12, or '11, and discovered that 02:04.072 --> 02:08.382 80% of you were freshmen and sophomores. 02:08.380 --> 02:12.330 There, then, I would inherited exactly the sampling bias 02:12.326 --> 02:15.896 error that we just observed. 02:15.900 --> 02:18.060 Freshmen and sophomores have handed in 02:18.063 --> 02:19.973 their schedules already. 02:19.970 --> 02:22.000 Consequently, they're certain about what 02:22.000 --> 02:23.330 classes they're taking. 02:23.330 --> 02:25.820 Consequently, they're more likely to 02:25.815 --> 02:28.615 have clickers already. 02:28.624 --> 02:34.994 So when we do psychological empirical research, we need to 02:34.990 --> 02:39.310 be extraordinarily careful that the means by which we are 02:39.308 --> 02:44.548 choosing subjects for experiments are, in fact, 02:44.550 --> 02:50.400 means that give us an accurate sample of what it is that 02:50.400 --> 02:52.840 we're interested in. 02:52.844 --> 02:56.944 An article appeared recently in the journal Behavioral and 02:56.940 --> 03:01.310 Brain Sciences with the title, "The Weirdest 03:01.310 --> 03:05.180 People in the World". 03:05.180 --> 03:13.080 And "WEIRD" here stands for Western, educated, 03:13.080 --> 03:17.150 industrialized, rich, and democratic. 03:17.154 --> 03:22.174 That is, American college students. 03:22.173 --> 03:28.023 People with access to online Internet polling. 03:28.020 --> 03:33.660 And what the article argued was that a great deal of 03:33.660 --> 03:39.870 psychological research that claims to make general 03:39.870 --> 03:42.720 assumptions, that claims to prove general assumptions 03:42.715 --> 03:50.245 about human nature, is biased in exactly the way my poll 03:50.250 --> 03:54.280 concerning what percentage of you have clickers was biased. 03:54.280 --> 04:00.490 It looks at a small sample of the population, those who are 04:00.490 --> 04:04.900 readily available to research laboratories on university 04:04.900 --> 04:10.880 campuses, for example, the psych introductory psych pool. 04:10.882 --> 04:14.502 And it bases its conclusions about human 04:14.500 --> 04:19.210 nature on that sample. 04:19.210 --> 04:23.680 Throughout the semester we need to be attentive, when we 04:23.676 --> 04:26.466 read psychological studies which are making claims about 04:26.470 --> 04:32.170 human nature, to the fact that it is possible that some of 04:32.170 --> 04:35.960 the things which we are told apply to human beings in 04:35.955 --> 04:42.745 general instead have at least being shown conclusively only 04:42.750 --> 04:46.580 to apply to WEIRD folk. 04:46.580 --> 04:49.390 Western, educated, industrialized, rich, and 04:49.386 --> 04:51.246 democratic. 04:51.250 --> 04:56.430 And in certain cases, the data would look very different if 04:56.430 --> 05:01.070 we looked at a different population. 05:01.066 --> 05:05.956 This is also the case in the classical texts we used. 05:05.962 --> 05:11.232 The process that Plato and Aristotle took on for 05:11.230 --> 05:14.720 themselves didn't look experimental in the way that 05:14.715 --> 05:17.765 experimental psychology looks experimental. 05:17.770 --> 05:22.880 But they looked at a sample of people around them, and on the 05:22.880 --> 05:27.410 basis of their experience, they drew conclusions about 05:27.410 --> 05:30.110 human nature. 05:30.110 --> 05:34.290 So little reminder of something that I said in the 05:34.285 --> 05:39.725 first lecture that sometimes gets lost. In some ways, this 05:39.728 --> 05:41.668 is a class about philosophy and the 05:41.673 --> 05:43.653 science of human nature. 05:43.645 --> 05:48.235 And in other ways, this is a class about Western philosophy 05:48.244 --> 05:52.024 and the science of Western human nature. 05:52.020 --> 05:58.490 And we need to be attentive to that when it becomes relevant. 05:58.494 --> 06:03.744 So what I want to do in today's lecture is to return 06:03.740 --> 06:08.820 to the issue that we took up last class, which is the issue 06:08.820 --> 06:13.930 about the multi-part nature of the human soul. 06:13.934 --> 06:16.944 And you'll recall that in the last lecture, we were 06:16.940 --> 06:22.460 introduced to Plato's great analogy of rational spirit and 06:22.464 --> 06:28.394 appetite, as exemplified by a charioteer, a calm horse and a 06:28.392 --> 06:29.612 wild horse. 06:29.610 --> 06:34.300 And also to a number of other distinctions. 06:34.300 --> 06:37.870 Left brain, right brain, which I managed to get wrong four 06:37.865 --> 06:41.645 times in a row, if I remember correctly. 06:41.650 --> 06:44.930 We were introduced to the difference between brainstem 06:44.930 --> 06:46.460 and upper brain. 06:46.460 --> 06:49.310 We were briefly introduced to Freud's idea of 06:49.310 --> 06:52.400 superego, ego, and id. 06:52.400 --> 06:57.580 But we didn't get to work in the dual processing tradition. 06:57.580 --> 07:01.290 And today's lecture will take up where the last lecture left 07:01.292 --> 07:05.942 off, with another way in which it is typical to distinguish 07:05.940 --> 07:07.830 parts of the soul. 07:07.830 --> 07:11.780 Let me say that there is very good reason to think that the 07:11.780 --> 07:16.530 research which I'm presenting today is not subject to the 07:16.530 --> 07:18.850 WEIRD objection. 07:18.854 --> 07:21.194 That is, there have been powerful cross-cultural 07:21.190 --> 07:24.460 demonstrations of nearly all the results which I'll be 07:24.462 --> 07:26.082 talking about today. 07:26.080 --> 07:29.930 And there are also good evolutionary reasons to think 07:29.930 --> 07:34.510 that the two systems which dual systems theorists posit 07:34.510 --> 07:39.100 are, in fact, going to be part of any human being because of 07:39.100 --> 07:43.800 the evolutionary process, which all of us underwent. 07:43.800 --> 07:48.550 So I wanted to start with a picture of Edward Thorndike 07:48.545 --> 07:51.445 not because he's important, but just because he's so 07:51.446 --> 07:53.396 fabulous-looking. 07:53.398 --> 07:57.438 This is from the New York Public Library Archive. 