1 00:00:00,000 --> 00:00:03,515 INTRODUCTION: The following content is presented by MIT 2 00:00:03,515 --> 00:00:06,026 OpenCourseWare under a Creative Commons license. 3 00:00:06,026 --> 00:00:08,537 Additional information about our license and MIT 4 00:00:08,537 --> 00:00:12,554 OpenCourseWare in general, is available at OCW.MIT.edu. 5 00:00:15,567 --> 00:00:23,180 PROFESSOR: Good afternoon here is the 6 00:00:23,180 --> 00:00:26,150 plan for today's lecture. 7 00:00:31,110 --> 00:00:38,360 There is out there somewhere, unless you're philosopher, a 8 00:00:38,360 --> 00:00:40,360 world, it's huge. 9 00:00:40,360 --> 00:00:43,500 A lot of stuff out there in the world. 10 00:00:43,500 --> 00:00:53,950 Their is inside use somewhere a long term memory, 11 00:00:53,950 --> 00:00:57,570 which is also huge. 12 00:00:57,570 --> 00:01:02,960 And and hopefully getting huger all the time. 13 00:01:05,580 --> 00:01:14,690 You cannot transcribe all of that world from the outside 14 00:01:14,690 --> 00:01:17,220 into your long term memory, as you may have already 15 00:01:17,220 --> 00:01:22,080 discovered if you've had a test this term. 16 00:01:22,080 --> 00:01:27,680 And in a move that ought to looks familiar to you by now. 17 00:01:27,680 --> 00:01:35,300 We can described that as a bottleneck of some sort. 18 00:01:35,300 --> 00:01:37,490 Not the same bottleneck that I was talking about in the 19 00:01:37,490 --> 00:01:41,440 context of attention, but a bottleneck nevertheless, that 20 00:01:41,440 --> 00:01:50,300 is involved in providing and putting a severe limitation on 21 00:01:50,300 --> 00:01:55,650 the ability to encode the world into long term memory. 22 00:01:55,650 --> 00:01:58,780 The first part of today's lecture is going to be about 23 00:01:58,780 --> 00:02:04,620 this problem of encoding and about the nature of this 24 00:02:04,620 --> 00:02:05,940 bottleneck. 25 00:02:05,940 --> 00:02:12,660 The second part is it that OK some stuff, a mere trickle of 26 00:02:12,660 --> 00:02:18,090 the world, managed to get through this bottleneck. 27 00:02:18,090 --> 00:02:19,540 It needs to stay there. 28 00:02:19,540 --> 00:02:22,240 It needs to stay there for a long time. 29 00:02:22,240 --> 00:02:27,160 You've got things in here, in this long term memory of 30 00:02:27,160 --> 00:02:35,450 yours, that are 15, 16 years old, maybe more at this point. 31 00:02:35,450 --> 00:02:42,900 How they stay here, how they are made firm, is the problem 32 00:02:42,900 --> 00:02:44,900 of consolidation. 33 00:02:44,900 --> 00:02:47,580 Which it would say underneath there if it wasn't that I put 34 00:02:47,580 --> 00:02:49,080 a screen in front. 35 00:02:49,080 --> 00:02:54,195 And finally, it is of absolutely no use to you have 36 00:02:54,195 --> 00:02:56,480 -- as you also may have discovered on some exam this 37 00:02:56,480 --> 00:02:57,720 term already -- 38 00:02:57,720 --> 00:03:02,020 to have a long term memory that is absolutely chock full 39 00:03:02,020 --> 00:03:06,080 of marvelous material if you can't get it back out. 40 00:03:06,080 --> 00:03:11,750 Unlike the story that we were telling before about 41 00:03:11,750 --> 00:03:14,940 perception, where there is a world of bottleneck and then 42 00:03:14,940 --> 00:03:18,880 some construct that was your perception of the world. 43 00:03:18,880 --> 00:03:22,640 In this case you've got a two way street, that you cannot 44 00:03:22,640 --> 00:03:25,780 appreciates the full contents of your long term memory at 45 00:03:25,780 --> 00:03:26,940 any given moment. 46 00:03:26,940 --> 00:03:29,545 You need to retrieve -- that's what it says on the far board 47 00:03:29,545 --> 00:03:32,050 -- you need to be able to retrieve material from long 48 00:03:32,050 --> 00:03:35,840 term memory, and bring it back as we will see. 49 00:03:35,840 --> 00:03:39,900 In effect to bring it back into the bottleneck. 50 00:03:43,710 --> 00:03:46,620 Got to be able to get back out into this limited capacity 51 00:03:46,620 --> 00:03:50,210 bottleneck where you can do things like write down the 52 00:03:50,210 --> 00:03:54,530 answer on a test, or tell me who's running for 53 00:03:54,530 --> 00:03:57,160 president this year. 54 00:03:57,160 --> 00:03:57,970 Nobody. 55 00:03:57,970 --> 00:03:59,500 AUDIENCE: Bush. 56 00:03:59,500 --> 00:04:00,520 PROFESSOR: Yeah that's one of them. 57 00:04:00,520 --> 00:04:01,560 Bush yeah. 58 00:04:01,560 --> 00:04:02,330 AUDIENCE: Kerry 59 00:04:02,330 --> 00:04:02,630 PROFESSOR: Kerry OK. 60 00:04:02,630 --> 00:04:03,960 And a few other characters too. 61 00:04:03,960 --> 00:04:04,240 Right. 62 00:04:04,240 --> 00:04:05,710 OK. 63 00:04:05,710 --> 00:04:09,600 Presumably except for the most politically obsessed among 64 00:04:09,600 --> 00:04:12,850 you, that was not an active thought in your mind 65 00:04:12,850 --> 00:04:13,980 until I asked it. 66 00:04:13,980 --> 00:04:17,460 But Kerry and Bush live in your long term memory, isn't 67 00:04:17,460 --> 00:04:18,830 that a scary thought? 68 00:04:18,830 --> 00:04:20,060 AUDIENCE: [LAUGHTER] 69 00:04:20,060 --> 00:04:24,090 PROFESSOR: And on demand, you could retrieve that back out 70 00:04:24,090 --> 00:04:28,320 into what we will call and your working memory. 71 00:04:28,320 --> 00:04:31,500 If we were just thinking about this and encoding terms, we 72 00:04:31,500 --> 00:04:36,660 might just call this short term memory -- stuff comes in 73 00:04:36,660 --> 00:04:39,870 from the outside, it's maintained for a little while 74 00:04:39,870 --> 00:04:44,190 somehow in a short term memory, and then somehow makes 75 00:04:44,190 --> 00:04:47,690 it into long term memory in ways that we'll talk about. 76 00:04:47,690 --> 00:04:52,900 But if we think about this in a two way, in a bi-directional 77 00:04:52,900 --> 00:04:56,750 kind of way, it's better perhaps to think of this as 78 00:04:56,750 --> 00:04:57,840 your working memory. 79 00:04:57,840 --> 00:05:01,800 Think of it as your computer desktop. 80 00:05:01,800 --> 00:05:03,750 It's got access to all the stuff that's 81 00:05:03,750 --> 00:05:05,410 on your hard drive. 82 00:05:05,410 --> 00:05:08,580 It's got access to the external world. 83 00:05:08,580 --> 00:05:10,530 You can tell it's got access to the external -- oh no we 84 00:05:10,530 --> 00:05:11,420 can't see it there. 85 00:05:11,420 --> 00:05:13,290 Well, you know, it's wireless and so it's got 86 00:05:13,290 --> 00:05:14,340 access to the web. 87 00:05:14,340 --> 00:05:15,970 And so that's the external world. 88 00:05:15,970 --> 00:05:18,630 But only a limited amount of it can be on the 89 00:05:18,630 --> 00:05:20,660 desktop any one time. 90 00:05:20,660 --> 00:05:24,710 Similarly, only a limited amount of either the world or 91 00:05:24,710 --> 00:05:28,430 the contents of your long term memory can be on your mental 92 00:05:28,430 --> 00:05:30,130 desktop at one time. 93 00:05:30,130 --> 00:05:31,260 That's working memory. 94 00:05:31,260 --> 00:05:37,560 That's an important junk of this bottleneck. 95 00:05:37,560 --> 00:05:43,800 Well what does it mean to say that a limited capacity 96 00:05:43,800 --> 00:05:45,510 bottleneck of some sort? 97 00:05:45,510 --> 00:05:49,780 Let me illustrate the limit to that capacity. 98 00:05:49,780 --> 00:05:53,970 We saw a particular example of that in what's known in the 99 00:05:53,970 --> 00:05:58,910 trade as visual short term memory last week, when I was 100 00:05:58,910 --> 00:06:01,160 putting up, you know, four little colored blobs and 101 00:06:01,160 --> 00:06:04,080 saying, OK now what changed. 102 00:06:04,080 --> 00:06:06,600 And there's a limit of about four objects that you 103 00:06:06,600 --> 00:06:08,480 can keep track of. 104 00:06:08,480 --> 00:06:17,280 But in trying to ask what's the limit of this bottleneck 105 00:06:17,280 --> 00:06:21,550 from the outside world into a more long term varieties of 106 00:06:21,550 --> 00:06:25,220 memory, one of the typical measures is what's called 107 00:06:25,220 --> 00:06:26,300 digit span. 108 00:06:26,300 --> 00:06:29,840 If you end up in the hospital because somebody bopped you 109 00:06:29,840 --> 00:06:32,330 over the head, one of the things you neurologist would 110 00:06:32,330 --> 00:06:34,180 do is read you a list of numbers, and ask you 111 00:06:34,180 --> 00:06:35,650 to repeat them back. 112 00:06:35,650 --> 00:06:39,830 And how many you can do is an index of whether you are OK. 113 00:06:39,830 --> 00:06:42,400 So let's see if you're OK. 114 00:06:42,400 --> 00:06:46,240 Except I won't do this with numbers, I will do it with 115 00:06:46,240 --> 00:06:46,870 color names. 116 00:06:46,870 --> 00:06:48,740 And we can come back to that in a minute. 117 00:06:48,740 --> 00:06:50,010 Where are my color names? 118 00:06:50,010 --> 00:06:50,780 Here are my color names. 119 00:06:50,780 --> 00:06:52,610 So what I'm going to do is read you a 120 00:06:52,610 --> 00:06:53,370 list of color names. 121 00:06:53,370 --> 00:06:57,230 I'll then say repeat, and you in glorious unison repeat this 122 00:06:57,230 --> 00:06:58,410 back to me. 123 00:06:58,410 --> 00:06:58,910 OK. 124 00:06:58,910 --> 00:07:03,610 So if I say, red, blue, purple, repeat. 125 00:07:03,610 --> 00:07:04,710 You say? 126 00:07:04,710 --> 00:07:07,440 AUDIENCE: Red, blue, purple. 127 00:07:07,440 --> 00:07:08,430 PROFESSOR: It never fails. 128 00:07:08,430 --> 00:07:10,720 AUDIENCE: [LAUGHTER] 129 00:07:10,720 --> 00:07:13,310 PROFESSOR: Right, [? Mara, ?] last year too? 130 00:07:13,310 --> 00:07:15,460 Always you get the repeat back too. 131 00:07:15,460 --> 00:07:18,280 And I'm not sure what it is about this that demands, that 132 00:07:18,280 --> 00:07:20,540 at least from some people, the repeats of repeat. 133 00:07:20,540 --> 00:07:21,970 But in any case good, good. 134 00:07:21,970 --> 00:07:25,800 You're not deeply brain damaged yet. 135 00:07:25,800 --> 00:07:26,670 Ready? 136 00:07:26,670 --> 00:07:27,990 We'll do some more of these. 137 00:07:27,990 --> 00:07:34,590 Green, pink, yellow, white yellow, repeat. 138 00:07:34,590 --> 00:07:41,340 AUDIENCE: Green, pink, yellow, white, yellow. 139 00:07:41,340 --> 00:07:42,180 PROFESSOR: OK. 140 00:07:42,180 --> 00:07:43,150 That was fine. 141 00:07:43,150 --> 00:07:45,740 All right, here we go. 142 00:07:45,740 --> 00:07:48,460 Oh, I forgot to mention. 143 00:07:48,460 --> 00:07:52,800 It turns out to be really easy to repeat back in essentially 144 00:07:52,800 --> 00:07:56,510 infinitely long list of these, if you write them down while 145 00:07:56,510 --> 00:07:57,582 I'm saying them. 146 00:07:57,582 --> 00:07:59,180 AUDIENCE: [LAUGHTER] 147 00:07:59,180 --> 00:08:01,300 PROFESSOR: There are a variety of demos that I'm going to do 148 00:08:01,300 --> 00:08:06,360 today that are of that form, and it's really boring. 149 00:08:06,360 --> 00:08:09,480 So, you know, OK we know you can do that, so don't do that, 150 00:08:09,480 --> 00:08:11,480 because that's boring. 151 00:08:11,480 --> 00:08:12,960 Be a nice honest person. 152 00:08:12,960 --> 00:08:14,070 OK ready? 153 00:08:14,070 --> 00:08:15,950 Where was I. I just did 5, right? 154 00:08:15,950 --> 00:08:16,770 Let's try 7. 155 00:08:16,770 --> 00:08:17,300 Ready? 156 00:08:17,300 --> 00:08:18,740 Oh and I told you when I'm going to stop. 157 00:08:18,740 --> 00:08:20,190 Oh what the heck. 158 00:08:20,190 --> 00:08:29,200 Pink, red, yellow, green, purple, pink, orange, repeat. 159 00:08:29,200 --> 00:08:35,980 AUDIENCE: Pink, red, yellow, green, purple, pink, orange. 160 00:08:35,980 --> 00:08:37,890 PROFESSOR: It's getting a little weak there at the end. 161 00:08:37,890 --> 00:08:41,590 But you will see, if you look at the handout, you will see 162 00:08:41,590 --> 00:08:45,090 that the answer to this question is given in the title 163 00:08:45,090 --> 00:08:47,690 of George Miller's classic paper on the subject. 164 00:08:47,690 --> 00:08:51,720 And is answer is 7 plus or minus 2, and some of you were 165 00:08:51,720 --> 00:08:53,760 on the minus side. 166 00:08:53,760 --> 00:08:56,070 Oh I should also say -- 167 00:08:56,070 --> 00:08:59,070 I should've said this at the beginning of the course -- 168 00:08:59,070 --> 00:09:02,670 people in intro psych classes, particularly by the time you 169 00:09:02,670 --> 00:09:05,060 get around starting to talk about cognitive things, like 170 00:09:05,060 --> 00:09:11,590 memory, take demos very personally and think 7 plus or 171 00:09:11,590 --> 00:09:15,230 minus 2, I got minus. 