07:57.444 --> 08:02.974 Edward Thorndike was a late 19th, early 20th century 08:02.970 --> 08:07.640 psychologist who did a lot of important associationist 08:07.640 --> 08:09.670 animal work. 08:09.670 --> 08:15.510 But he also did research, publishing an article in 1922, 08:15.510 --> 08:21.610 that in some ways can be seen as the founding work of the 08:21.610 --> 08:25.510 reasoning tradition that Jonathan Evans described in 08:25.510 --> 08:29.410 the article that we read for last class. 08:29.405 --> 08:35.335 So he describes, in his 1922 article, an experiment that he 08:35.340 --> 08:39.940 does, which is titled "The Effect of Changed Data Upon 08:39.935 --> 08:41.135 Reasoning". 08:41.144 --> 08:45.444 And what he's interested in there is the question of 08:45.440 --> 08:50.180 whether problems that are posed to people that are 08:50.180 --> 08:54.910 formally identical, but that differ in how that formal 08:54.910 --> 09:00.140 material is presented, are processed differently. 09:00.140 --> 09:05.280 So, for example, he asked people, either what is the 09:05.275 --> 09:08.955 square of x plus y, (first, outer, inner, last, I think is 09:08.963 --> 09:15.133 how you would do it) and he asked people, what is the 09:15.126 --> 09:18.926 square of b1 plus b2? 09:18.930 --> 09:23.150 Whereas people found the first question easy, they found the 09:23.150 --> 09:25.480 second question much more difficult. 09:25.482 --> 09:28.892 Success rates on this were up around 90%. 09:28.892 --> 09:31.522 Success rates on this were considerably lower. 09:31.520 --> 09:36.080 Or he asked them, what's the square of a squared x cubed, 09:36.080 --> 09:40.150 versus what's the square of r subscript 1 to the 8th r 09:40.146 --> 09:42.776 subscript 11 to the second? 09:42.780 --> 09:45.020 And so on. 09:45.016 --> 09:48.906 Presenting people with problems that were formally 09:48.910 --> 09:53.250 identical but which differed in terms of the complexity of 09:53.250 --> 09:58.080 the characters used to represent them produced a 09:58.082 --> 10:00.552 massive decrement in performance. 10:03.202 --> 10:09.692 Fast forward to a period in the 1960s, '70s, and '80s, in 10:09.692 --> 10:14.962 which people began to study syllogistic reasoning, and a 10:14.959 --> 10:21.179 number of special instances of this phenomenon emerge. 10:21.180 --> 10:25.770 So in Jonathan Evans piece that you read for last class, 10:25.770 --> 10:29.050 you were presented with examples like this. 10:29.052 --> 10:35.782 A syllogism that was valid and believable - that is, whose 10:35.784 --> 10:39.104 formal properties guaranteed that if the premises were 10:39.102 --> 10:42.622 true, the conclusion was true, and as a matter of fact 10:42.624 --> 10:45.084 conclusion was true. 10:45.084 --> 10:49.944 And arguments that were valid but unbelievable--arguments 10:49.940 --> 10:52.710 where the structure of the argument guaranteed that if 10:52.706 --> 10:56.156 the premises were true, the conclusion was true, but where 10:56.160 --> 10:59.130 the conclusion was false. 10:59.130 --> 11:02.120 So, for example, you might be told: No Greek 11:02.120 --> 11:05.450 tragedies are comedies. 11:05.450 --> 11:09.190 Some Greek comedies are plays. 11:09.185 --> 11:12.755 And asked whether it followed from that that some Greek 11:12.760 --> 11:14.900 plays are not Greek tragedies. 11:14.900 --> 11:17.840 No Greek tragedies are comedies, some Greek plays are 11:17.838 --> 11:20.848 comedies, therefore, some Greek plays are not Greek 11:20.850 --> 11:25.560 tragedies, and 90% of people were able to see that that 11:25.560 --> 11:28.020 argument was valid. 11:28.020 --> 11:31.180 By contrast, you might be given an argument, equally 11:31.180 --> 11:35.180 valid, but with an unbelievable conclusion; like: 11:35.180 --> 11:37.840 No Russian novels are short. 11:37.840 --> 11:40.670 Some novels by Dostoevsky are short. 11:40.670 --> 11:42.960 Therefore, some novels by Dostoevsky 11:42.960 --> 11:44.320 are not Russian novels. 11:44.320 --> 11:45.550 That's a valid argument. 11:45.550 --> 11:48.410 It's a valid argument with a false conclusion, because one 11:48.405 --> 11:51.825 of the premises is false. 11:51.827 --> 11:56.457 But as a result of the conclusion being implausible, 11:56.460 --> 12:02.530 only 55% of people were able to recognize that the 12:02.532 --> 12:05.972 conclusion followed from the premises. 12:05.970 --> 12:12.810 Notice, however, that these two arguments are 12:12.813 --> 12:13.703 structurally identical. 12:13.695 --> 12:19.215 They both have a form, no A's are B's, some C's are B's, 12:19.220 --> 12:23.520 therefore, some C's are not A's. 12:23.520 --> 12:28.170 Formal properties alone don't determine our 12:28.174 --> 12:31.964 ability to judge validity. 12:31.956 --> 12:39.026 And, in fact, it is also true that formal properties alone 12:39.030 --> 12:44.390 don't determine our ability to judge invalidity. 12:44.390 --> 12:48.240 So whereas valid arguments with plausible conclusions are 12:48.240 --> 12:52.330 judged to be valid roughly 90% of the time, we just heard 12:52.330 --> 12:56.410 that valid arguments with implausible conclusions are 12:56.410 --> 12:59.940 judged to be valid considerably less often. 12:59.943 --> 13:05.913 An opposite error arises in the case of invalid arguments. 13:05.914 --> 13:09.794 Invalid arguments with plausible conclusions are 13:09.790 --> 13:12.520 judged to be valid. 13:12.520 --> 13:16.900 It's judged to be the case that the form of the argument 13:16.900 --> 13:20.150 guarantees the truth of the conclusion. 13:20.150 --> 13:23.140 Whereas it turns out that what guarantees the truth of the 13:23.140 --> 13:26.980 conclusion, in that case, is nothing other than facts about 13:26.980 --> 13:29.570 the world, as opposed to facts about the 13:29.574 --> 13:33.044 structure of the argument. 13:33.035 --> 13:41.625 In light of this Thorndikean condition, researcher after 13:41.632 --> 13:46.562 researcher came up with research paradigms that 13:46.557 --> 13:51.637 demonstrated what we've just been talking about. 