172 00:09:15,230 --> 00:09:15,820 I'm doomed. 173 00:09:15,820 --> 00:09:19,060 AUDIENCE: [LAUGHTER] 174 00:09:19,060 --> 00:09:20,140 PROFESSOR: You're probably fine. 175 00:09:20,140 --> 00:09:23,150 I mean if you, if it was the red, yellow, purple, you're 176 00:09:23,150 --> 00:09:25,750 saying what was the one after red. 177 00:09:25,750 --> 00:09:27,520 You know I admit there might be an issue there. 178 00:09:32,060 --> 00:09:35,410 The advantage of a 300 person intro class is we're looking 179 00:09:35,410 --> 00:09:38,140 at the average, and what you're doing on any given 180 00:09:38,140 --> 00:09:43,410 single isolated demo trial is not actually deeply relevant 181 00:09:43,410 --> 00:09:46,560 to whether or not you're going to pass physics for instance. 182 00:09:46,560 --> 00:09:47,390 OK, one more. 183 00:09:47,390 --> 00:09:48,850 Ready? 184 00:09:48,850 --> 00:09:58,130 Blue, green, purple, red, pink, yellow, green, blue, 185 00:09:58,130 --> 00:10:01,320 white, orange, repeat. 186 00:10:01,320 --> 00:10:07,570 AUDIENCE: Blue, green, purple, red, 187 00:10:07,570 --> 00:10:11,790 green, yellow [LAUGHTER] 188 00:10:11,790 --> 00:10:13,040 PROFESSOR: Yeah, yeah, yeah. 189 00:10:15,230 --> 00:10:16,850 You can hear. 190 00:10:16,850 --> 00:10:18,140 That was 10. 191 00:10:18,140 --> 00:10:21,360 And you could here it in the group, and you could probably 192 00:10:21,360 --> 00:10:25,160 feel it in yourself, that, you know, up to 6, 7, 8, you're 193 00:10:25,160 --> 00:10:26,160 yeah, we're good here. 194 00:10:26,160 --> 00:10:28,860 Oh man if I start taking in this next one, that 195 00:10:28,860 --> 00:10:30,180 I'm losing them here. 196 00:10:30,180 --> 00:10:31,650 Something bad is happening. 197 00:10:31,650 --> 00:10:38,430 So there's this capacity limit on what you can get into -- 198 00:10:38,430 --> 00:10:40,350 well let's used this jargon for the time being -- into 199 00:10:40,350 --> 00:10:43,210 some sort of short term memory, and then spit out 200 00:10:43,210 --> 00:10:44,660 immediately to me. 201 00:10:44,660 --> 00:10:49,430 Now 7 plus or minus to what? 202 00:10:49,430 --> 00:10:50,670 Let's try another one. 203 00:10:50,670 --> 00:10:51,830 You ready? 204 00:10:51,830 --> 00:10:54,090 Same game. 205 00:10:54,090 --> 00:11:03,510 Red, red, red, red, blue, blue, blue, blue, green, 206 00:11:03,510 --> 00:11:07,620 green, green, green, repeat. 207 00:11:07,620 --> 00:11:09,000 AUDIENCE: Red -- 208 00:11:09,000 --> 00:11:10,440 [LAUGHTER] 209 00:11:10,440 --> 00:11:13,030 PROFESSOR: You know you can do this. 210 00:11:13,030 --> 00:11:14,910 That's 12 items. 211 00:11:14,910 --> 00:11:19,790 How come I all of a sudden manage to boost your memories 212 00:11:19,790 --> 00:11:22,310 so effectively? 213 00:11:22,310 --> 00:11:25,480 Yeah classification, or what Miller originally, and the 214 00:11:25,480 --> 00:11:28,760 field had stuck with the notion of chunking. 215 00:11:28,760 --> 00:11:32,190 If you can break things into meaningful chunks , it's 7 216 00:11:32,190 --> 00:11:34,430 plus or minus 2 chunks. 217 00:11:34,430 --> 00:11:37,660 And what's important is how you managed to cut the world 218 00:11:37,660 --> 00:11:40,180 into chunks. 219 00:11:40,180 --> 00:11:43,380 This by the way explains why I don't do this demo in the 220 00:11:43,380 --> 00:11:49,800 traditional way of giving you digit strings. 221 00:11:49,800 --> 00:11:54,020 Because years ago I discovered that if you give somebody a 10 222 00:11:54,020 --> 00:11:59,010 digit, you know, give 300 MIT undergraduates a 10 digit 223 00:11:59,010 --> 00:12:01,910 string, you know, a bunch of them crud out on you, but then 224 00:12:01,910 --> 00:12:04,460 there's a handful of people who just raddle it right off. 225 00:12:04,460 --> 00:12:06,210 And you say oh my goodness, these people have 226 00:12:06,210 --> 00:12:07,920 this amazing memory. 227 00:12:07,920 --> 00:12:09,550 And you ask them how'd you do it? 228 00:12:09,550 --> 00:12:11,110 And they say oh well that was obvious. 229 00:12:11,110 --> 00:12:13,040 It was the natural log of 16. 230 00:12:13,040 --> 00:12:16,700 AUDIENCE: [LAUGHTER] 231 00:12:16,700 --> 00:12:20,760 PROFESSOR: MIT students tend to junk numbers rather more 232 00:12:20,760 --> 00:12:23,400 readily than the rest of the population. 233 00:12:23,400 --> 00:12:26,980 They don't tend to junk color names particularly adeptly 234 00:12:26,980 --> 00:12:30,020 I've found, and so you get the right 7 235 00:12:30,020 --> 00:12:32,320 plus or minus 2 answer. 236 00:12:32,320 --> 00:12:38,270 Now this sort of chunking, you should think of this more like 237 00:12:38,270 --> 00:12:38,940 a rate right. 238 00:12:38,940 --> 00:12:42,400 It's not 7 plus or minus 2 things ever, ever, ever. 239 00:12:42,400 --> 00:12:46,230 It's 7 plus or minus 2 things in some unit of time that you 240 00:12:46,230 --> 00:12:47,670 can manage to deal with. 241 00:12:47,670 --> 00:12:51,270 Otherwise as soon as I had presented 7 plus or minus 2 242 00:12:51,270 --> 00:12:54,400 units of information in this lecture, you'd be like dead 243 00:12:54,400 --> 00:12:55,610 for the rest of the day. 244 00:12:55,610 --> 00:12:58,570 Or maybe, you know, the first lecture shot your brain and 245 00:12:58,570 --> 00:13:00,460 that was it for the term. 246 00:13:00,460 --> 00:13:02,810 That's obviously not the case. 247 00:13:02,810 --> 00:13:06,310 What the question is how big a chunk can you manage to cut 248 00:13:06,310 --> 00:13:09,770 things into as it's coming in. 249 00:13:09,770 --> 00:13:15,690 One of the places you can see that is in language as you're 250 00:13:15,690 --> 00:13:19,740 trying to keep track for instance, what I'm saying. 251 00:13:19,740 --> 00:13:26,430 So if I ask you to repeat this sentence, memory is a 252 00:13:26,430 --> 00:13:28,900 fascinating topic that would be on the final 253 00:13:28,900 --> 00:13:31,180 exam, you would say? 254 00:13:31,180 --> 00:13:33,220 AUDIENCE: Memory is a fascinating topic that would 255 00:13:33,220 --> 00:13:35,480 be on the final exam. 256 00:13:35,480 --> 00:13:39,330 PROFESSOR: But if I lettuce, quadrangle forms only only 257 00:13:39,330 --> 00:13:42,380 column bug de la forms aftunic reply quintilian. 258 00:13:42,380 --> 00:13:43,866 Kelly You say? 259 00:13:43,866 --> 00:13:46,550 AUDIENCE: [LAUGHTER] 260 00:13:46,550 --> 00:13:49,460 PROFESSOR: Say it roughly the same number of syllables. 261 00:13:49,460 --> 00:13:51,070 But ouughh. 262 00:13:51,070 --> 00:13:53,300 You know, you can't repeat that back, because 263 00:13:53,300 --> 00:13:54,530 you can't chunk it. 264 00:13:54,530 --> 00:13:57,830 Well you can, but the chunks that you were making were too 265 00:13:57,830 --> 00:14:00,860 small to survive the rate at which 266 00:14:00,860 --> 00:14:03,080 they were being presented. 267 00:14:03,080 --> 00:14:07,230 So everybody here, if forced to, would've produce some 268 00:14:07,230 --> 00:14:08,030 little bit of it. 269 00:14:08,030 --> 00:14:13,530 But people would not be able to produce the whole thing, 270 00:14:13,530 --> 00:14:14,580 because you can't chunk it into 271 00:14:14,580 --> 00:14:16,240 meaningful chunks that work. 272 00:14:16,240 --> 00:14:20,350 This is also the route presumably of the near 273 00:14:20,350 --> 00:14:25,310 universal experience that when you go to some other country 274 00:14:25,310 --> 00:14:28,480 where they speak another language, that you nominally 275 00:14:28,480 --> 00:14:31,700 speak because you took four years of it in high school. 276 00:14:31,700 --> 00:14:35,670 You discover that those Parisians or those Spaniards 277 00:14:35,670 --> 00:14:38,390 or the Japanese or whoever talk really, 278 00:14:38,390 --> 00:14:40,470 really, really fast. 279 00:14:40,470 --> 00:14:42,710 What's the chance that everybody else in the world 280 00:14:42,710 --> 00:14:44,500 talks really, really, really fast. 281 00:14:44,500 --> 00:14:46,270 You know, it's not true of course. 282 00:14:46,270 --> 00:14:50,350 And those of you who are non-native English speakers, 283 00:14:50,350 --> 00:14:54,870 and have now arrived at MIT, are thinking no, you know, 284 00:14:54,870 --> 00:14:58,770 back in, you know, Kazakhstan they all spoke really slowly 285 00:14:58,770 --> 00:14:59,480 or whatever. 286 00:14:59,480 --> 00:15:02,440 But these English speakers are wackos. 287 00:15:02,440 --> 00:15:05,890 The problem is you are in your native tongue, or in any town 288 00:15:05,890 --> 00:15:07,730 but you are really fluent in. 289 00:15:07,730 --> 00:15:10,280 You are very good at cutting into 290 00:15:10,280 --> 00:15:12,200 meaningful chunks rapidly. 291 00:15:12,200 --> 00:15:15,100 If you're busy trying to say, what was that word. 292 00:15:15,100 --> 00:15:15,450 I know. 293 00:15:15,450 --> 00:15:17,450 I damn it that was the week I was asleep in 294 00:15:17,450 --> 00:15:18,415 high school or something. 295 00:15:18,415 --> 00:15:18,690 AUDIENCE: [LAUGHER] 296 00:15:18,690 --> 00:15:19,790 PROFESSOR: No that was the year I was 297 00:15:19,790 --> 00:15:20,310 asleep in high school. 298 00:15:20,310 --> 00:15:21,560 Anyway. 299 00:15:24,890 --> 00:15:30,390 The chunks go by too fast like in the nonsense that I was 300 00:15:30,390 --> 00:15:34,140 producing for you. 301 00:15:34,140 --> 00:15:41,560 You're set up to get chunks through in some domains much 302 00:15:41,560 --> 00:15:42,540 better than in others. 303 00:15:42,540 --> 00:15:46,890 And in fact, let me illustrate that here. 304 00:15:46,890 --> 00:15:48,180 By where did I hide? 305 00:15:48,180 --> 00:15:50,920 That looks like it's probably it. 306 00:15:50,920 --> 00:15:51,430 All right. 307 00:15:51,430 --> 00:15:53,850 Got to memorize some more stuff here. 308 00:15:53,850 --> 00:15:54,550 Come on. 309 00:15:54,550 --> 00:15:55,170 You want to go. 310 00:15:55,170 --> 00:15:58,070 And there we go. 311 00:15:58,070 --> 00:16:00,270 Too much light. 312 00:16:00,270 --> 00:16:04,670 Let's kill the stage lights. 313 00:16:04,670 --> 00:16:05,030 Good. 314 00:16:05,030 --> 00:16:07,310 OK. 315 00:16:07,310 --> 00:16:10,210 I want you to memorize all these pictures. 316 00:16:10,210 --> 00:16:12,230 You ready? 317 00:16:12,230 --> 00:16:14,240 There's a picture. 318 00:16:14,240 --> 00:16:16,320 And there's a picture. 319 00:16:16,320 --> 00:16:18,250 And that's a picture. 320 00:16:18,250 --> 00:16:21,800 And there's the back end of a horse or two. 321 00:16:21,800 --> 00:16:23,810 And there's a Christmas tree or something. 322 00:16:23,810 --> 00:16:28,110 And I don't know I must've rated a housing site. 323 00:16:28,110 --> 00:16:30,050 But, you know, these are definitely pictures. 324 00:16:30,050 --> 00:16:31,430 And that's a picture. 325 00:16:31,430 --> 00:16:33,570 And that looks like a picture. 326 00:16:33,570 --> 00:16:35,760 And oh look there's tomatoes. 327 00:16:35,760 --> 00:16:36,920 OK. 328 00:16:36,920 --> 00:16:37,650 Got them all? 329 00:16:37,650 --> 00:16:38,600 AUDIENCE: Yes. 330 00:16:38,600 --> 00:16:41,740 PROFESSOR: OK now what I want you to do is, I'm going to 331 00:16:41,740 --> 00:16:44,400 show you some pictures that are either old or new, and you 332 00:16:44,400 --> 00:16:48,920 just say yes if you've seen it before, no if you havent. 333 00:16:48,920 --> 00:16:50,520 AUDIENCE: No. 334 00:16:50,520 --> 00:16:51,980 No. 335 00:16:51,980 --> 00:16:53,280 Yes. 336 00:16:53,280 --> 00:16:54,780 Yes. 337 00:16:54,780 --> 00:16:55,970 No. 338 00:16:55,970 --> 00:17:00,060 Yes Yes, No. 339 00:17:00,060 --> 00:17:01,260 No. 340 00:17:01,260 --> 00:17:02,120 Yes. 341 00:17:02,120 --> 00:17:04,690 No. 342 00:17:04,690 --> 00:17:05,782 PROFESSOR: OK. 343 00:17:05,782 --> 00:17:10,020 The important point is that you're very good at this. 344 00:17:10,020 --> 00:17:15,080 The interesting side point is that this particular one is 345 00:17:15,080 --> 00:17:16,600 flipped left, right, reverse. 346 00:17:16,600 --> 00:17:18,730 So if you were sitting there saying why are these people 347 00:17:18,730 --> 00:17:20,320 next to me changing their mind. 348 00:17:20,320 --> 00:17:23,390 It's because some people noticed that it was left, 349 00:17:23,390 --> 00:17:24,390 right, reverse. 350 00:17:24,390 --> 00:17:29,990 But if I had lots of time, I could have done this with a 351 00:17:29,990 --> 00:17:31,330 lot of pictures. 