13:51.640 --> 13:59.470 So, famously, the Wason selection task asks people to 13:59.466 --> 14:04.726 determine which of four cards they need to turn over to 14:04.733 --> 14:06.583 verify the truth of a statement. 14:06.580 --> 14:09.210 So, for example, I might give you a statement, "if there's 14:09.205 --> 14:13.865 an A on the one side, there's a three on the other." And I 14:13.867 --> 14:17.327 might present you with four cards, and ask you which ones 14:17.330 --> 14:19.530 you need to turn over. 14:19.530 --> 14:22.620 Obviously, you need to turn over the A, and everybody 14:22.620 --> 14:23.780 recognizes that. 14:23.780 --> 14:25.910 You need to check whether there's a three 14:25.910 --> 14:27.640 on the other side. 14:27.644 --> 14:31.944 Obviously, you don't need to turn over the D. You know that 14:31.935 --> 14:34.125 the cards have a letter on one side, and a 14:34.130 --> 14:35.450 number on the other. 14:35.450 --> 14:37.890 So there's no worry that there's an A on the 14:37.891 --> 14:39.701 other side of here. 14:39.695 --> 14:43.805 But people have a tendency to think that you need to turn 14:43.805 --> 14:46.915 over the three, and that you don't need to 14:46.920 --> 14:48.610 turn over the seven. 14:48.610 --> 14:50.910 But look out. 14:50.910 --> 14:55.280 Right on that other side of the seven was an A. and the 14:55.280 --> 14:58.820 statement turns out to be false. 14:58.824 --> 15:04.444 The ones you need to check are the A and the seven. 15:04.440 --> 15:09.450 People find this task relatively difficult. 15:09.450 --> 15:13.370 But here's a structurally identical task that people 15:13.370 --> 15:15.650 find relatively easy. 15:15.650 --> 15:21.390 "If a person is drinking beer, the person must be over 21." I 15:21.390 --> 15:23.460 show you four cards. 15:23.460 --> 15:28.660 The beer drinker, the soda drinker, the over-21 year old, 15:28.660 --> 15:31.480 and the 17 year old. 15:31.478 --> 15:34.768 And every single one of you could I take it get a job as a 15:34.766 --> 15:40.066 bouncer, walk in, discover that the 17 year old is 15:40.070 --> 15:42.480 drinking a beer. 15:42.480 --> 15:46.290 And thereby, learn exactly what was hard to see in 15:46.290 --> 15:50.040 earlier case, that the cards you need to turn over are the 15:50.040 --> 15:57.810 first and the last, rather than the first and the third. 15:57.806 --> 16:02.496 Now, there have been all sorts of explanations hypothesized 16:02.502 --> 16:06.592 for why it is that we find the second of these tasks easier 16:06.590 --> 16:11.460 than the first. Perhaps the most famous of these is a 16:11.460 --> 16:15.030 hypothesis advanced by the evolutionary psychologists 16:15.030 --> 16:20.430 Leda Cosmides and John Tooby, who argue that we have within 16:20.430 --> 16:25.300 ourselves what they call a cheater detection module, and 16:25.295 --> 16:28.525 that we're enormously sensitive to cases that 16:28.528 --> 16:33.868 involve violations of normative rules. 16:33.874 --> 16:38.884 So you notice that the second one is different from the 16:38.876 --> 16:41.526 first in two ways. 16:41.530 --> 16:45.790 One is that it's socially embedded, whereas the first is 16:45.790 --> 16:46.660 purely abstract. 16:46.660 --> 16:50.110 And we'll discover, as we continue our readings this 16:50.114 --> 16:55.414 semester, that social embedding awakens reasoning 16:55.410 --> 16:58.350 processes that aren't present otherwise. 16:58.350 --> 17:04.480 In fact, we learned that last week, with the eyes study when 17:04.480 --> 17:09.400 we learned that people are more pro-social, more likely 17:09.400 --> 17:13.520 to engage in morally normative behavior, when they're 17:13.520 --> 17:16.910 presented visually, with eyes, in part, because it awakens 17:16.910 --> 17:20.880 and activates a kind of social understanding which all of us 17:20.878 --> 17:21.358 have. 17:21.364 --> 17:24.024 So the first difference between the bottom and the top 17:24.020 --> 17:29.130 is that the bottom one invokes sociality, whereas the top one 17:29.130 --> 17:31.490 is purely abstract. 17:31.490 --> 17:33.830 And the second difference between the bottom and top is, 17:33.830 --> 17:37.280 of course, that this is a normative rule. 17:37.280 --> 17:39.810 It's about how things ought to be. 17:39.810 --> 17:41.700 Whereas this is a descriptive rule. 17:41.700 --> 17:45.220 It's about how things are. 17:45.220 --> 17:49.050 So the lesson that we can take away from the Wason selection 17:49.046 --> 17:51.936 task, for those of you who are interested in it, there have 17:51.937 --> 17:54.507 been thousands and thousands of variations done on it, 17:54.510 --> 17:57.660 which are extremely interesting in sorting out 17:57.660 --> 18:01.370 exactly where people are good and not so good at the task. 18:01.372 --> 18:06.002 The lesson that we can take away from it is the lesson 18:06.000 --> 18:09.900 that has been emphasized throughout this lecture and 18:09.903 --> 18:11.453 the previous one. 18:11.450 --> 18:16.460 That we have ways of processing information that 18:16.460 --> 18:20.410 don't merely track formal properties. 18:20.408 --> 18:25.638 And some of those ways of processing information involve 18:25.640 --> 18:32.140 bringing on line, so to speak, what Plato would call parts of 18:32.140 --> 18:35.300 the soul which had not been previously 18:35.300 --> 18:39.410 attentive to the situation. 18:39.410 --> 18:40.590 Now-- 18:40.590 --> 18:42.550 Is there a question in the back? 18:42.550 --> 18:43.540 STUDENT: Yeah, sorry. 18:43.536 --> 18:47.106 For the last one, what exactly were the subjects asked to do? 18:47.110 --> 18:48.960 I don't really understand the experimental design. 18:48.959 --> 18:51.599 PROFESSOR: So the experimental design for both of these cases 18:51.600 --> 18:52.500 was as follows. 18:52.495 --> 18:54.725 You're given a sentence that you need to 18:54.730 --> 18:56.640 verify the truth of. 18:56.640 --> 19:00.360 So you're asked is it true that if there's an a on one 19:00.356 --> 19:01.846 side, there's a three on the other? 19:01.854 --> 19:05.004 Is it true that if a person is drinking beer, the person must 19:05.000 --> 19:06.590 be over 21? 19:06.590 --> 19:10.470 And you're told that you have four items in front of you, 19:10.473 --> 19:14.893 and you're asked, which ones do you need to turn over to 19:14.890 --> 19:16.370 verify the statement? 19:16.370 --> 19:19.000 So to verify the statement, if there's an a on one side, 19:18.998 --> 19:21.298 there's a three on the other side, you need to turn over 19:21.300 --> 19:24.390 the a and the seven, not the a and the three. 19:24.390 --> 19:26.840 And that should become obvious to you if you look at this 19:26.835 --> 19:30.365 case, where to determine whether the person was 19:30.365 --> 19:33.885 drinking beer, the person must be over 21, you need to turn 19:33.890 --> 19:38.760 over the card belonging to the beer drinker, and the card 19:38.760 --> 19:43.400 belonging to the person who is in fact not over 21. 19:43.400 --> 19:44.690 So does that clarify? 19:44.685 --> 19:45.985 Excellent. 19:45.990 --> 19:47.120 OK. 19:47.120 --> 19:52.670 So dual processing accounts attempt to provide a general 19:52.670 --> 19:56.890 explanation for what's going on in the cases I've just 19:56.890 --> 20:00.300 described, and in the cases I'm about to describe. 20:00.300 --> 20:05.870 They suggest that we have two relatively autonomous 20:05.871 --> 20:10.031 mechanisms for processing information, and they're 20:10.030 --> 20:13.920 called all sorts of things, but what's become the most 20:13.920 --> 20:18.010 normative way of describing them, is to call them simply 20:18.010 --> 20:20.910 system one and system two. 20:20.910 --> 20:24.790 So whereas system one is evolutionarily primitive--it 20:24.790 --> 20:28.940 makes use of parts of the brain that came into our 20:28.935 --> 20:32.085 evolutionary process relatively early in the 20:32.087 --> 20:36.437 game--system two is evolutionarily relatively 20:36.440 --> 20:39.510 recent, it involves higher cortical function. 20:39.510 --> 20:43.300 System one is unconscious, or preconscious, whereas system 20:43.300 --> 20:45.170 two is conscious. 20:45.174 --> 20:49.344 System one operates automatically, whereas system 20:49.340 --> 20:53.130 two is consciously controlled. 20:53.130 --> 20:56.610 System one is effortless, it happens without our trying. 20:56.611 --> 21:00.731 System two is effortful, in the sense that it involves an 21:00.734 --> 21:01.924 expenditure of cognitive energy; 21:01.917 --> 21:04.387 you have to pay attention. 21:04.390 --> 21:08.970 System one is super fast. It processes information almost 21:08.970 --> 21:10.490 instantaneously. 21:10.490 --> 21:12.680 System two is, relatively speaking-- 21:12.680 --> 21:15.900 slow, the information that we get through system two takes 21:15.900 --> 21:19.380 considerably longer: seconds rather than milliseconds. 21:19.384 --> 21:21.264 System one is associative. 21:21.256 --> 21:24.316 It recognizes patterns in the world. 21:24.320 --> 21:26.170 System two is rule-based. 21:26.170 --> 21:28.700 It can apply principles. 21:28.700 --> 21:32.050 System one is, people sometimes say, reflexive -- 21:32.050 --> 21:36.440 it happens without reflection, which is what 21:36.435 --> 21:40.415 underlies system two. 21:40.420 --> 21:46.070 So the distinction between system one and system two is 21:46.066 --> 21:51.886 the result of many decades of work by many people. 21:51.890 --> 21:55.600 Here's a chart from a different paper by Jonathan 21:55.600 --> 21:57.660 Evans in which he enumerates-- 21:57.655 --> 22:01.025 and this will be available to you on the V2 site-- 22:01.030 --> 22:06.220 some of the many researchers whose work went into talking 22:06.220 --> 22:09.100 about system one and system two. 22:09.100 --> 22:14.030 And I want to point out to you one important thing here, 22:14.030 --> 22:17.490 which is that it's a bit misleading to speak of system 22:17.486 --> 22:22.426 one and system two as if they are individual things. 22:22.428 --> 22:26.988 System one is sometimes called the autonomous set of 22:26.990 --> 22:31.120 subsystems. The idea: there is the visual processing, and the 22:31.120 --> 22:33.360 auditory processing, and there's processing that gives 22:33.355 --> 22:37.165 us very specific information about things like faces, or 22:37.170 --> 22:40.220 the average length of lines, or that enables us to 22:40.220 --> 22:43.280 recognize something as predator or prey. 22:43.280 --> 22:46.720 All of those systems have the characteristics 22:46.722 --> 22:48.132 that system one does. 22:48.130 --> 22:50.030 They're quick, they're unconscious, they're 22:50.034 --> 22:52.694 evolutionarily primitive, they come online without 22:52.690 --> 22:53.700 reflection. 22:53.700 --> 22:58.110 But they are not, each of them, they are not altogether 22:58.110 --> 23:01.790 a coherent system. 23:01.790 --> 23:07.240 So dual processing accounts are a way of trying to make 23:07.236 --> 23:11.906 sense of a set of phenomena, some of which have to do with 23:11.907 --> 23:17.667 the processing of reasoning, and some of which, as we 23:17.665 --> 23:22.415 learned in the absolutely lovely Nobel Prize speech of 23:22.420 --> 23:26.610 Daniel Kahneman, which I asked you to watch for today, some 23:26.605 --> 23:33.665 of which takes place in a more general domain of reasoning. 23:33.674 --> 23:36.424 And I've reproduced for you here Kahneman's beautiful 23:36.421 --> 23:39.811 chart explaining his understanding of the relation 23:39.805 --> 23:42.055 between system one and system two. 23:42.060 --> 23:45.610 Where he talks about the similarities between 23:45.606 --> 23:48.856 perception and intuition on the one hand and 23:48.860 --> 23:50.410 reasoning on the other. 23:50.410 --> 23:55.700 And again, I'll leave this slide for you on the website. 