352 00:17:31,330 --> 00:17:37,810 The largest report in the literature is 10,000 pictures 353 00:17:37,810 --> 00:17:42,130 shown to the usual collection of college undergraduates. 354 00:17:42,130 --> 00:17:47,930 And some days later, the students still remembered I 355 00:17:47,930 --> 00:17:53,760 think it's about 80% accuracy on that. 356 00:17:53,760 --> 00:17:57,560 If I read you 10,000 color names -- well there aren't 357 00:17:57,560 --> 00:18:01,380 10,000 color names -- if I read you 10,000 words or 358 00:18:01,380 --> 00:18:05,890 almost any other material that I would care to present to 359 00:18:05,890 --> 00:18:08,900 you, you would do nowhere near as well. 360 00:18:08,900 --> 00:18:11,830 We're not entirely clear on how you do this. 361 00:18:11,830 --> 00:18:13,610 This is quite a remarkable ability. 362 00:18:13,610 --> 00:18:17,880 But it does speak to the ability of picture 363 00:18:17,880 --> 00:18:19,870 information. 364 00:18:19,870 --> 00:18:22,030 Familiarity with a picture. 365 00:18:22,030 --> 00:18:24,380 Getting through this bottleneck into some sort of 366 00:18:24,380 --> 00:18:27,840 long term memory with remarkable efficiency, that 367 00:18:27,840 --> 00:18:31,030 you are somehow specialized for doing that. 368 00:18:31,030 --> 00:18:34,020 There's also a little bit of a cheat here. 369 00:18:34,020 --> 00:18:35,440 It's only a little bit of a cheat, but it's 370 00:18:35,440 --> 00:18:36,570 an interesting cheat. 371 00:18:36,570 --> 00:18:40,280 This was a recognition test. 372 00:18:40,280 --> 00:18:44,340 The color one, the previous demos were recall test. 373 00:18:44,340 --> 00:18:47,840 I was asking you what, you know, pull out of your memory 374 00:18:47,840 --> 00:18:48,600 what was there. 375 00:18:48,600 --> 00:18:51,840 Here I was saying was this there. 376 00:18:51,840 --> 00:18:52,990 And you know this. 377 00:18:52,990 --> 00:18:57,200 But, from your test taking experience you know that, OK 378 00:18:57,200 --> 00:18:59,960 if I give you a choice between fill in the blank and multiple 379 00:18:59,960 --> 00:19:01,280 choice, which do you want to do? 380 00:19:01,280 --> 00:19:03,170 AUDIENCE: Multiple choice. 381 00:19:03,170 --> 00:19:04,710 PROFESSOR: Multiple choice, because that's 382 00:19:04,710 --> 00:19:06,470 a recognition test. 383 00:19:06,470 --> 00:19:09,560 Fill in the blank is recall test. 384 00:19:09,560 --> 00:19:14,760 The distinction is OK here's your long term memory. 385 00:19:14,760 --> 00:19:18,610 Here's the target that you're trying to find in long term 386 00:19:18,610 --> 00:19:20,380 memory somewhere. 387 00:19:20,380 --> 00:19:25,960 If I'm doing a recall test, I'm saying, you know, go find 388 00:19:25,960 --> 00:19:29,520 this thing in all the vast halls of your long term 389 00:19:29,520 --> 00:19:33,200 memory, and you send some little probe out into long 390 00:19:33,200 --> 00:19:35,170 term memory that may or may not find it. 391 00:19:35,170 --> 00:19:40,060 If I say is this in there. 392 00:19:40,060 --> 00:19:42,910 That's an obviously simpler matching kind of task. 393 00:19:42,910 --> 00:19:48,740 And you are typically much better at doing recognition 394 00:19:48,740 --> 00:19:54,290 than you are at doing recall. 395 00:19:54,290 --> 00:19:59,410 So you have presumably special purpose mechanisms in your 396 00:19:59,410 --> 00:20:04,660 system that are really good at doing some tasks. 397 00:20:04,660 --> 00:20:09,700 What do you do in the case of those unfortunate tasks. 398 00:20:09,700 --> 00:20:12,090 Actually I don't need that anymore at all. 399 00:20:12,090 --> 00:20:14,490 In those case of those unfortunate tasks which you 400 00:20:14,490 --> 00:20:16,650 were not well designed to do. 401 00:20:16,650 --> 00:20:21,130 For instance the memorization of neuroanatomical terms for a 402 00:20:21,130 --> 00:20:24,730 midterm in intro psych, which is only the beginning. 403 00:20:24,730 --> 00:20:29,130 How many people over here are pre-meds or think they might 404 00:20:29,130 --> 00:20:29,930 be pre-meds? 405 00:20:29,930 --> 00:20:33,440 You guys are the ones who are going to devote years of your 406 00:20:33,440 --> 00:20:38,080 life to the memorization of things that you were not built 407 00:20:38,080 --> 00:20:39,450 to memorize. 408 00:20:39,450 --> 00:20:43,320 Like where all the bones are, and stuff like that. 409 00:20:43,320 --> 00:20:45,870 So how do you do that? 410 00:20:45,870 --> 00:20:50,110 Let's try memorizing some nonsense. 411 00:20:54,940 --> 00:20:55,680 Go to sleep. 412 00:20:55,680 --> 00:20:56,840 Why don't you want to go to sleep. 413 00:20:56,840 --> 00:20:58,730 AUDIENCE: [LAUGHTER] 414 00:20:58,730 --> 00:21:01,240 PROFESSOR: It's like your children. 415 00:21:01,240 --> 00:21:02,490 Mara, what do you do? 416 00:21:04,840 --> 00:21:06,060 I don't want to know OK. 417 00:21:06,060 --> 00:21:07,180 You mean I'm supposed to hit it? 418 00:21:07,180 --> 00:21:09,550 AUDIENCE: [LAUGHTER] 419 00:21:09,550 --> 00:21:10,920 PROFESSOR: Oh OK. 420 00:21:10,920 --> 00:21:13,360 All right. 421 00:21:13,360 --> 00:21:15,860 It went to sleep. 422 00:21:15,860 --> 00:21:17,130 It happens from time to time. 423 00:21:17,130 --> 00:21:18,250 Anyway where were we. 424 00:21:18,250 --> 00:21:20,250 Oh yes nonsense. 425 00:21:20,250 --> 00:21:23,040 This is a line of work that starts really with a guy named 426 00:21:23,040 --> 00:21:26,870 Ebbinghaus in Germany in the late 19th century, who wanted 427 00:21:26,870 --> 00:21:28,840 to understand memory processes. 428 00:21:28,840 --> 00:21:33,030 He's in isolation from these issues about meaning. 429 00:21:33,030 --> 00:21:38,560 And so what he did was he got himself actually for starters 430 00:21:38,560 --> 00:21:43,050 to memorize nonsense syllables that allegedly had no meaning. 431 00:21:43,050 --> 00:21:46,390 It's extremely hard to find material that has no meaning. 432 00:21:46,390 --> 00:21:50,300 But I will attempt to read you some now. 433 00:21:50,300 --> 00:21:52,560 Grab a hunk of paper. 434 00:21:52,560 --> 00:21:54,800 I don't want your writing them down while I read them to you, 435 00:21:54,800 --> 00:21:58,320 but I do want you to be in a position to write all the 436 00:21:58,320 --> 00:22:02,660 words down when I tell you it's time to write. 437 00:22:02,660 --> 00:22:06,430 Oh while you're doing that, let me make my annual 438 00:22:06,430 --> 00:22:08,000 disclaimer. 439 00:22:08,000 --> 00:22:12,760 MIT is a hugely multicultural institution. 440 00:22:12,760 --> 00:22:15,600 You speak a vast range of languages. 441 00:22:15,600 --> 00:22:18,960 To my knowledge what I am saying is meaningless. 442 00:22:18,960 --> 00:22:22,190 If it turns out that it is meaningful, particularly if it 443 00:22:22,190 --> 00:22:26,060 turns out that it's like deeply profane, do let me know 444 00:22:26,060 --> 00:22:29,700 later and I will change my nonsense yet again. 445 00:22:29,700 --> 00:22:32,760 I have over the years said a number of absolutely 446 00:22:32,760 --> 00:22:34,960 remarkable things. 447 00:22:34,960 --> 00:22:39,710 Including something in Mali so good that the young women who 448 00:22:39,710 --> 00:22:43,040 told me I couldn't say it, wouldn't tell me what 449 00:22:43,040 --> 00:22:44,900 it was I had said. 450 00:22:44,900 --> 00:22:45,190 But -- 451 00:22:45,190 --> 00:22:45,920 AUDIENCE: [LAUGHTER] 452 00:22:45,920 --> 00:22:49,180 PROFESSOR: -- they looked at my list and they just crossed 453 00:22:49,180 --> 00:22:50,350 these things out. 454 00:22:50,350 --> 00:22:51,350 Anyway. 455 00:22:51,350 --> 00:22:55,370 So it's mostly gone. 456 00:22:55,370 --> 00:22:57,250 So all right. 457 00:22:57,250 --> 00:22:58,370 You're ready to write? 458 00:22:58,370 --> 00:23:02,180 I'm going to read you a bunch of nonsense syllables, and 459 00:23:02,180 --> 00:23:06,960 you're going to write them all down. 460 00:23:06,960 --> 00:23:09,320 Ready OK. 461 00:23:09,320 --> 00:23:11,670 Fip. 462 00:23:11,670 --> 00:23:14,230 Dut. 463 00:23:14,230 --> 00:23:16,800 Moche. 464 00:23:16,800 --> 00:23:19,300 Yill. 465 00:23:19,300 --> 00:23:21,950 Saz. 466 00:23:21,950 --> 00:23:24,380 Tirt. 467 00:23:24,380 --> 00:23:26,530 Varl. 468 00:23:26,530 --> 00:23:29,430 Bince. 469 00:23:29,430 --> 00:23:32,040 Jucks. 470 00:23:32,040 --> 00:23:34,520 Gouf. 471 00:23:34,520 --> 00:23:37,250 Zauce. 472 00:23:37,250 --> 00:23:39,890 Rab. 473 00:23:39,890 --> 00:23:40,180 OK. 474 00:23:40,180 --> 00:23:41,410 Write down everything you can remember. 475 00:23:41,410 --> 00:23:51,070 AUDIENCE: [UNINTELLIGIBLE] 476 00:23:51,070 --> 00:23:52,460 PROFESSOR: Now don't copy off your neighbor. 477 00:24:03,170 --> 00:24:12,500 On your hand out on the second page I believe, you will find 478 00:24:12,500 --> 00:24:17,250 the axes for what's called the serial position curves. 479 00:24:17,250 --> 00:24:19,620 Which is the way we want to represent the data from this 480 00:24:19,620 --> 00:24:21,680 experiment. 481 00:24:21,680 --> 00:24:26,440 We want to represent percentage recall, which in 482 00:24:26,440 --> 00:24:29,460 this case will be in sort of number of hands, as a function 483 00:24:29,460 --> 00:24:31,470 of position in the list. 484 00:24:31,470 --> 00:24:38,210 Where in the list that I read did the word show up. 485 00:24:38,210 --> 00:24:40,980 And I will tell you the answer at ahead of time, and then 486 00:24:40,980 --> 00:24:44,190 we'll check if it worked. 487 00:24:44,190 --> 00:24:50,110 The answer is that the data will look like this. 488 00:24:50,110 --> 00:24:56,090 There will be what's known as a primacy effect, where you'll 489 00:24:56,090 --> 00:24:59,130 remember the first words in the list quite well. 490 00:24:59,130 --> 00:25:07,500 And a recency affect, with some preservation of the last 491 00:25:07,500 --> 00:25:08,810 words in the list. 492 00:25:08,810 --> 00:25:12,130 That at least is my assertion. 493 00:25:12,130 --> 00:25:16,490 Let us actually check the data here. 494 00:25:16,490 --> 00:25:19,240 How many people got fib? 495 00:25:22,570 --> 00:25:22,800 Alright. 496 00:25:22,800 --> 00:25:25,760 Data point turns out to be about there I think. 497 00:25:25,760 --> 00:25:27,190 How many people got dut? 498 00:25:30,250 --> 00:25:32,550 Few, fewer. 499 00:25:32,550 --> 00:25:35,560 How many people got -- well you see your falling a little 500 00:25:35,560 --> 00:25:36,590 below the line here. 501 00:25:36,590 --> 00:25:38,810 I am worried about you guys. 502 00:25:38,810 --> 00:25:42,530 How many got tirt? 503 00:25:42,530 --> 00:25:46,120 Ouhh how interesting. 504 00:25:46,120 --> 00:25:49,270 How many people got varl? 505 00:25:49,270 --> 00:25:52,510 Ouhh that's pretty pathetic. 506 00:25:52,510 --> 00:25:55,440 How many people got zauce. 507 00:25:55,440 --> 00:25:56,200 Oh coming back. 508 00:25:56,200 --> 00:25:57,390 Oh that's pretty good actually. 509 00:25:57,390 --> 00:25:58,670 That's above the line here. 510 00:25:58,670 --> 00:26:02,260 And how many people got rab? 511 00:26:02,260 --> 00:26:04,720 That's sort of thereish. 512 00:26:04,720 --> 00:26:05,940 OK. 513 00:26:05,940 --> 00:26:07,090 What's with tirt? 514 00:26:07,090 --> 00:26:09,280 [LAUGHTER] 515 00:26:09,280 --> 00:26:10,700 You guys did bet it. 516 00:26:10,700 --> 00:26:14,310 Well you know, teaching concourse this morning, I 517 00:26:14,310 --> 00:26:17,520 discovered that one of my words has currently acquired a 518 00:26:17,520 --> 00:26:20,280 slang meaning that I'd never heard of before. 519 00:26:20,280 --> 00:26:25,940 So tirt you just got lucky on, or tirt suddenly means -- 520 00:26:25,940 --> 00:26:27,580 AUDIENCE: [UNINTELLIGIBLE] 521 00:26:27,580 --> 00:26:30,320 PROFESSOR: What's a tirt? 522 00:26:30,320 --> 00:26:30,960 What? 523 00:26:30,960 --> 00:26:33,460 AUDIENCE: [UNINTELLIGIBLE] 524 00:26:33,460 --> 00:26:34,370 PROFESSOR: Tert. 525 00:26:34,370 --> 00:26:37,015 Oh oh tert like as in tertiary or something? 526 00:26:37,015 --> 00:26:37,350 AUDIENCE: Yes. 527 00:26:37,350 --> 00:26:40,390 PROFESSOR: Oh, see I'm spelling it t-i-r-t, and it 528 00:26:40,390 --> 00:26:42,490 never occurred to me. 529 00:26:42,490 --> 00:26:43,390 All right, all right. 530 00:26:43,390 --> 00:26:45,750 We're going to have to do something about tirt. 531 00:26:45,750 --> 00:26:47,000 Anyway. 