23:55.696 --> 24:00.306 So what I want to do in the next part of the lecture, is 24:00.307 --> 24:04.097 to move from the discussion of Evans, which in some ways was 24:04.100 --> 24:07.090 left over from last lecture though connected to this one, 24:07.094 --> 24:11.254 and talk a little bit about the work of Daniel Kahneman 24:11.250 --> 24:15.190 and his collaborator, Amos Tversky. 24:15.190 --> 24:16.960 So you're now going to get the second 24:16.956 --> 24:19.366 chance to use your clickers. 24:19.370 --> 24:22.470 We're only doing one other try today, and if this one works, 24:22.470 --> 24:25.600 we'll have a whole slew next Tuesday. 24:25.600 --> 24:28.810 OK, so this is the famous Asian disease problem from 24:28.812 --> 24:31.082 Kahneman and Tversky, and it runs as follows. 24:31.080 --> 24:35.360 A terrible disease has struck 600 people in your town. 24:35.360 --> 24:39.140 Without treatment, they're all doomed. 24:39.140 --> 24:42.940 You are the mayor, and there are two courses 24:42.940 --> 24:46.680 of treatment available. 24:46.680 --> 24:52.130 If your last name begins with the letters A through L, 24:52.130 --> 24:54.820 you're going to need to read the information that I'm going 24:54.816 --> 24:55.756 to put in the green box. 24:55.760 --> 25:00.170 So with you head pointing over to this side, and only read 25:00.173 --> 25:02.023 what happens in the green box. 25:02.015 --> 25:06.025 If your last name begins M through Z, you're going to 25:06.030 --> 25:08.360 read information in the blue box. 25:08.360 --> 25:11.580 And let me tell you, we will use these color conventions 25:11.576 --> 25:13.946 throughout, if this ends up working, and that the A 25:13.948 --> 25:18.038 through L group will have only numbers '1' and '2' for yes 25:18.042 --> 25:21.162 and no, and the M through Z group is going to have numbers 25:21.162 --> 25:23.232 '3' and '4' for yes and no. 25:23.230 --> 25:27.030 So if you're an M through Z-er, look at the blue, and 25:27.034 --> 25:30.674 get your fingers ready on three and four, and if you're 25:30.674 --> 25:33.254 an A through L-er, look at the green box and get your fingers 25:33.245 --> 25:35.705 ready on one and two. 25:35.710 --> 25:39.940 I asked Marvin Chun how to do slides like this, and this was 25:39.940 --> 25:40.630 his suggestion. 25:40.630 --> 25:41.560 OK. 25:41.556 --> 25:42.476 So ready? 25:42.482 --> 25:48.022 I'm going to tell you about plan A and plan B, so look at 25:48.020 --> 25:51.010 your side of the board and not the other. 25:51.010 --> 25:51.810 OK? 25:51.810 --> 26:02.940 So read about plan A. And now read about plan B. OK. 26:02.942 --> 26:10.302 Get your clickers out, and if you are on the green team, use 26:10.300 --> 26:14.460 '1' to indicate if you'd choose plan A, and '2' to 26:14.457 --> 26:18.857 indicate that you'd choose plan B. And if you're on the 26:18.858 --> 26:25.088 blue team, use '1' to indicate that you choose plan A and '2' 26:25.086 --> 26:25.936 to indicate that you 26:25.938 --> 26:28.158 choose plan B. OK. 26:28.156 --> 26:29.526 I'm going to put on the timer. 26:29.527 --> 26:32.657 We want 64 to 66, zillions and zillions 26:32.660 --> 26:33.850 and zillions of responses. 26:33.850 --> 26:38.860 And let's see how the numbers come out. 26:38.860 --> 26:40.300 OK. 26:40.300 --> 26:42.310 Here are our numbers. 26:42.310 --> 26:43.710 OK. 26:43.710 --> 26:51.940 Those of you on the green team, 40% of you chose plan A. 26:51.940 --> 26:58.590 Those of you on the blue team, only 25% of you chose plan A. 26:58.590 --> 27:00.470 Ah, you know what? 27:00.470 --> 27:02.700 This is actually not -- 27:02.695 --> 27:03.135 So. 27:03.140 --> 27:04.120 We-- 27:04.124 --> 27:07.274 The relevant size of the bar is relevant but this is 27:07.268 --> 27:08.468 divided into a 100. 27:08.470 --> 27:10.260 I need to learn a little bit more 27:10.260 --> 27:11.110 about how to use clickers. 27:11.110 --> 27:14.170 So let me now regroup and make my point again. 27:14.166 --> 27:18.896 As you'll notice, on the green team, the relative preference 27:18.903 --> 27:23.743 for plan A exceeded the relative preference for plan 27:23.740 --> 27:28.490 B. Whereas in the very small second half of the class, 27:28.486 --> 27:32.816 which consists of, the problem is that 60% of you are in A 27:32.821 --> 27:34.981 through L, and only 40% of you are M through Z 27:34.975 --> 27:36.145 (this is our problem). 27:36.148 --> 27:37.058 OK. 27:37.060 --> 27:41.850 But in this group, if I could quickly do 26 times 0.4, we 27:41.847 --> 27:43.737 would be able to find out what that what absolute 27:43.735 --> 27:44.505 percentage it was. 27:44.510 --> 27:47.460 In this group, the relative preference is for plan B 27:47.456 --> 27:49.916 rather than plan A. 27:49.920 --> 27:56.060 Notice, however, that plan A and plan B are identical. 27:56.062 --> 27:57.522 There's 600 people. 27:57.523 --> 28:02.983 And under plan A, 200 people will live; which means, 400 28:02.984 --> 28:04.724 people will die. 28:04.720 --> 28:09.850 And on this side, there are 600 people; if we go with plan 28:09.847 --> 28:14.507 A, 400 people will die, which means it's certain that 200 28:14.510 --> 28:16.780 people will live. 28:16.775 --> 28:21.685 However, the results which you all showed are, in fact, 28:21.692 --> 28:25.492 exactly the typical set of results. 28:25.485 --> 28:31.105 Typically, people presented with a problem that involves a 28:31.110 --> 28:35.850 choice between certainty and probability framed in terms of 28:35.850 --> 28:40.810 its positive outcomes will go with the certain rather than 28:40.809 --> 28:42.759 with risky plan. 28:42.760 --> 28:46.890 Whereas people who, told the same thing, where they're 28:46.894 --> 28:52.114 given to focus on the certainty that 400 people will 28:52.105 --> 28:59.195 die tend to go with the probabilistic option. 28:59.200 --> 29:04.180 Notice, again, plan A and plan B on the two sides 29:04.182 --> 29:06.622 are exactly the same. 29:06.623 --> 29:10.713 Just one is framed in terms of who will live, the other in 29:10.712 --> 29:12.782 terms of who will die. 29:12.782 --> 29:17.592 And the result is almost a complete inversion of people's 29:17.592 --> 29:19.852 preferences. 29:19.852 --> 29:24.462 And we get these sorts of framing effects over and over. 29:24.464 --> 29:27.514 Here's a study by Kahneman and Tversky's student, Eldar 29:27.507 --> 29:30.437 Shafir, from the early 1990s. 29:30.440 --> 29:33.080 You go to an ice cream store, and you're hoping to get 29:33.080 --> 29:35.340 yourself two flavors of ice cream. 29:35.340 --> 29:39.630 One is a good flavor, the other is an excellent flavor, 29:39.630 --> 29:42.010 but it has high cholesterol. 29:42.010 --> 29:46.880 And you discover that you only have enough money to buy one 29:46.882 --> 29:48.812 of the ice creams. 29:48.806 --> 29:54.726 So if I asked you which one would you choose, 28% of 29:54.725 --> 30:00.225 people choose the good flavor, and 72% of people choose the 30:00.231 --> 30:03.921 excellent flavor with high cholesterol. 30:03.920 --> 30:13.450 But if I ask you which one do you give up, 55% of the people 30:13.454 --> 30:19.484 give up the one with the good flavor, and 45% of the people 30:19.480 --> 30:24.550 give up the one with excellent flavor, high cholesterol. 30:24.552 --> 30:28.562 So even though, in this case, 28%-- 30:28.564 --> 30:32.284 even though these are exactly the same question, which do 30:32.282 --> 30:32.732 you choose? 30:32.734 --> 30:36.704 If you choose A, you give up B. If you choose B, you give 30:36.702 --> 30:41.192 up A. These numbers don't match up. 30:41.190 --> 30:45.100 When you're asked which one you choose, the excellent 30:45.100 --> 30:50.130 looms large, and so you go for it, neglecting the high 30:50.130 --> 30:51.090 cholesterol. 30:51.090 --> 30:53.900 When you're asked which one you give up, the high 30:53.895 --> 30:55.715 cholesterol looms larger. 30:58.500 --> 31:01.740 This phenomenon occurs over and over again. 31:01.740 --> 31:07.070 Suppose you're going to a movie, and when you get to the 31:07.072 --> 31:10.542 theater, you discover that you've lost something from 31:10.540 --> 31:12.370 your wallet. 31:12.370 --> 31:17.840 What you've lost from your wallet is either a $10 bill or 31:17.840 --> 31:21.600 a $10 movie ticket which you had purchased last 31:21.602 --> 31:23.922 night for the movie. 31:23.924 --> 31:27.104 You get to the theater, you open your wallet to go in, 31:27.100 --> 31:29.860 either, in the first case, to pull out a $10 bill to buy the 31:29.860 --> 31:33.600 ticket, or in the second case, to pull out your "Admit 1" 31:33.600 --> 31:36.960 ticket to let yourself in, and you discover that you've lost 31:36.960 --> 31:39.330 the item that would have enabled you to 31:39.328 --> 31:42.258 get into the theater. 31:42.256 --> 31:46.156 However, you have another $10 bill in your pocket, and the 31:46.160 --> 31:53.010 question is: do you buy another movie ticket? 31:53.005 --> 31:55.405 For people who've lost a $10 bill on the way to the movie 31:55.410 --> 31:59.190 theater, 90% of them say, "I lost a $10 bill, but so what. 31:59.190 --> 32:02.620 I'm going to buy a new ticket." For people who have 32:02.620 --> 32:07.370 lost a $10 movie ticket which is of exactly the same value 32:07.370 --> 32:14.050 as a $10 bill, only 42% said that they would spend the $10 32:14.050 --> 32:15.660 to buy another ticket. 32:18.985 --> 32:24.885 So framing is one of the examples of a heuristic, or 32:24.890 --> 32:26.970 bias, which Kahneman and Tversky 32:26.966 --> 32:30.616 focus on in their work. 32:30.622 --> 32:34.952 We will return to some other examples later in the 32:34.952 --> 32:38.062 semester, when we read the work of Cass Sunstein. 32:38.062 --> 32:41.402 What I want to point you to now is a particular example, 32:41.400 --> 32:45.750 which is going to serve as our segue into the idea of alief. 32:45.750 --> 32:48.380 And that's the distinction between frequency and 32:48.380 --> 32:49.750 probability. 32:49.751 --> 32:53.711 So suppose you're trying to get a red ball because getting 32:53.713 --> 32:57.013 a red ball will help you win a prize. 32:57.014 --> 33:02.254 And you have a choice about whether you want to draw from 33:02.250 --> 33:06.730 this box over here, which has nine white balls and one red 33:06.732 --> 33:14.242 one, or this box over here, which has eight red balls and 33:14.235 --> 33:15.995 92 white ones. 33:16.002 --> 33:20.562 So here you have a 10% chance, here you have an 8% chance. 33:20.560 --> 33:25.270 And you're going to be drawing from the boxes blindfolded. 33:25.270 --> 33:28.710 An image for which I don't suggest doing Google search. 33:32.561 --> 33:33.251 So, I found blindfolded justice. 33:33.250 --> 33:34.500 That did a little better. 33:36.830 --> 33:37.740 All right. 33:37.744 --> 33:39.584 So you have your choice. 33:39.580 --> 33:42.170 Do you want to let this box be the one -- 33:42.174 --> 33:44.264 you're trying to get a red ball, -- do you want to let 33:44.260 --> 33:46.650 this box be the one, or do you want to let 33:46.650 --> 33:48.550 this box be the one? 33:48.550 --> 33:53.570 Obviously, rationally, you've got a better chance over here. 33:53.571 --> 33:59.531 But people are in fact pulled in two directions. 33:59.525 --> 34:03.315 You had a 10% chance over here, but my goodness, there's 34:03.324 --> 34:05.084 8 balls over here! 34:05.080 --> 34:05.960 8 8! 34:05.960 --> 34:06.840 1 8! 34:06.840 --> 34:07.720 1 8! 34:07.720 --> 34:08.130 More! 34:08.130 --> 34:08.540 More! 34:08.540 --> 34:10.480 More! 34:10.484 --> 34:14.794 What's going on here, I suggest, is that whereas you 34:14.785 --> 34:20.885 have a belief that here, you have 10% chance, you have what 34:20.889 --> 34:26.059 I call an alief that here there are eight, whereas over 34:26.058 --> 34:29.338 there, there is one. 34:29.335 --> 34:36.135 So alief is a notion that I've actually discussed on 34:36.140 --> 34:41.840 bloggingheads.tv with Paul Bloom. 