532 00:26:49,490 --> 00:26:50,590 OK. 533 00:26:50,590 --> 00:26:53,860 The question here is going to be how did you do it? 534 00:26:53,860 --> 00:26:57,510 Part of the answer is clearly any word that happened to make 535 00:26:57,510 --> 00:27:03,590 some sort of meaningful association to you is going to 536 00:27:03,590 --> 00:27:07,280 have a preferential chance of getting into long term memory. 537 00:27:07,280 --> 00:27:09,700 Where did my beautiful little map of the world go? 538 00:27:09,700 --> 00:27:11,330 It must be underneath there. 539 00:27:11,330 --> 00:27:15,090 You go up. 540 00:27:15,090 --> 00:27:20,540 One way to think about this is sort of by analogy to 541 00:27:20,540 --> 00:27:21,500 immigration. 542 00:27:21,500 --> 00:27:27,100 You know here's America, and here are all of our various 543 00:27:27,100 --> 00:27:29,560 forebearances and stuff trying to get into the country. 544 00:27:29,560 --> 00:27:32,650 And there's this choke point at immigration. 545 00:27:32,650 --> 00:27:37,250 You know, if you got your Uncle Charlie already in the 546 00:27:37,250 --> 00:27:43,520 country, he can pull you in ways that if you have no 547 00:27:43,520 --> 00:27:46,200 relations here it's harder to get in. 548 00:27:46,200 --> 00:27:49,680 So if tirt reminded you of tertiary, that's Uncle 549 00:27:49,680 --> 00:27:52,300 Tertiary in long term memory. 550 00:27:52,300 --> 00:27:53,900 Sort of pulling you through. 551 00:27:53,900 --> 00:27:57,120 But the question is what do you do if you don't have that? 552 00:27:57,120 --> 00:27:59,700 So the things at the beginning. 553 00:27:59,700 --> 00:28:02,300 Anybody got any intuition about what they were doing 554 00:28:02,300 --> 00:28:06,750 with fip and dut and stuff like that? 555 00:28:06,750 --> 00:28:07,640 Yeah, Yeah, OK. 556 00:28:07,640 --> 00:28:10,800 AUDIENCE: I was writing down this fip and dut. 557 00:28:10,800 --> 00:28:11,310 PROFESSOR: Oh OK. 558 00:28:11,310 --> 00:28:13,130 Apart from the fact that you were -- no, no, you presumably 559 00:28:13,130 --> 00:28:15,060 were being a good person and not writing them down when I 560 00:28:15,060 --> 00:28:16,700 read them first though right? 561 00:28:16,700 --> 00:28:17,050 OK. 562 00:28:17,050 --> 00:28:22,660 So what were you doing while you were waiting to 563 00:28:22,660 --> 00:28:23,400 dump them back out? 564 00:28:23,400 --> 00:28:23,850 Yeah. 565 00:28:23,850 --> 00:28:25,643 AUDIENCE: I mean I was going over them in my head. 566 00:28:25,643 --> 00:28:25,980 Rehearsing them. 567 00:28:25,980 --> 00:28:26,240 PROFESSOR: Yeah. 568 00:28:26,240 --> 00:28:26,530 OK. 569 00:28:26,530 --> 00:28:28,230 You were sitting there rehearsing. 570 00:28:28,230 --> 00:28:29,920 How many people had the experience that they were 571 00:28:29,920 --> 00:28:33,600 rehearsing, and discovering that oh man there's more stuff 572 00:28:33,600 --> 00:28:35,360 here than I can rehearse? 573 00:28:35,360 --> 00:28:37,060 That's the capacity limit again. 574 00:28:37,060 --> 00:28:40,920 But there's a rehearsal loop for auditorium material 575 00:28:40,920 --> 00:28:45,090 sometimes called a phonological loop. 576 00:28:45,090 --> 00:28:48,710 Does it say phonological loop on the handout anywhere? 577 00:28:48,710 --> 00:28:49,010 No. 578 00:28:49,010 --> 00:28:50,060 Well. 579 00:28:50,060 --> 00:28:55,300 You know, phono like phonograph, phonological. 580 00:28:55,300 --> 00:28:58,520 Where you repeat this stuff to hold it in 581 00:28:58,520 --> 00:28:59,470 this working memory. 582 00:28:59,470 --> 00:29:02,250 Keep it alive in working memory. 583 00:29:02,250 --> 00:29:07,940 And in that way sort of keep it alive. 584 00:29:11,790 --> 00:29:14,330 Anybody got a different experience for zauce and rab 585 00:29:14,330 --> 00:29:15,580 at the end of there? 586 00:29:18,880 --> 00:29:20,490 I'm sorry I heard something that sounded promising. 587 00:29:20,490 --> 00:29:21,790 Yeah? 588 00:29:21,790 --> 00:29:21,931 AUDIENCE: They were just there. 589 00:29:21,931 --> 00:29:21,960 PROFESSOR: They 590 00:29:21,960 --> 00:29:27,610 were just there is actually the right intuition. 591 00:29:27,610 --> 00:29:32,360 It was almost like you could still hear them. 592 00:29:32,360 --> 00:29:34,110 They didn't need to be preserved yet. 593 00:29:34,110 --> 00:29:37,530 They were still fresh and hadn't rotted. 594 00:29:37,530 --> 00:29:41,430 The claim though is that the process that gives you the 595 00:29:41,430 --> 00:29:43,810 recency effect is different than the process that gives 596 00:29:43,810 --> 00:29:45,250 you the primacy effect. 597 00:29:45,250 --> 00:29:48,890 Let's give that a try. 598 00:29:48,890 --> 00:29:52,770 What I'll do is I'll read you another list of nonsense like 599 00:29:52,770 --> 00:29:54,320 the first one. 600 00:29:54,320 --> 00:29:57,870 I'm going to ask you to write them down again, but this time 601 00:29:57,870 --> 00:30:00,630 at the end instead of saying write I'll say count. 602 00:30:00,630 --> 00:30:06,380 When I say count, I want you to count backwards from 431 by 603 00:30:06,380 --> 00:30:07,750 3's out loud. 604 00:30:07,750 --> 00:30:09,650 You know all in beautiful unison. 605 00:30:09,650 --> 00:30:11,800 Right got that? 606 00:30:11,800 --> 00:30:15,450 That would make a lot of noise, so when I do this, keep 607 00:30:15,450 --> 00:30:15,910 an eye on me. 608 00:30:15,910 --> 00:30:18,600 When I do this, write everything down. 609 00:30:18,600 --> 00:30:21,720 OK? 610 00:30:21,720 --> 00:30:22,330 OK. 611 00:30:22,330 --> 00:30:23,540 They kind of got the instructions. 612 00:30:23,540 --> 00:30:25,770 All right, where's the rest of my nonsense here? 613 00:30:28,550 --> 00:30:29,310 AUDIENCE: [UNINTELLIGIBLE] 614 00:30:29,310 --> 00:30:30,780 PROFESSOR: OK. 615 00:30:33,810 --> 00:30:35,060 OK ready? 616 00:30:38,100 --> 00:30:40,690 Sip. 617 00:30:40,690 --> 00:30:43,480 Forth. 618 00:30:43,480 --> 00:30:46,010 Lig. 619 00:30:46,010 --> 00:30:48,680 Vop. 620 00:30:48,680 --> 00:30:51,360 Hearn. 621 00:30:51,360 --> 00:30:53,980 Mope. 622 00:30:53,980 --> 00:30:56,620 Jick. 623 00:30:56,620 --> 00:30:59,200 Tinned. 624 00:30:59,200 --> 00:31:01,730 Mez. 625 00:31:01,730 --> 00:31:04,500 Wamp. 626 00:31:04,500 --> 00:31:07,570 Flob. 627 00:31:07,570 --> 00:31:09,330 Gone. 628 00:31:09,330 --> 00:31:09,570 OK. 629 00:31:09,570 --> 00:31:11,160 Count 431. 630 00:31:11,160 --> 00:31:13,130 AUDIENCE: 431. 631 00:31:13,130 --> 00:31:14,380 [UNINTELLIGIBLE] 632 00:31:18,740 --> 00:31:19,130 PROFESSOR: OK. 633 00:31:19,130 --> 00:31:20,230 Write them all down. 634 00:31:20,230 --> 00:31:21,460 The giggles would probably do just fine. 635 00:31:21,460 --> 00:31:41,460 AUDIENCE: [LAUGHTER] 636 00:31:41,460 --> 00:31:41,790 PROFESSOR: OK. 637 00:31:41,790 --> 00:31:47,080 The claim here is that what that should do, did you feel 638 00:31:47,080 --> 00:31:49,670 yourself rehearsing the first ones again? 639 00:31:49,670 --> 00:31:50,590 AUDIENCE: Yeah. 640 00:31:50,590 --> 00:31:53,290 PROFESSOR: The claim is that the primacy affect, which is 641 00:31:53,290 --> 00:31:57,810 due to some sort of rehearsal into long term memory. 642 00:31:57,810 --> 00:32:01,270 That that primacy effect should still be there. 643 00:32:01,270 --> 00:32:05,170 But the recency effect due to some sort of it's just 644 00:32:05,170 --> 00:32:09,240 thereness, should have been disrupted by the counting and 645 00:32:09,240 --> 00:32:10,950 the chordals and stuff like that. 646 00:32:10,950 --> 00:32:11,310 OK. 647 00:32:11,310 --> 00:32:14,800 So let's check here, where's my list gone to? 648 00:32:14,800 --> 00:32:15,210 Here list. 649 00:32:15,210 --> 00:32:17,220 I should remember this by now. 650 00:32:17,220 --> 00:32:17,570 OK. 651 00:32:17,570 --> 00:32:21,270 How many people got sip? 652 00:32:21,270 --> 00:32:21,800 And look. 653 00:32:21,800 --> 00:32:24,060 That's almost exactly the same number. 654 00:32:24,060 --> 00:32:27,450 How many people got forth? 655 00:32:27,450 --> 00:32:27,630 Ooh. 656 00:32:27,630 --> 00:32:30,100 A lot of forths. 657 00:32:30,100 --> 00:32:34,550 How many people got jick? 658 00:32:34,550 --> 00:32:35,990 Ouhh deeply pathetic. 659 00:32:35,990 --> 00:32:37,870 AUDIENCE: [LAUGHTER] 660 00:32:37,870 --> 00:32:42,040 PROFESSOR: How many people got mez? 661 00:32:42,040 --> 00:32:43,260 Few more. 662 00:32:43,260 --> 00:32:45,900 How many people got flob? 663 00:32:45,900 --> 00:32:49,390 Ouhh that's pretty pathetic. 664 00:32:49,390 --> 00:32:52,470 How many people got gone? 665 00:32:52,470 --> 00:32:53,190 No gone? 666 00:32:53,190 --> 00:32:55,430 Gone was pretty well gone. 667 00:32:55,430 --> 00:32:57,280 These were supposed to be circles, So they'd be a 668 00:32:57,280 --> 00:32:58,210 different data set. 669 00:32:58,210 --> 00:33:02,080 So anyway, nice primacy effect. 670 00:33:02,080 --> 00:33:04,170 Recency effect gone. 671 00:33:04,170 --> 00:33:04,650 AUDIENCE: [UNINTELLIGIBLE] 672 00:33:04,650 --> 00:33:05,330 PROFESSOR: Yeah. 673 00:33:05,330 --> 00:33:06,780 There are other words in there. 674 00:33:06,780 --> 00:33:10,850 This is the phenomenon of MIT students really wanting to do 675 00:33:10,850 --> 00:33:12,340 well on these sort of things right? 676 00:33:12,340 --> 00:33:13,390 AUDIENCE: [LAUGHTER] 677 00:33:13,390 --> 00:33:14,870 PROFESSOR: Man, he didn't ask me. 678 00:33:14,870 --> 00:33:17,640 It's like, you know, you have the same experience for keeps 679 00:33:17,640 --> 00:33:18,480 on the exam. 680 00:33:18,480 --> 00:33:21,560 You sit there and you study chapter 3 until you're blue. 681 00:33:21,560 --> 00:33:25,430 And oh man they only asked about chapter 2. 682 00:33:25,430 --> 00:33:26,280 So yeah, yeah, yeah. 683 00:33:26,280 --> 00:33:28,990 I'm glad you got whatever it was wamp or -- 684 00:33:28,990 --> 00:33:30,370 AUDIENCE: [LAUGHTER] 685 00:33:30,370 --> 00:33:34,130 PROFESSOR: The worst thing about this you realize is that 686 00:33:34,130 --> 00:33:37,400 the stuff that you rehearsed into long term memory may stay 687 00:33:37,400 --> 00:33:39,350 there for a surprisingly long time. 688 00:33:39,350 --> 00:33:42,990 I had a guy come up to me on the subway some years back. 689 00:33:42,990 --> 00:33:45,700 Say to me, you don't know me right? 690 00:33:45,700 --> 00:33:48,720 I said that's right I don't know you. 691 00:33:48,720 --> 00:33:50,910 He said fip, dut. 692 00:33:50,910 --> 00:33:53,160 AUDIENCE: [LAUGHTER] 693 00:33:53,160 --> 00:33:55,650 PROFESSOR: I said you are a very strange person. 694 00:33:55,650 --> 00:34:00,410 But I now know where I know you from, or don't you from or 695 00:34:00,410 --> 00:34:04,060 something like that. 696 00:34:04,060 --> 00:34:13,660 You may have now warped your brain on ongoing basis. 697 00:34:13,660 --> 00:34:13,820 No. 698 00:34:13,820 --> 00:34:14,945 I don't want you to go up. 699 00:34:14,945 --> 00:34:15,925 Get back here. 700 00:34:15,925 --> 00:34:16,410 Stop. 701 00:34:16,410 --> 00:34:17,220 Come back down. 702 00:34:17,220 --> 00:34:19,360 You go up. 703 00:34:19,360 --> 00:34:19,740 All right. 704 00:34:19,740 --> 00:34:25,680 So back on the first page of the handout, you have what was 705 00:34:25,680 --> 00:34:28,670 sometimes called a standard model for getting from the 706 00:34:28,670 --> 00:34:30,100 outside world -- 707 00:34:30,100 --> 00:34:32,960 whoops my bottleneck disappeared again -- 708 00:34:32,960 --> 00:34:35,740 into long term memory. 709 00:34:35,740 --> 00:34:41,030 Which was stuff comes into a short term memory. 710 00:34:41,030 --> 00:34:43,990 It needs to be preserved in short term memory by something 711 00:34:43,990 --> 00:34:45,270 like rehearsal. 712 00:34:45,270 --> 00:34:48,860 If it lasts long enough in short term memory, it gets 713 00:34:48,860 --> 00:34:54,980 into long term memory where it manages to stay. 714 00:34:54,980 --> 00:34:58,360 There are certain problems with this kind of a model, 715 00:34:58,360 --> 00:35:00,300 which you ought to be able to anticipate now. 716 00:35:00,300 --> 00:35:02,820 But we can illustrate easily enough. 717 00:35:02,820 --> 00:35:06,070 How many people here had breakfast? 718 00:35:06,070 --> 00:35:06,730 OK. 719 00:35:06,730 --> 00:35:09,340 Person with the MIT shirt on there. 720 00:35:09,340 --> 00:35:10,690 What did you have for breakfast? 721 00:35:10,690 --> 00:35:11,500 AUDIENCE: Bagel. 722 00:35:11,500 --> 00:35:12,170 PROFESSOR: Bagels. 723 00:35:12,170 --> 00:35:12,870 Anything else? 724 00:35:12,870 --> 00:35:14,350 AUDIENCE: Bagel and cream cheese. 