34:41.840 --> 34:45.820 And though we were filming in different places, here I am in 34:45.823 --> 34:48.843 the game room at my house with our geo-puzzles, here's Paul 34:48.836 --> 34:53.176 in the study, it appears that we were separated at birth, 34:53.180 --> 35:00.640 because that's at the same time as that photo. 35:00.640 --> 35:02.550 All right. 35:02.550 --> 35:07.700 So suppose I take you to the Grand Canyon, and I bring you 35:07.696 --> 35:12.036 out on the glass walkway that extends 4,000 feet above the 35:12.036 --> 35:13.916 roaring river below. 35:13.920 --> 35:16.880 And you step out there with me, and you 35:16.875 --> 35:20.015 voluntarily remain there. 35:20.022 --> 35:23.522 I take it if you voluntarily remain on a glass surface 35:23.524 --> 35:27.274 4,000 feet above a roaring river, you believe that 35:27.270 --> 35:29.930 surface is safe. 35:29.926 --> 35:34.946 Nonetheless, I will wager that most of you would shudder and 35:34.950 --> 35:37.060 shiver and shake. 35:37.060 --> 35:41.570 And you'd do so because you have what I would call an 35:41.568 --> 35:46.228 alief that says to you, "I am 4,000 feet in the air, with 35:46.230 --> 35:51.110 nothing holding me here, and I'm going to tremble." 35:51.110 --> 35:53.980 Or suppose we're watching a Western movie -- anybody 35:53.980 --> 35:59.210 recognize that gentleman who was president 35:59.205 --> 36:02.015 when you were born? 36:02.015 --> 36:05.105 When you watch a movie, a Western movie, and the bullets 36:05.106 --> 36:07.566 are flying off the stage, obviously you believe that 36:07.572 --> 36:08.412 you're safe. 36:08.410 --> 36:12.140 You don't think, "oh, what a good thing it was that the 36:12.136 --> 36:14.896 bullets didn't come off the screen this time". 36:14.900 --> 36:20.770 Nonetheless, particularly in 3D, you will 36:20.770 --> 36:22.330 bend your head down. 36:22.333 --> 36:24.783 If the green slime is coming off the screen, you'll tremble 36:24.780 --> 36:25.430 in your seat. 36:25.430 --> 36:27.790 If Anna Karenina is about to die, you will cry. 36:27.794 --> 36:31.694 Not because you believe that you're in danger, but because 36:31.686 --> 36:33.696 you have an alief. 36:33.700 --> 36:37.660 How many of you set your watch five minutes fast, and then 36:37.660 --> 36:40.140 subtract back down? 36:40.136 --> 36:42.226 When you do it, you believe that it's ten 36:42.232 --> 36:43.842 but you have an alief-- 36:43.840 --> 36:45.370 you look down at it -- 36:45.370 --> 36:45.720 10:05! 36:45.720 --> 36:46.780 It enters your visual system , it gets processed really fast 36:46.775 --> 36:49.835 and it says "10:05! 36:49.840 --> 36:51.700 Hurry!" 36:51.704 --> 36:56.524 Or suppose you're watching a rerun of your favorite team on 36:56.524 --> 37:01.584 television, and you know that if one of the guys on the team 37:01.575 --> 37:05.555 tries to steal second, he's going to be thrown out. 37:05.560 --> 37:09.870 And so you yell at the television screen, "Don't run! 37:09.868 --> 37:11.778 Don't run!" Why? 37:11.780 --> 37:14.350 Because you believe that your voice is going to go through 37:14.348 --> 37:20.538 the television screen, back in time, to first base to reach 37:20.540 --> 37:21.280 the runner? 37:21.280 --> 37:21.700 No. 37:21.700 --> 37:25.100 You have a belief that it's a rerun, and your alief says 37:25.095 --> 37:26.655 "don't run". 37:26.656 --> 37:29.456 Suppose you're on a diet, and you see this beautiful piece 37:29.458 --> 37:32.248 of chocolate cake and you have the belief that it's 37:32.254 --> 37:33.994 undesirable. 37:33.990 --> 37:38.380 Your alief system, in the form of your Platonic horse may 37:38.376 --> 37:40.896 nonetheless pull you towards it. 37:40.900 --> 37:46.230 Suppose I present you with this delicious cake, those are 37:46.228 --> 37:50.648 Tootsie rolls, this is a perfectly clean and sterilized 37:50.650 --> 37:54.250 pan, that's coconut there, in fact, this one has the same 37:54.252 --> 37:57.772 ingredients as that one, exactly. 37:57.766 --> 37:59.226 You believe me, right? 37:59.230 --> 38:01.060 I'm your professor, you're here, listening 38:01.059 --> 38:01.729 to what I tell you. 38:01.725 --> 38:03.215 I tell you, this is edible! 38:03.222 --> 38:05.222 Tootsie rolls, coconut. 38:05.218 --> 38:07.208 Nonetheless, I take it that your alief 38:07.214 --> 38:09.914 system kicks into gear. 38:09.910 --> 38:13.570 Suppose I ask you to sign this contract: "I hereby assert 38:13.570 --> 38:16.810 that my soul belongs only to you, O Satan". 38:16.810 --> 38:20.470 And I write at the bottom, "This is not a legal contract. 38:20.470 --> 38:25.520 It's just a prompt in a psychology experiment." You 38:25.520 --> 38:27.690 will, nonetheless, be reluctant to sign, not because 38:27.694 --> 38:30.994 you don't believe that this is a legal contract, right? 38:30.990 --> 38:31.240 [high-piched voice] 38:31.239 --> 38:34.099 "Oh my goodness, if I sign my soul over to the devil, I can 38:34.103 --> 38:39.393 tell it's the devil--it's parchment." No. 38:39.390 --> 38:42.760 You believe that there's nothing to it, but 38:42.757 --> 38:44.317 nonetheless, you hesitate. 38:44.320 --> 38:49.660 Suppose I take you to Monica Bonvicini's bathroom, which 38:49.660 --> 38:54.200 is, as you can see, completely opaque from the outside. 38:54.202 --> 38:57.402 You stand outside this public restroom, you peer into it, 38:57.400 --> 39:00.660 you see that there's no way to see inside. 39:00.660 --> 39:03.540 You belief is that you are totally protected. 39:03.542 --> 39:08.172 Nonetheless, when you go in to use the facilities, 39:08.170 --> 39:10.100 it looks like that. 39:10.100 --> 39:16.470 And your alief makes it rather difficult to do what you had 39:16.467 --> 39:21.437 gone in with the intention of doing. 39:21.440 --> 39:28.000 Suppose we have a bag of sugar and two glasses of water. 39:27.997 --> 39:32.577 And you take a spoonful of the sugar, and you put one in this 39:32.578 --> 39:36.208 glass and label it "Sugar," and one in this glass and 39:36.208 --> 39:41.658 label it "Poison." You took the sugar, you put it in the 39:41.660 --> 39:44.770 glasses, you put the labels on it. 39:44.770 --> 39:48.170 Nonetheless, people are reluctant to 39:48.174 --> 39:50.214 drink from this glass. 