725 00:35:14,350 --> 00:35:15,060 PROFESSOR: Bagels and cream cheese. 726 00:35:15,060 --> 00:35:15,540 Oh OK. 727 00:35:15,540 --> 00:35:15,960 That's good. 728 00:35:15,960 --> 00:35:19,010 We want a little bit of a memory load here. 729 00:35:19,010 --> 00:35:21,380 Because I need to be able to explain that the reason that 730 00:35:21,380 --> 00:35:24,440 she knows that she had bagels and cream cheese for breakfast 731 00:35:24,440 --> 00:35:29,520 is that on the way out of her living arrangements, she was 732 00:35:29,520 --> 00:35:31,370 going bagels and cream cheese, bagels and cream cheese, Bagel 733 00:35:31,370 --> 00:35:32,516 and cream cheese. 734 00:35:32,516 --> 00:35:33,040 AUDIENCE: [LAUGHTER] 735 00:35:33,040 --> 00:35:34,850 PROFESSOR: Right? 736 00:35:34,850 --> 00:35:35,190 OK. 737 00:35:35,190 --> 00:35:37,610 That's good. 738 00:35:37,610 --> 00:35:38,200 No. 739 00:35:38,200 --> 00:35:38,990 She wasn't doing that. 740 00:35:38,990 --> 00:35:41,030 You weren't doing that, but you remember what you had for 741 00:35:41,030 --> 00:35:44,050 breakfast too. 742 00:35:44,050 --> 00:35:48,800 Obviously this rehearsal thing while it's important for 743 00:35:48,800 --> 00:35:52,390 getting nonsense from the world into long term memory, 744 00:35:52,390 --> 00:35:56,440 is not the only way to get through this bottleneck. 745 00:35:56,440 --> 00:35:59,300 Again things with meaning manage to 746 00:35:59,300 --> 00:36:02,540 get there more readily. 747 00:36:02,540 --> 00:36:13,380 And we can see that that fact tells us something perhaps 748 00:36:13,380 --> 00:36:19,050 about the way that our long term memory is put together. 749 00:36:19,050 --> 00:36:21,480 Now here what we're talking about is what's know as 750 00:36:21,480 --> 00:36:24,220 explicit long term memories as distinct from implicit. 751 00:36:24,220 --> 00:36:27,660 Explicit long term memory are the memories that you can get 752 00:36:27,660 --> 00:36:29,830 to if I ask you about them. 753 00:36:29,830 --> 00:36:32,500 That you can recover. 754 00:36:32,500 --> 00:36:35,180 So if I say, you know, who's running for President, you 755 00:36:35,180 --> 00:36:37,900 have an explicit memory that the answer is Kerry/Bush and 756 00:36:37,900 --> 00:36:42,610 whoever else you can name down the line there. 757 00:36:42,610 --> 00:36:50,350 If I asked you how do you say the word president, what are 758 00:36:50,350 --> 00:36:53,630 the tongue motions that are required. 759 00:36:53,630 --> 00:36:56,600 If you say president, you can now monitor it and figure out 760 00:36:56,600 --> 00:36:56,910 what they are. 761 00:36:56,910 --> 00:37:01,000 But you have no notion ahead of time of how you do that. 762 00:37:01,000 --> 00:37:03,730 You certainly remember how to do it in the sense that you 763 00:37:03,730 --> 00:37:05,130 can produce it on demand. 764 00:37:05,130 --> 00:37:06,660 That would be an implicit memory. 765 00:37:06,660 --> 00:37:09,290 And how you ride a bicycle. 766 00:37:09,290 --> 00:37:12,320 There are a whole slew of things like motor memories 767 00:37:12,320 --> 00:37:13,610 that are implicit. 768 00:37:13,610 --> 00:37:15,520 But here what we're talking about is getting into an 769 00:37:15,520 --> 00:37:17,790 explicit long term memory. 770 00:37:17,790 --> 00:37:24,460 It's useful to divide explicit long term memory into episodic 771 00:37:24,460 --> 00:37:33,770 memories for the episodes of your life, and semantic 772 00:37:33,770 --> 00:37:38,050 memories, your body of knowledge and facts and things 773 00:37:38,050 --> 00:37:38,700 of that sort. 774 00:37:38,700 --> 00:37:41,950 There not like walled off from each other. 775 00:37:41,950 --> 00:37:49,420 If I say, what's the capital of the United States, you dig 776 00:37:49,420 --> 00:37:52,860 out of your semantic memory the answer Washington DC. 777 00:37:52,860 --> 00:37:55,560 If you've been to Washington, you might then start digging 778 00:37:55,560 --> 00:37:59,000 out associated episodic memories of being 779 00:37:59,000 --> 00:38:02,330 there and so on. 780 00:38:02,330 --> 00:38:07,040 But within semantic memory, the notion here -- remember 781 00:38:07,040 --> 00:38:10,980 the Uncle Charlie or whoever he was notion, that things can 782 00:38:10,980 --> 00:38:16,960 reach out to the world and help pull things into long 783 00:38:16,960 --> 00:38:18,990 term memory -- 784 00:38:18,990 --> 00:38:23,600 is related to the way that we think that your long term up 785 00:38:23,600 --> 00:38:25,830 memory might be organized. 786 00:38:25,830 --> 00:38:28,100 And one of the popular ways of thinking about this is sort of 787 00:38:28,100 --> 00:38:34,640 a giant network of associations called a semantic 788 00:38:34,640 --> 00:38:35,890 network often. 789 00:38:39,450 --> 00:38:43,290 If I say cat, the first word that comes to your mind is? 790 00:38:43,290 --> 00:38:44,440 AUDIENCE: Meow. 791 00:38:44,440 --> 00:38:44,670 PROFESSOR: Meow. 792 00:38:44,670 --> 00:38:45,530 AUDIENCE: Dog. 793 00:38:45,530 --> 00:38:46,740 PROFESSOR: Dog. 794 00:38:46,740 --> 00:38:47,220 AUDIENCE: Mouse. 795 00:38:47,220 --> 00:38:47,670 PROFESSOR: OK. 796 00:38:47,670 --> 00:38:53,500 So there's a note in your long term memory that is cat in 797 00:38:53,500 --> 00:38:56,300 some sense in this story. 798 00:38:56,300 --> 00:39:01,730 It's close neighbors are things like dog 799 00:39:01,730 --> 00:39:07,550 and meow and -- 800 00:39:07,550 --> 00:39:10,020 what did we have -- mouse. 801 00:39:10,020 --> 00:39:12,890 And now it's important to realize this is not some sort 802 00:39:12,890 --> 00:39:16,630 of Linnaean taxonomy of, you know, species and geneus and 803 00:39:16,630 --> 00:39:18,650 all that sort of good stuff. 804 00:39:18,650 --> 00:39:20,380 You know, even though it's going to be connected with 805 00:39:20,380 --> 00:39:26,600 animal up here somewhere, animal and fur and stuff. 806 00:39:26,600 --> 00:39:28,400 But you might also have connections that 807 00:39:28,400 --> 00:39:32,550 go like meow, mow. 808 00:39:32,550 --> 00:39:37,110 AUDIENCE: [LAUGHTER] 809 00:39:37,110 --> 00:39:38,395 PROFESSOR: Actually that goes back here. 810 00:39:38,395 --> 00:39:41,280 I actually have a cat whose name is Chairman Mao by 811 00:39:41,280 --> 00:39:43,460 roughly this association. 812 00:39:43,460 --> 00:39:44,630 AUDIENCE: [LAUGHTER] 813 00:39:44,630 --> 00:39:47,330 PROFESSOR: You know, then, you know, China. 814 00:39:47,330 --> 00:39:51,840 AUDIENCE: [LAUGHTER] 815 00:39:51,840 --> 00:39:53,590 PROFESSOR: Plates. 816 00:39:53,590 --> 00:39:54,440 AUDIENCE: [LAUGHTER] 817 00:39:54,440 --> 00:39:59,940 PROFESSOR: So, you know, nobody has the notion that you 818 00:39:59,940 --> 00:40:05,170 could somehow map this out for any given individual. 819 00:40:05,170 --> 00:40:08,160 It's going to be a vastly complex and continuously 820 00:40:08,160 --> 00:40:12,360 changing set of associations. 821 00:40:12,360 --> 00:40:17,510 But there is evidence for this notion of proximity by 822 00:40:17,510 --> 00:40:18,170 association. 823 00:40:18,170 --> 00:40:23,340 By meaning in long term memory. 824 00:40:23,340 --> 00:40:28,460 And one of the ways you might find that out is by doing what 825 00:40:28,460 --> 00:40:31,700 are called priming experiments, which I might as 826 00:40:31,700 --> 00:40:33,990 well mention because I see that's -- 827 00:40:33,990 --> 00:40:36,260 is that on the handout properly. 828 00:40:36,260 --> 00:40:38,360 Well if it isn't, it should be. 829 00:40:38,360 --> 00:40:41,270 It probably says priming somewhere. 830 00:40:41,270 --> 00:40:44,330 Priming experiment. 831 00:40:44,330 --> 00:40:45,300 Evidence from priming. 832 00:40:45,300 --> 00:40:46,550 Look at that. 833 00:40:49,450 --> 00:40:57,140 On my computer screen here, I'll put up a string of 834 00:40:57,140 --> 00:41:01,060 letters, and you tell me whether or not it's a word. 835 00:41:01,060 --> 00:41:01,400 Right? 836 00:41:01,400 --> 00:41:04,270 So I put up. 837 00:41:04,270 --> 00:41:05,230 And you say? 838 00:41:05,230 --> 00:41:06,690 AUDIENCE: [UNINTELLIGIBLE] 839 00:41:06,690 --> 00:41:07,220 PROFESSOR: No. 840 00:41:07,220 --> 00:41:10,400 Even though it's now in your long term memory right. 841 00:41:10,400 --> 00:41:11,650 But if I put up. 842 00:41:15,550 --> 00:41:16,100 Yes. 843 00:41:16,100 --> 00:41:17,960 OK now. 844 00:41:17,960 --> 00:41:27,300 If the next word was something like truck, unless you have a 845 00:41:27,300 --> 00:41:29,480 dog in your truck or something like that. 846 00:41:33,150 --> 00:41:36,020 Well all right, let's compare this. 847 00:41:36,020 --> 00:41:40,350 The next word might be truck, or the next word might be cat. 848 00:41:40,350 --> 00:41:42,630 You'll be faster to say cat. 849 00:41:42,630 --> 00:41:44,930 You'll be faster to confirm that cat is a word 850 00:41:44,930 --> 00:41:46,790 than truck is a word. 851 00:41:46,790 --> 00:41:48,010 Why is that? 852 00:41:48,010 --> 00:41:52,960 Well seemingly when you go in to your semantic network to 853 00:41:52,960 --> 00:42:01,080 discover that dog is a word, you light up the dog node, and 854 00:42:01,080 --> 00:42:05,200 the activities spreads -- it's called spreading activation -- 855 00:42:05,200 --> 00:42:10,830 away from that node to the neighboring nodes. 856 00:42:10,830 --> 00:42:15,920 As a result, when you get the word cat, cat's already a 857 00:42:15,920 --> 00:42:17,560 little activated. 858 00:42:17,560 --> 00:42:19,840 The bell's been rung a little bit already. 859 00:42:19,840 --> 00:42:23,430 And your quicker to confirm that cat is a word than truck. 860 00:42:23,430 --> 00:42:27,440 Which is you know, down here in, you know, in the next 861 00:42:27,440 --> 00:42:29,670 county somewhere. 862 00:42:33,650 --> 00:42:36,480 In principle I suppose you could map out somebody's 863 00:42:36,480 --> 00:42:39,720 semantic network with a technique like this. 864 00:42:39,720 --> 00:42:42,580 But it would be a lifetime's endeavor ever, and not clear 865 00:42:42,580 --> 00:42:43,540 what the point is. 866 00:42:43,540 --> 00:42:47,350 But you could you can use this sort of a technique to show 867 00:42:47,350 --> 00:42:49,640 that some things live closer to each 868 00:42:49,640 --> 00:42:50,870 other than other things. 869 00:42:50,870 --> 00:42:55,260 Oh and by the way, when you go and reach in to retrieve -- 870 00:42:55,260 --> 00:42:56,950 this is sort of jumping to the retrieval part -- but when you 871 00:42:56,950 --> 00:43:04,550 go reach in to retrieve something from one node, you 872 00:43:04,550 --> 00:43:07,910 might goof and pull something from the neighboring node. 873 00:43:07,910 --> 00:43:11,290 And you know this, or you have experienced this if you are 874 00:43:11,290 --> 00:43:13,300 not an only child perhaps, even if you are. 875 00:43:13,300 --> 00:43:15,430 How many do you have siblings? 876 00:43:15,430 --> 00:43:16,460 Keep your hands up. 877 00:43:16,460 --> 00:43:19,240 How many of those have ever had your parents call you by 878 00:43:19,240 --> 00:43:21,190 the sibling's name? 879 00:43:21,190 --> 00:43:22,470 Almost everybody. 880 00:43:22,470 --> 00:43:23,640 Including some people who didn't have 881 00:43:23,640 --> 00:43:24,390 their hands up before. 882 00:43:24,390 --> 00:43:24,590 AUDIENCE: [LAUGHTER] 883 00:43:24,590 --> 00:43:27,480 PROFESSOR: Who didn't know that they had siblings until 884 00:43:27,480 --> 00:43:30,690 that happened. 885 00:43:30,690 --> 00:43:33,880 Or worse, you know, you get called by the name of the dog 886 00:43:33,880 --> 00:43:35,640 or something like that. 887 00:43:35,640 --> 00:43:39,330 This is not because whatever you may believe, you know, 888 00:43:39,330 --> 00:43:42,890 that your parents are unusually dim people. 889 00:43:42,890 --> 00:43:48,660 It's because all those terms presumably live very close to 890 00:43:48,660 --> 00:43:50,950 each in this semantic memory. 891 00:43:50,950 --> 00:43:53,490 And when particularly under some pressure you reach into 892 00:43:53,490 --> 00:43:56,180 grab kid number -- all my kids have the same name. 893 00:43:56,180 --> 00:44:01,750 Their name is Ben Phillip Simon, whatever your name is. 894 00:44:01,750 --> 00:44:06,470 And it works for all of them. 895 00:44:06,470 --> 00:44:08,460 Now there are certain drawbacks to this. 896 00:44:08,460 --> 00:44:13,400 It turns out to be decidedly unfortunate if in a moment of 897 00:44:13,400 --> 00:44:17,220 passion you murmur the name of the last girlfriend. 898 00:44:17,220 --> 00:44:17,560 AUDIENCE: [LAUGHTER] 899 00:44:17,560 --> 00:44:19,730 PROFESSOR: No matter how close they live in 900 00:44:19,730 --> 00:44:21,140 your semantic network. 901 00:44:24,320 --> 00:44:27,620 We can explain it, but we can't excuse it if you know 902 00:44:27,620 --> 00:44:38,490 what I. So big world, small desktop, small working memory, 903 00:44:38,490 --> 00:44:41,350 big long term memory. 904 00:44:44,670 --> 00:44:48,000 What's going on when the stuff in long term memory gets 905 00:44:48,000 --> 00:44:51,650 turned into something long term? 906 00:44:51,650 --> 00:44:56,180 It's really there for a long, long time. 907 00:44:56,180 --> 00:45:01,190 If you get a chance, David Pillemer's article that's in 908 00:45:01,190 --> 00:45:05,760 the website under the memory category about very long term 909 00:45:05,760 --> 00:45:09,090 memories, is fun to read as he goes and probes people's 910 00:45:09,090 --> 00:45:15,570 oldest memories from way, way back. 911 00:45:15,570 --> 00:45:17,480 How does stuff get consolidated. 912 00:45:17,480 --> 00:45:19,780 Well we know some things about this. 913 00:45:19,780 --> 00:45:20,970 We know quite a lot about this. 914 00:45:20,970 --> 00:45:24,840 But I want to tell you one bit, a couple of bits, that 915 00:45:24,840 --> 00:45:27,590 are important on a sort of a physiological front. 916 00:45:27,590 --> 00:45:30,170 So here's the brain again. 917 00:45:30,170 --> 00:45:32,940 Underneath the temporal lobe, not in the cortex of the 918 00:45:32,940 --> 00:45:40,480 temporal lobe, but underneath the temporal lobe struc the 919 00:45:40,480 --> 00:45:46,160 hippocampus, and to a lesser extent the amygdala. 920 00:45:46,160 --> 00:45:48,420 I'll just mention. 921 00:45:48,420 --> 00:45:51,310 But really we're talking about hippocampus here. 922 00:45:51,310 --> 00:45:55,470 Parts of the so called Limbic System that are vitally 923 00:45:55,470 --> 00:45:58,280 important in this process of consolidation. 924 00:45:58,280 --> 00:46:05,830 And we know that in the first instance from perhaps the most 925 00:46:05,830 --> 00:46:10,720 famous patient in the neuropsychological literature. 926 00:46:10,720 --> 00:46:15,530 He's a patient known in the literature as HM. 927 00:46:15,530 --> 00:46:17,380 He's the inspiration for the film 928 00:46:17,380 --> 00:46:18,710 Memento if you saw Memento. 929 00:46:21,530 --> 00:46:28,470 When he was a young man he had a bike accident that was 930 00:46:28,470 --> 00:46:33,330 probably the cause of his epilepsy. 931 00:46:33,330 --> 00:46:36,140 So he was an epileptic. 932 00:46:36,140 --> 00:46:39,580 You may recall I said earlier in the course that epilepsy is 933 00:46:39,580 --> 00:46:43,150 a sort of an electrical storm often started from a damaged 934 00:46:43,150 --> 00:46:46,360 piece of tissue in the brain. 935 00:46:46,360 --> 00:46:49,490 If it's not responding to drugs, which this was not, one 936 00:46:49,490 --> 00:46:53,920 of the treatment is to go and try to excise, to take out, 937 00:46:53,920 --> 00:46:55,670 the generator. 938 00:46:55,670 --> 00:47:02,850 The evidence in HM's case pointed to structures deep in 939 00:47:02,850 --> 00:47:03,730 the temporal lobe. 940 00:47:03,730 --> 00:47:06,540 Particularly the hippocampus and amygdala. 941 00:47:06,540 --> 00:47:12,410 And what was done in the late 50's, in the late 50's or 942 00:47:12,410 --> 00:47:14,240 early 60's, anybody remember? 943 00:47:14,240 --> 00:47:18,290 Anyway a long time ago, at this point, up in Montreal was 944 00:47:18,290 --> 00:47:24,010 that these structures were largely destroyed on both 945 00:47:24,010 --> 00:47:25,150 sides of his brain. 946 00:47:25,150 --> 00:47:25,870 That's important. 947 00:47:25,870 --> 00:47:29,670 This is a bilateral lesion. 948 00:47:29,670 --> 00:47:31,940 And this did have the effect of largely 949 00:47:31,940 --> 00:47:34,470 controlling his seizures. 950 00:47:34,470 --> 00:47:36,750 The seizure problem went away. 951 00:47:36,750 --> 00:47:42,140 However, he's a one of a kind patient, because the side 952 00:47:42,140 --> 00:47:46,340 effects, the unintentional consequences of this lesion, 953 00:47:46,340 --> 00:47:50,830 was so devastating that nobody could ethically 954 00:47:50,830 --> 00:47:52,410 ever do this again. 955 00:47:52,410 --> 00:47:59,660 At least not to a human patient. 956 00:47:59,660 --> 00:48:03,930 So he's still alive. 957 00:48:03,930 --> 00:48:06,890 And in fact, he's in a nursing facility near Boston. 958 00:48:06,890 --> 00:48:12,190 Because he was brought to Boston essentially as a 959 00:48:12,190 --> 00:48:15,040 psychological subject. 960 00:48:15,040 --> 00:48:17,690 Been unable to live on his own since the surgery. 961 00:48:17,690 --> 00:48:18,870 What's his problem? 962 00:48:18,870 --> 00:48:19,880 Well, he really has two. 963 00:48:19,880 --> 00:48:24,940 But the most relevant one for the present purposes is that 964 00:48:24,940 --> 00:48:29,370 he is simply unable to form new 965 00:48:29,370 --> 00:48:31,360 explicit long term memories. 966 00:48:35,670 --> 00:48:41,540 Take his interaction with a once upon a time TA in this 967 00:48:41,540 --> 00:48:44,910 course name John Gabrieli now a Professor at Stanford soon 968 00:48:44,910 --> 00:48:47,080 to be a Professor back here at MIT. 969 00:48:47,080 --> 00:48:50,290 But who did, I think he did his doctral work studying HM. 970 00:48:50,290 --> 00:48:52,880 But he certainly studied HM as a graduate student. 971 00:48:52,880 --> 00:48:59,420 He'd come into the lab to meet HM. 972 00:48:59,420 --> 00:49:01,600 Say hi, you know, have we've met before. 973 00:49:01,600 --> 00:49:02,610 HM said no. 974 00:49:02,610 --> 00:49:03,360 I don't think so. 975 00:49:03,360 --> 00:49:06,270 He's perfectly intact short term memory, which is fine for 976 00:49:06,270 --> 00:49:08,530 conducting a conversation. 977 00:49:08,530 --> 00:49:14,650 Also decently intact memories, this long term memory from 978 00:49:14,650 --> 00:49:16,870 before the operation. 979 00:49:16,870 --> 00:49:19,550 So it's not that we've somehow wiped out his 980 00:49:19,550 --> 00:49:20,680 storehouse of memories. 981 00:49:20,680 --> 00:49:23,710 It's the ability to put new stuff in here that's gone. 982 00:49:23,710 --> 00:49:25,690 So in comes Gabrieli. 983 00:49:25,690 --> 00:49:26,620 Have we ever met before? 984 00:49:26,620 --> 00:49:26,840 No. 985 00:49:26,840 --> 00:49:31,160 Hi I'm John Gabrieli, you know, it's nice to meet you. 986 00:49:31,160 --> 00:49:32,600 I'm HM, well I suppose he didn't call 987 00:49:32,600 --> 00:49:34,590 himself HM, but anyway. 988 00:49:34,590 --> 00:49:38,710 You use initials to protect the privacy of the patient. 989 00:49:38,710 --> 00:49:40,710 Not necessarily the initials of the patient. 990 00:49:40,710 --> 00:49:42,220 Something to label them. 991 00:49:42,220 --> 00:49:45,220 So anyway, fine have a nice conversation. 992 00:49:45,220 --> 00:49:46,570 Gabrieli goes out. 993 00:49:46,570 --> 00:49:47,480 Comes back in. 994 00:49:47,480 --> 00:49:49,270 Says hi, have we ever met before. 995 00:49:49,270 --> 00:49:50,900 No I don't think so. 996 00:49:50,900 --> 00:49:51,230 Well hi. 997 00:49:51,230 --> 00:49:52,690 I'm John Gabrieli nice to meet. 998 00:49:52,690 --> 00:49:54,240 Have the same conversation. 999 00:49:54,240 --> 00:49:56,820 You can do this over and over again. 1000 00:50:00,320 --> 00:50:03,690 In a more control kind of way, do a digit span kind of 1001 00:50:03,690 --> 00:50:04,440 experiment. 1002 00:50:04,440 --> 00:50:06,340 Give him a few digits to remember. 1003 00:50:06,340 --> 00:50:07,080 No problem. 1004 00:50:07,080 --> 00:50:09,360 He's got the same 7 plus or minus 2 that the rest of us 1005 00:50:09,360 --> 00:50:12,000 have pretty much. 1006 00:50:12,000 --> 00:50:15,810 Distract him momentarily, and it's just gone. 1007 00:50:15,810 --> 00:50:17,980 You've had this experience yourself if you try to get 1008 00:50:17,980 --> 00:50:21,740 from the phone book to the phone with a phone number. 1009 00:50:21,740 --> 00:50:24,670 And somebody says, didd you do x, y and z? 1010 00:50:24,670 --> 00:50:25,740 Ahh. 1011 00:50:25,740 --> 00:50:27,740 There goes the phone number. 1012 00:50:27,740 --> 00:50:27,990 AUDIENCE: [LAUGHTER] 1013 00:50:27,990 --> 00:50:31,700 PROFESSOR: That's his continuous experience, and has 1014 00:50:31,700 --> 00:50:34,540 been for 50 years at this point. 1015 00:50:34,540 --> 00:50:36,090 A few things have gotten in. 1016 00:50:36,090 --> 00:50:38,670 He happens to know that men have landed on 1017 00:50:38,670 --> 00:50:41,030 the moon for instance. 1018 00:50:41,030 --> 00:50:43,530 Perhaps because they're a little bits of the structure 1019 00:50:43,530 --> 00:50:45,930 that are maintained perhaps from another route. 1020 00:50:45,930 --> 00:50:47,020 We're not quite sure. 1021 00:50:47,020 --> 00:50:52,990 But basically no new episodic memories. 1022 00:50:52,990 --> 00:50:56,690 What this is telling us is that the hippocampus in 1023 00:50:56,690 --> 00:51:00,320 particular, other research indicates is critical for the 1024 00:51:00,320 --> 00:51:02,270 act of consolidation. 1025 00:51:02,270 --> 00:51:04,620 It is not the memory itself. 1026 00:51:04,620 --> 00:51:08,770 It is the mechanism that allows you to make those 1027 00:51:08,770 --> 00:51:11,260 memories solid in some fashion. 1028 00:51:11,260 --> 00:51:11,620 Yes? 1029 00:51:11,620 --> 00:51:14,160 AUDIENCE: But, does he know he has this disorder? 1030 00:51:14,160 --> 00:51:18,230 PROFESSOR: He knows he's got a memory problem. 1031 00:51:18,230 --> 00:51:19,870 And this actually is a good point. 1032 00:51:19,870 --> 00:51:22,470 He ties into his other problem, which is in some 1033 00:51:22,470 --> 00:51:25,560 sense an off setting benefits. 1034 00:51:25,560 --> 00:51:30,110 The Limbic System, certainly the amygdala, very important 1035 00:51:30,110 --> 00:51:32,170 in the mediation of emotion. 1036 00:51:32,170 --> 00:51:34,960 His experience of emotion is greatly flattened. 1037 00:51:34,960 --> 00:51:38,900 Flattened affect in the jargon of the trade. 1038 00:51:38,900 --> 00:51:40,160 So. 1039 00:51:40,160 --> 00:51:40,900 Which is good. 1040 00:51:40,900 --> 00:51:42,560 Because he does know that there's something wrong, but 1041 00:51:42,560 --> 00:51:44,570 it doesn't bother him that much. 1042 00:51:44,570 --> 00:51:46,260 You know what bothers him? 1043 00:51:46,260 --> 00:51:50,970 What bothers him is looking in the mirror. 1044 00:51:50,970 --> 00:51:52,300 Now why should that be? 1045 00:51:52,300 --> 00:51:53,470 Well imagine this. 1046 00:51:53,470 --> 00:51:57,240 His lesion happened when he's in his 20s. 1047 00:51:57,240 --> 00:52:00,040 He's still got the long term memories that 1048 00:52:00,040 --> 00:52:02,160 he had in his 20s. 1049 00:52:02,160 --> 00:52:03,710 Basically your age. 1050 00:52:03,710 --> 00:52:07,520 Now you go back to your room, you look in the mirror, and 1051 00:52:07,520 --> 00:52:11,470 what looks out at you is a 70 year old guy or women as the 1052 00:52:11,470 --> 00:52:12,820 case may be. 1053 00:52:12,820 --> 00:52:14,570 I suppose it doesn't particularly matter. 1054 00:52:14,570 --> 00:52:16,460 It's going to be a disaster under either 1055 00:52:16,460 --> 00:52:17,870 circumstance right? 1056 00:52:17,870 --> 00:52:21,360 You're going to think I'm in a sci-fi movie, and I 1057 00:52:21,360 --> 00:52:23,400 don't have the part. 1058 00:52:23,400 --> 00:52:26,320 You know the good guy with the laser gun part 1059 00:52:26,320 --> 00:52:27,610 that I really wanted. 1060 00:52:27,610 --> 00:52:30,930 You know I'm in big trouble here. 1061 00:52:30,930 --> 00:52:35,110 And this is disturbing to him. 1062 00:52:35,110 --> 00:52:37,660 Not as disturbing as it would be to you again, because his 1063 00:52:37,660 --> 00:52:40,610 emotional responses has been very severely blunted. 