39:50.214 --> 39:52.634 Moreover, they're not just reluctant to drink from it 39:52.630 --> 39:53.940 when it says poison. 39:53.940 --> 39:56.720 They're reluctant to drink from it when they've written 39:56.719 --> 39:59.309 the words Not Poison on it. 39:59.312 --> 39:59.812 Why? 39:59.805 --> 40:01.275 Because the word poison is there. 40:01.284 --> 40:04.104 Running into your alief system. 40:04.100 --> 40:04.670 Good sugar. 40:04.670 --> 40:07.140 Mmm drink sugar. 40:07.140 --> 40:07.360 Poison? 40:07.355 --> 40:07.995 Poison? 40:08.000 --> 40:09.930 Don't drink the poison! 40:09.933 --> 40:11.323 --OK. 40:11.322 --> 40:17.322 Suppose that I have a kitchen, and I'm interested in making 40:17.318 --> 40:19.618 some kitty litter cake, so I have my cake 40:19.615 --> 40:22.065 pans stored over here. 40:22.073 --> 40:26.003 And a chef comes in and says to me, it would be much more 40:26.000 --> 40:30.420 efficient if the cake pan were over on the right. 40:30.420 --> 40:33.230 He's very pleased with what he's done. 40:33.230 --> 40:37.640 And when I ask him to get the cake pan, he says, "I'm so 40:37.640 --> 40:40.270 happy that we moved the muffin tin cake 40:40.265 --> 40:41.925 to the right cabinet". 40:41.930 --> 40:45.540 And those two get it, believing it could be on the 40:45.540 --> 40:49.470 right, exactly where it used to be. 40:49.470 --> 40:52.550 His belief is that it's on the right. 40:52.550 --> 40:57.180 His alief is a lagging habit. 40:57.180 --> 41:00.360 Any of you who has ever rearranged your room knows 41:00.357 --> 41:01.007 this feeling. 41:01.010 --> 41:04.780 Any of you who has ever put your cell phone in your hand, 41:04.778 --> 41:06.798 and then looked for your cell phone-- 41:06.796 --> 41:09.516 I've had terrors of having lost my children, and then 41:09.524 --> 41:10.864 realized they were on my shoulder. 41:14.326 --> 41:19.176 So to have an alief is roughly to have a representationally 41:19.180 --> 41:24.690 mediated propensity to respond to an apparent stimulus in a 41:24.693 --> 41:25.643 particular way. 41:25.635 --> 41:26.105 Right? 41:26.106 --> 41:28.986 So the apparent stimulus of the glass staircase, the 41:28.990 --> 41:32.860 apparent stimulus of the kitty litter cake, you have a 41:32.860 --> 41:36.940 propensity, that's either innate, as in the case of the 41:36.938 --> 41:44.148 glass, or the fudge shaped like feces, or a habitual 41:44.145 --> 41:47.855 propensity, as when you arrange the kitchen. 41:47.855 --> 41:53.535 These are habitual ways of responding to the world that 41:53.540 --> 41:56.030 activate the sort of lower-level systems that we've 41:56.030 --> 41:57.670 been talking about. 41:57.670 --> 42:01.020 And importantly--although we can recognize alief most 42:01.020 --> 42:04.610 easily by looking at these kinds of discordant cases, the 42:04.606 --> 42:08.226 cases where belief tells you to do one thing, and alief 42:08.230 --> 42:09.820 tells you to do the other. 42:09.815 --> 42:15.045 In fact, alief is active all the time. 42:15.050 --> 42:19.740 Every time I've used my right thumb to push the key on this, 42:19.740 --> 42:22.420 I've done it out of alief. 42:22.420 --> 42:25.920 Fortunately, it's one that corresponds with what I intend 42:25.920 --> 42:30.990 to do, but it is certainly the case that an enormous 42:30.986 --> 42:36.666 proportion of our actions are governed by alief. 42:36.668 --> 42:38.918 The question is this. 42:38.917 --> 42:43.367 Given that I just showed you that there are hundreds and 42:43.370 --> 42:48.080 hundreds of ways of describing what I adverted to with the 42:48.078 --> 42:54.048 notion of alief, why introduce this new term? 42:57.412 --> 43:02.262 The story has something to do with alief itself. 43:02.260 --> 43:07.850 So every 20 years or so, the United States government 43:07.850 --> 43:09.320 introduces a dollar coin. 43:09.320 --> 43:12.870 Here's the one they introduced in 1921. 43:12.870 --> 43:16.250 Here's the one they introduced in 1972. 43:16.253 --> 43:20.943 Here's the one they introduced later in the 1980s. 43:20.940 --> 43:22.970 Here's the one they introduced in 1980, 43:22.965 --> 43:24.545 with Susan B. Anthony. 43:24.551 --> 43:29.161 Here's the Sacajawea one they introduced in 2003, and in a 43:29.160 --> 43:32.840 massive fit of public relations genius, here's the 43:32.840 --> 43:38.410 Millard Fillmore golden coin, which will be issued soon. 43:38.410 --> 43:40.400 What's going on here? 43:40.400 --> 43:44.340 What's going on here is that it's hard to get people to 43:44.340 --> 43:49.190 make use of something if it doesn't fit into the currency 43:49.190 --> 43:53.250 system which they have already. 43:53.250 --> 43:57.730 Dollar coins don't fit naturally into the ways that 43:57.730 --> 44:02.260 Americans use money. 44:02.260 --> 44:07.140 Likewise, talk of system one, system two, relatively 44:07.135 --> 44:12.485 autonomous systems, heuristics biases and so on, don't fit 44:12.485 --> 44:16.285 naturally into the way that we have of 44:16.290 --> 44:18.490 talking about ourselves. 44:18.486 --> 44:23.256 We talk about ourselves in terms of beliefs and desires. 44:23.260 --> 44:29.680 And in order to make use of the gold coin that is the 44:29.680 --> 44:34.320 recognition of the multipartite soul, we need a 44:34.320 --> 44:38.280 notion that fits into our conceptual currency. 44:38.280 --> 44:44.280 And that's the role that my hope is alief will play. 44:44.280 --> 44:48.360 So alief is going to return to us in later lectures. 44:48.364 --> 44:51.604 We'll hear about it again in the context of the harmony of 44:51.598 --> 44:54.758 the soul and in other domains, we'll hear about heuristic and 44:54.760 --> 44:58.630 biases and we'll hear about the multipart soul. 44:58.630 --> 45:01.480 I look forward to seeing all of you on Tuesday for harmony 45:01.483 --> 45:02.733 and happiness.