1064 00:52:40,610 --> 00:52:43,230 So much so that he needs to be asked 1065 00:52:43,230 --> 00:52:46,420 whether or not he's sick. 1066 00:52:46,420 --> 00:52:48,040 You know do you have a stomach ache? 1067 00:52:48,040 --> 00:52:50,010 No that you mentioned it, yes I do. 1068 00:52:50,010 --> 00:52:52,870 It's not that he couldn't feel the pain, it's just without 1069 00:52:52,870 --> 00:52:56,200 the emotional overlay so, you know. 1070 00:52:56,200 --> 00:52:57,020 I see blue. 1071 00:52:57,020 --> 00:52:58,565 I have pain what's the big deal? 1072 00:52:58,565 --> 00:52:58,980 AUDIENCE: [LAUGHTER] 1073 00:52:58,980 --> 00:52:59,380 PROFESSOR: Yes? 1074 00:52:59,380 --> 00:53:01,710 AUDIENCE: [UNINTELLIGIBLE] motor memory 1075 00:53:01,710 --> 00:53:03,010 PROFESSOR: Yes. 1076 00:53:03,010 --> 00:53:05,170 Nice point. 1077 00:53:05,170 --> 00:53:09,180 So for example, I don't know how many of you 1078 00:53:09,180 --> 00:53:12,140 have hung out at -- 1079 00:53:12,140 --> 00:53:14,100 you know get rid of the brain here. 1080 00:53:14,100 --> 00:53:16,140 Who needs the brain? 1081 00:53:16,140 --> 00:53:21,020 So any you ever done a mirror maze? 1082 00:53:21,020 --> 00:53:22,670 Some people may have actually built one of 1083 00:53:22,670 --> 00:53:23,510 these once upon a time. 1084 00:53:23,510 --> 00:53:26,070 So imagine you have like a star. 1085 00:53:26,070 --> 00:53:26,630 Oops. 1086 00:53:26,630 --> 00:53:29,140 Imagine you could actually draw a star. 1087 00:53:29,140 --> 00:53:32,550 AUDIENCE: [LAUGHTER] 1088 00:53:32,550 --> 00:53:33,000 PROFESSOR: All right. 1089 00:53:33,000 --> 00:53:34,250 Well. 1090 00:53:36,480 --> 00:53:39,120 So you make this out of like aluminum foil, and you go and 1091 00:53:39,120 --> 00:53:40,730 trace it with a metal stylus. 1092 00:53:40,730 --> 00:53:44,520 And if you go off the aluminum foil it makes a nasty noise. 1093 00:53:44,520 --> 00:53:46,180 Because you're breaking a circuit or 1094 00:53:46,180 --> 00:53:47,190 something like that. 1095 00:53:47,190 --> 00:53:49,590 Anybody ever build one of those? 1096 00:53:49,590 --> 00:53:51,630 Oh what's the world coming to. 1097 00:53:51,630 --> 00:53:52,700 It's easy right? 1098 00:53:52,700 --> 00:53:53,840 Anybody can do this. 1099 00:53:53,840 --> 00:53:59,210 What people can't do regular readily is do this looking at 1100 00:53:59,210 --> 00:54:01,140 the image in a mirror. 1101 00:54:01,140 --> 00:54:01,390 Right? 1102 00:54:01,390 --> 00:54:05,310 Because then think they got [UNINTELLIGIBLE]. 1103 00:54:05,310 --> 00:54:07,630 AUDIENCE: [LAUGHTER] 1104 00:54:07,630 --> 00:54:10,340 PROFESSOR: But you learn to do this. 1105 00:54:10,340 --> 00:54:14,550 And HM has been trained in one set of experiments to do this. 1106 00:54:14,550 --> 00:54:18,230 In multiple sessions he got better. 1107 00:54:18,230 --> 00:54:21,160 If you asked him have you ever done this before on the nth 1108 00:54:21,160 --> 00:54:22,920 time that he did it. 1109 00:54:22,920 --> 00:54:23,540 The answer is no. 1110 00:54:23,540 --> 00:54:25,390 I don't think so. 1111 00:54:25,390 --> 00:54:27,900 And he goes [UNINTELLIGIBLE]. 1112 00:54:27,900 --> 00:54:29,740 Hey, you know, that was really good. 1113 00:54:29,740 --> 00:54:31,840 Most people asked how did you do that? 1114 00:54:31,840 --> 00:54:32,090 Yeah. 1115 00:54:32,090 --> 00:54:33,500 I've always been good at that kind of thing. 1116 00:54:33,500 --> 00:54:35,320 Or, you know, just lucky. 1117 00:54:35,320 --> 00:54:39,190 No episodic memory, no explicit memory of 1118 00:54:39,190 --> 00:54:40,000 ever having done it. 1119 00:54:40,000 --> 00:54:43,500 But yes a motor memory that allows them to do it. 1120 00:54:43,500 --> 00:54:48,890 Similarly, your sense of familiarity runs 1121 00:54:48,890 --> 00:54:51,390 on different pathways. 1122 00:54:51,390 --> 00:54:56,880 So if I bumped into you wandering around the halls, I 1123 00:54:56,880 --> 00:54:58,590 may well have the sense that I've seen 1124 00:54:58,590 --> 00:55:01,040 you somewhere before. 1125 00:55:01,040 --> 00:55:08,320 And I might guess that that's, you know, I've seen you in 1126 00:55:08,320 --> 00:55:10,400 intro psych, because that's where I see a lot of people 1127 00:55:10,400 --> 00:55:12,060 whose names I don't really quite know 1128 00:55:12,060 --> 00:55:13,670 and stuff like that. 1129 00:55:13,670 --> 00:55:14,780 So that would work. 1130 00:55:14,780 --> 00:55:17,690 And HM gets that sense of familiarity too. 1131 00:55:17,690 --> 00:55:23,040 So after a while actually with Gabrieli, Gabrieli comes in, 1132 00:55:23,040 --> 00:55:24,050 have we ever met before? 1133 00:55:24,050 --> 00:55:24,200 Yeah. 1134 00:55:24,200 --> 00:55:25,680 You looks familiar. 1135 00:55:25,680 --> 00:55:26,570 Do you know my name? 1136 00:55:26,570 --> 00:55:26,970 Know you're name? 1137 00:55:26,970 --> 00:55:28,900 Forgotten your name. 1138 00:55:28,900 --> 00:55:29,900 Who do you think I am? 1139 00:55:29,900 --> 00:55:31,400 I think you're probably somebody from 1140 00:55:31,400 --> 00:55:33,280 my high school class. 1141 00:55:33,280 --> 00:55:34,340 Now why would that be. 1142 00:55:34,340 --> 00:55:36,980 Well Gabrieli at the time was a young 20 1143 00:55:36,980 --> 00:55:37,970 something year old guy. 1144 00:55:37,970 --> 00:55:42,090 This guy's got a 20 something year old long term memory. 1145 00:55:42,090 --> 00:55:44,660 A reasonable guess of somebody who looks more or less 1146 00:55:44,660 --> 00:55:48,600 familiar, but isn't in my current collection of 1147 00:55:48,600 --> 00:55:49,870 people I know well. 1148 00:55:49,870 --> 00:55:52,200 It's yeah maybe from high school, or 1149 00:55:52,200 --> 00:55:52,890 something like that. 1150 00:55:52,890 --> 00:55:58,310 We're about the same age as the 70 year old HM to the 20 1151 00:55:58,310 --> 00:56:01,570 something year old Gabrieli. 1152 00:56:01,570 --> 00:56:04,130 So those sort of things he does learn. 1153 00:56:04,130 --> 00:56:09,690 It's that new explicit long term memories that require the 1154 00:56:09,690 --> 00:56:14,020 hippocampus to be consolidated. 1155 00:56:14,020 --> 00:56:15,330 How long does this take? 1156 00:56:15,330 --> 00:56:18,240 Let me say a quick word about that, and then we'll take a 1157 00:56:18,240 --> 00:56:20,070 quick break. 1158 00:56:20,070 --> 00:56:24,630 You can measure the time course of consolidation by 1159 00:56:24,630 --> 00:56:28,190 doing an experiment of the following sort. 1160 00:56:28,190 --> 00:56:31,340 You can draw bad rats. 1161 00:56:34,390 --> 00:56:36,230 OK there's a rat. 1162 00:56:36,230 --> 00:56:37,880 He's on a little pedestal. 1163 00:56:37,880 --> 00:56:43,310 If you were a rat on a small pedestal not too far above the 1164 00:56:43,310 --> 00:56:46,350 floor, what are you going to do? 1165 00:56:46,350 --> 00:56:48,450 AUDIENCE: Jump off. 1166 00:56:48,450 --> 00:56:49,290 PROFESSOR: Jump off, right. 1167 00:56:49,290 --> 00:56:49,730 I'm a rat. 1168 00:56:49,730 --> 00:56:51,810 Man I got to go look around at stuff. 1169 00:56:51,810 --> 00:56:54,380 So I jump off. 1170 00:56:54,380 --> 00:56:58,600 Well what they didn't tell me back at home was that this 1171 00:56:58,600 --> 00:57:01,660 particular floor is electrified. 1172 00:57:01,660 --> 00:57:06,420 So when I jump off my little toes get an unpleasant, not 1173 00:57:06,420 --> 00:57:08,800 damaging, but an unpleasant shock. 1174 00:57:08,800 --> 00:57:12,000 So now the guy puts me back on here. 1175 00:57:12,000 --> 00:57:12,790 What do I do? 1176 00:57:12,790 --> 00:57:14,210 AUDIENCE: Jump off. 1177 00:57:14,210 --> 00:57:15,140 PROFESSOR: I don't jump off. 1178 00:57:15,140 --> 00:57:16,460 I'm not a stupid rat. 1179 00:57:16,460 --> 00:57:17,420 I'm just a rat. 1180 00:57:17,420 --> 00:57:19,690 And so I stay there. 1181 00:57:19,690 --> 00:57:21,000 OK, well. 1182 00:57:21,000 --> 00:57:26,610 I'm a rat who's in an experiment. 1183 00:57:26,610 --> 00:57:29,210 And the question is what percentage of the 1184 00:57:29,210 --> 00:57:37,860 time do I jump off? 1185 00:57:41,220 --> 00:57:48,390 And the variable here is after I jump off, and after I get my 1186 00:57:48,390 --> 00:57:52,770 toes shocked, I'm also going to get a dose of what's called 1187 00:57:52,770 --> 00:57:55,520 electroconvulsive shock. 1188 00:57:55,520 --> 00:57:59,900 Which is basically electrical current run through the brain. 1189 00:57:59,900 --> 00:58:03,100 You know you can kill people or animals that way. 1190 00:58:03,100 --> 00:58:05,220 Obviously that's not the point here. 1191 00:58:05,220 --> 00:58:08,920 This is to disrupt the on going electrical 1192 00:58:08,920 --> 00:58:10,760 activity of the brain. 1193 00:58:10,760 --> 00:58:12,590 Not to scramble the structure. 1194 00:58:12,590 --> 00:58:14,280 Not to cook proteins. 1195 00:58:14,280 --> 00:58:17,750 But just to disrupt any ongoing electrical activity. 1196 00:58:20,460 --> 00:58:24,410 So this is the delay to ECS. 1197 00:58:26,940 --> 00:58:30,180 Rat steps down when does he get his ECS? 1198 00:58:30,180 --> 00:58:34,310 If he gets the ECS immediately, put him back on 1199 00:58:34,310 --> 00:58:37,090 the platform. 1200 00:58:37,090 --> 00:58:39,760 Basically a 100% of the time the rat will step down the 1201 00:58:39,760 --> 00:58:40,800 next time to. 1202 00:58:40,800 --> 00:58:41,560 Why? 1203 00:58:41,560 --> 00:58:45,300 Not because the ECS is like fun or something like that. 1204 00:58:45,300 --> 00:58:49,740 But because the ECS has wiped out the memory. 1205 00:58:49,740 --> 00:58:53,540 The memory is not consolidated and the rat simply doesn't 1206 00:58:53,540 --> 00:58:55,950 know that his little feet got fried. 1207 00:58:55,950 --> 00:58:58,740 And he steps down again. 1208 00:58:58,740 --> 00:59:02,550 As the delay gets longer, you get an essentially exponential 1209 00:59:02,550 --> 00:59:09,040 looking curve with a time course in this case of seconds 1210 00:59:09,040 --> 00:59:13,730 that represents the time of consolidation of that memory. 1211 00:59:13,730 --> 00:59:16,380 At least to the point of supporting the behavior of not 1212 00:59:16,380 --> 00:59:18,010 stepping down. 1213 00:59:18,010 --> 00:59:22,840 but it's probably better to think it's not the case that 1214 00:59:22,840 --> 00:59:25,730 all memories consolidate within seconds. 1215 00:59:25,730 --> 00:59:29,410 There's evidence from human studies. 1216 00:59:29,410 --> 00:59:30,820 Now why would you do human 1217 00:59:30,820 --> 00:59:33,190 electroconvulsive shock studies? 1218 00:59:33,190 --> 00:59:37,770 The reason is that electroconvulsive shock in 1219 00:59:37,770 --> 00:59:42,990 humans is electroconvulsive therapy. 1220 00:59:42,990 --> 00:59:45,730 AUDIENCE: [LAUGHTER] 1221 00:59:45,730 --> 00:59:47,000 PROFESSOR: It sounds very odd. 1222 00:59:47,000 --> 00:59:47,890 It sounds very odd. 1223 00:59:47,890 --> 00:59:51,300 And those of you who've seen the now quite old movie to 1224 00:59:51,300 --> 00:59:53,360 kill, Not To Kill A Mockingbird, One Flew Over The 1225 00:59:53,360 --> 00:59:54,060 Cuckoo's Nest. 1226 00:59:54,060 --> 00:59:56,790 One of those bird movies. 1227 00:59:56,790 --> 01:00:00,370 One Flew Over The Cuckoo's Nest probably have a dim view 1228 01:00:00,370 --> 01:00:03,260 of electroconvulsive therapy. 1229 01:00:03,260 --> 01:00:05,190 But that's badly earned. 1230 01:00:05,190 --> 01:00:09,040 Int he case of intractable both depression. 1231 01:00:09,040 --> 01:00:14,820 Depression that's not being broken up successfully by 1232 01:00:14,820 --> 01:00:18,980 medication or by psychotherapy. 1233 01:00:18,980 --> 01:00:22,350 Severe depression of that sort carries with it a very serious 1234 01:00:22,350 --> 01:00:24,030 risk of mortality from suicide. 1235 01:00:24,030 --> 01:00:27,660 So this is not a, you know, casual disorder. 1236 01:00:27,660 --> 01:00:31,350 It turns out that electroconvulsive therapy is 1237 01:00:31,350 --> 01:00:37,750 often very effective in breaking up such depressions. 1238 01:00:37,750 --> 01:00:42,130 And is actually is useful part of the spectrum of treatments 1239 01:00:42,130 --> 01:00:47,120 that are available in this case. 1240 01:00:47,120 --> 01:00:50,590 Like most medical treatment, it's not entirely cost free. 1241 01:00:50,590 --> 01:00:55,010 It is the case that there are memory consequences of 1242 01:00:55,010 --> 01:00:56,450 electroconvulsive therapy. 1243 01:00:56,450 --> 01:01:00,080 It's part of what actually gives it a somewhat bad 1244 01:01:00,080 --> 01:01:01,720 reputation. 1245 01:01:01,720 --> 01:01:06,000 And you can actually see those memory consequences. 1246 01:01:06,000 --> 01:01:08,750 You can see traces of them going back years. 1247 01:01:14,320 --> 01:01:18,410 This is in studies for instance of memory for TV 1248 01:01:18,410 --> 01:01:19,850 shows seen only once. 1249 01:01:19,850 --> 01:01:21,920 You know sort of thing that you might dimly remember under 1250 01:01:21,920 --> 01:01:23,410 the best of circumstances. 1251 01:01:23,410 --> 01:01:27,580 These rather limited kind of memories can be damaged by a 1252 01:01:27,580 --> 01:01:29,780 electroconvulsive therapy years later. 1253 01:01:29,780 --> 01:01:34,620 As if this exponential has a very long tail. 1254 01:01:34,620 --> 01:01:39,010 And when you want to call it consolidated is perhaps a 1255 01:01:39,010 --> 01:01:41,330 function of may be something like how 1256 01:01:41,330 --> 01:01:42,580 important the memory is. 1257 01:01:42,580 --> 01:01:47,700 If I do this again, I'm going to be hurt right now. 1258 01:01:47,700 --> 01:01:50,800 That memory can be made solid enough relatively quickly that 1259 01:01:50,800 --> 01:01:52,240 you don't go and do it again. 1260 01:01:52,240 --> 01:01:57,060 On the other hand, if the memory is, was it Bridgette 1261 01:01:57,060 --> 01:02:02,460 Loves Bernie, or Benji. 1262 01:02:02,460 --> 01:02:04,400 All right that kind of good. 1263 01:02:04,400 --> 01:02:07,340 You know no stakes kind of memory is apparently still 1264 01:02:07,340 --> 01:02:08,430 vulnerable. 1265 01:02:08,430 --> 01:02:10,510 At least still partially vulnerable going back for a 1266 01:02:10,510 --> 01:02:11,520 very long time. 1267 01:02:11,520 --> 01:02:14,160 The hippocampus seems to be behaving like something like 1268 01:02:14,160 --> 01:02:15,420 scaffolding. 1269 01:02:15,420 --> 01:02:19,370 It holds the bits of the memory in place until it's 1270 01:02:19,370 --> 01:02:21,980 firm enough to stand on its own. 1271 01:02:21,980 --> 01:02:28,860 And then you can really call it in the sense of sort of a 1272 01:02:28,860 --> 01:02:30,620 permanent memory. 1273 01:02:30,620 --> 01:02:34,230 Now that doesn't solve the issue of getting 1274 01:02:34,230 --> 01:02:36,110 stuff out of memory. 1275 01:02:36,110 --> 01:02:41,120 And I will say a word or two about that momentarily. 1276 01:02:41,120 --> 01:02:43,280 But first you can wipe out your short term memory by 1277 01:02:43,280 --> 01:02:44,700 stretching for a second. 1278 01:02:58,160 --> 01:02:59,410 [SIDE CONVERSATION WITH STUDENT] 1279 01:04:23,160 --> 01:04:24,410 OK. 1280 01:04:28,350 --> 01:04:32,640 Looking at the handout, let me just make sure that I 1281 01:04:32,640 --> 01:04:35,630 explained the jargon here. 1282 01:04:35,630 --> 01:04:39,100 So a loss of memory as everybody knows is an amnesia. 1283 01:04:39,100 --> 01:04:46,250 A loss of memory for events prior to the point of injury 1284 01:04:46,250 --> 01:04:49,320 would be a retrograde amnesia. 1285 01:04:49,320 --> 01:04:52,220 That's not what HM has. 1286 01:04:52,220 --> 01:04:55,220 HM has an anterograde amnesia. 1287 01:04:55,220 --> 01:04:59,940 A problem with his memory subsequent to the lesion, 1288 01:04:59,940 --> 01:05:03,440 subsequent to the injury. 1289 01:05:03,440 --> 01:05:07,490 We won't do this is a demo, but if I were to take a 1290 01:05:07,490 --> 01:05:11,900 baseball bat and pop her on the head. 1291 01:05:11,900 --> 01:05:14,350 Well apart from the fact that I probably wouldn't be back on 1292 01:05:14,350 --> 01:05:15,600 a Thursday 1293 01:05:15,600 --> 01:05:18,910 AUDIENCE: [LAUGHTER] 1294 01:05:18,910 --> 01:05:23,980 PROFESSOR: When she came to, she might not remember having 1295 01:05:23,980 --> 01:05:27,540 actually been hit, because the set of memories from 1296 01:05:27,540 --> 01:05:31,100 immediately before that whop would have been wiped out in a 1297 01:05:31,100 --> 01:05:32,790 retrograde amnesia. 1298 01:05:32,790 --> 01:05:37,530 It is also possible that she wouldn't remember things for a 1299 01:05:37,530 --> 01:05:40,180 period of time after she had regained consciousness, if I 1300 01:05:40,180 --> 01:05:42,560 managed to knock her out completely. 1301 01:05:42,560 --> 01:05:44,810 There might be a period of time after she regained 1302 01:05:44,810 --> 01:05:49,140 consciousness before her memory would -- 1303 01:05:49,140 --> 01:05:51,630 she might have been talking and wandering around and say, 1304 01:05:51,630 --> 01:05:54,380 I'm fine, I'm fine. 1305 01:05:54,380 --> 01:05:56,550 But have no recollection of that either. 1306 01:05:56,550 --> 01:06:00,130 That would be an anterograde amnesia. 1307 01:06:00,130 --> 01:06:03,350 Sorry I didn't really mean to hit you on the head. 1308 01:06:03,350 --> 01:06:04,600 You'll be fine. 1309 01:06:07,460 --> 01:06:12,420 So let me say a few words in the remaining time about 1310 01:06:12,420 --> 01:06:17,160 retrieval from memory. 1311 01:06:17,160 --> 01:06:18,400 Like any of these topics, there's a 1312 01:06:18,400 --> 01:06:19,330 great deal to be said. 1313 01:06:19,330 --> 01:06:21,980 But probably the most important thing to understand 1314 01:06:21,980 --> 01:06:25,590 about retrieval from memory is that retrieval particularly if 1315 01:06:25,590 --> 01:06:34,890 any rich sort of memory is not the recovery, the replay, of 1316 01:06:34,890 --> 01:06:45,960 the tape for some true guaranteed image of what you 1317 01:06:45,960 --> 01:06:47,970 experienced or what you learned or 1318 01:06:47,970 --> 01:06:49,050 something like that. 1319 01:06:49,050 --> 01:06:53,490 This sort of image here of a scaffolding holding together 1320 01:06:53,490 --> 01:06:58,100 that memory is also an image of a notion that any memory is 1321 01:06:58,100 --> 01:07:02,220 probably stored multipley and in different bits of brain. 1322 01:07:02,220 --> 01:07:04,070 Different aspects of different bits of brain. 1323 01:07:04,070 --> 01:07:08,700 And if you want to recall it, that what you're doing is 1324 01:07:08,700 --> 01:07:11,940 reconstructing it. 1325 01:07:11,940 --> 01:07:15,900 You can probably get some intuition about this if you 1326 01:07:15,900 --> 01:07:21,670 asked yourself about famous stories about you in your 1327 01:07:21,670 --> 01:07:28,550 childhood that are famous in your family that you remember. 1328 01:07:28,550 --> 01:07:32,440 Not just the stories they tell about you from when you hit 1329 01:07:32,440 --> 01:07:33,830 with the bat and stuff like that. 1330 01:07:33,830 --> 01:07:36,550 But stories that you remember is sort of shaping incidents 1331 01:07:36,550 --> 01:07:37,450 in your childhood. 1332 01:07:37,450 --> 01:07:42,670 But that also get retold every time the family 1333 01:07:42,670 --> 01:07:44,660 gets together, right. 1334 01:07:44,660 --> 01:07:48,810 So in the Jewish tradition is a holiday called Passover. 1335 01:07:48,810 --> 01:07:51,220 The tradition is that you drink four cups of wine. 1336 01:07:51,220 --> 01:07:53,390 It's not really a brilliant tradition if you happen to be 1337 01:07:53,390 --> 01:07:56,260 about four years old. 1338 01:07:56,260 --> 01:08:01,140 And I vividly remember my sister getting into the nice 1339 01:08:01,140 --> 01:08:06,790 sweet wine that you drink and conching out at the table. 1340 01:08:06,790 --> 01:08:09,650 And I know whose house we were at. 1341 01:08:09,650 --> 01:08:13,220 This is a tale that has been retold to my sister's no doubt 1342 01:08:13,220 --> 01:08:18,410 the light every year for the intervening 40 plus years. 1343 01:08:18,410 --> 01:08:22,740 And it is absolutely unclear to me given what I 1344 01:08:22,740 --> 01:08:25,150 subsequently learned about memory, whether what I am 1345 01:08:25,150 --> 01:08:28,740 remembering is the original event at this point or 1346 01:08:28,740 --> 01:08:31,340 reconstruction based on all the family stories. 1347 01:08:31,340 --> 01:08:33,600 I think I can still remember the original thing, but 1348 01:08:33,600 --> 01:08:34,920 there's no real way of knowing. 1349 01:08:34,920 --> 01:08:41,890 What you are remembering is a reconstruction. 1350 01:08:41,890 --> 01:08:44,680 Before I talk more about that, let me load you up with some 1351 01:08:44,680 --> 01:08:47,210 more words. 1352 01:08:47,210 --> 01:08:50,010 These aren't going to be nonsense unless 1353 01:08:50,010 --> 01:08:51,790 I can't find them. 1354 01:08:51,790 --> 01:08:52,350 There we go. 1355 01:08:52,350 --> 01:08:54,580 There's some nice words. 1356 01:08:54,580 --> 01:08:56,980 Which words do I want to use. 1357 01:08:56,980 --> 01:08:57,780 Oh these are good words. 1358 01:08:57,780 --> 01:08:59,350 OK. 1359 01:08:59,350 --> 01:09:01,310 I'm going to read you a list of words, and then we'll do a 1360 01:09:01,310 --> 01:09:02,300 recognition test. 1361 01:09:02,300 --> 01:09:05,200 Don't have to write anything down this time. 1362 01:09:05,200 --> 01:09:07,730 I'm going to read you a list of words, and you're just 1363 01:09:07,730 --> 01:09:10,940 going to tell me the word was on the list or the word was 1364 01:09:10,940 --> 01:09:11,520 not on the list. 1365 01:09:11,520 --> 01:09:13,870 Ready? 1366 01:09:13,870 --> 01:09:16,970 So try to remember all of these. 1367 01:09:16,970 --> 01:09:19,350 Fury. 1368 01:09:19,350 --> 01:09:21,660 Rage. 1369 01:09:21,660 --> 01:09:24,080 Enrage. 1370 01:09:24,080 --> 01:09:26,400 Carpet. 1371 01:09:26,400 --> 01:09:28,630 Club. 1372 01:09:28,630 --> 01:09:31,170 Emotion. 1373 01:09:31,170 --> 01:09:33,460 Ire. 1374 01:09:33,460 --> 01:09:35,860 Mountain. 1375 01:09:35,860 --> 01:09:38,120 Fight. 1376 01:09:38,120 --> 01:09:40,420 Fear. 1377 01:09:40,420 --> 01:09:42,830 Mean. 1378 01:09:42,830 --> 01:09:45,110 Mad. 1379 01:09:45,110 --> 01:09:47,450 Place. 1380 01:09:47,450 --> 01:09:49,480 Hate. 1381 01:09:49,480 --> 01:09:51,350 And road. 1382 01:09:51,350 --> 01:09:51,740 OK. 1383 01:09:51,740 --> 01:09:54,930 You got all those nicely stored in memory? 1384 01:09:54,930 --> 01:09:55,290 OK. 1385 01:09:55,290 --> 01:09:57,780 So let's see, where's my test list here? 1386 01:09:57,780 --> 01:09:59,880 Looks like my test list. 1387 01:09:59,880 --> 01:10:01,100 OK. 1388 01:10:01,100 --> 01:10:02,970 Did you hear the word fight? 1389 01:10:02,970 --> 01:10:03,490 AUDIENCE: Yes. 1390 01:10:03,490 --> 01:10:04,950 PROFESSOR: Science? 1391 01:10:04,950 --> 01:10:05,840 AUDIENCE: No. 1392 01:10:05,840 --> 01:10:07,160 PROFESSOR: Rage? 1393 01:10:07,160 --> 01:10:08,240 AUDIENCE: Yes. 1394 01:10:08,240 --> 01:10:09,210 PROFESSOR: Fear? 1395 01:10:09,210 --> 01:10:10,260 AUDIENCE: Yes. 1396 01:10:10,260 --> 01:10:11,820 PROFESSOR: Emotion? 1397 01:10:11,820 --> 01:10:12,920 AUDIENCE: Yes. 1398 01:10:12,920 --> 01:10:14,120 PROFESSOR: Wrath? 1399 01:10:14,120 --> 01:10:15,080 AUDIENCE: No. 1400 01:10:15,080 --> 01:10:15,940 PROFESSOR: Club? 1401 01:10:15,940 --> 01:10:16,950 AUDIENCE: Yes. 1402 01:10:16,950 --> 01:10:17,990 PROFESSOR: Anger? 1403 01:10:17,990 --> 01:10:18,990 AUDIENCE: [UNINTELLIGIBLE] 1404 01:10:18,990 --> 01:10:22,250 PROFESSOR: Well. 1405 01:10:24,840 --> 01:10:25,780 Road? 1406 01:10:25,780 --> 01:10:26,700 AUDIENCE: Yes. 1407 01:10:26,700 --> 01:10:27,580 PROFESSOR: Ire? 1408 01:10:27,580 --> 01:10:28,560 AUDIENCE: Yes. 1409 01:10:28,560 --> 01:10:30,020 PROFESSOR: Hatred? 1410 01:10:30,020 --> 01:10:30,790 AUDIENCE: No. 1411 01:10:30,790 --> 01:10:33,340 PROFESSOR: OK. 1412 01:10:33,340 --> 01:10:34,730 How about enrage? 1413 01:10:34,730 --> 01:10:35,320 AUDIENCE: Yes. 1414 01:10:35,320 --> 01:10:36,630 PROFESSOR: OK. 1415 01:10:36,630 --> 01:10:41,500 so there were a couple in there where you could hear 1416 01:10:41,500 --> 01:10:45,310 people hesitating, or you could hear a 1417 01:10:45,310 --> 01:10:49,100 diversion of opinion. 1418 01:10:49,100 --> 01:10:52,700 And those are the critical elements there. 1419 01:10:52,700 --> 01:10:55,880 The most critical element in this case is the word anger, 1420 01:10:55,880 --> 01:10:59,000 which a fair number of people asserted it was on the list. 1421 01:10:59,000 --> 01:11:00,540 It was not in fact on the list. 1422 01:11:00,540 --> 01:11:03,800 And even those people who were asserting that it was not on 1423 01:11:03,800 --> 01:11:07,840 the list, there was a measurable 1424 01:11:07,840 --> 01:11:10,280 what's going on there. 1425 01:11:10,280 --> 01:11:13,270 This is in effect known as the Deese-Roediger-McDermott 1426 01:11:13,270 --> 01:11:17,770 effect or on the handout the Deese-Roediger-McDermott demo. 1427 01:11:17,770 --> 01:11:21,710 Roediger and Mcdermott discovered it. 1428 01:11:21,710 --> 01:11:24,030 And then like all good phenomena in psychology 1429 01:11:24,030 --> 01:11:27,600 discovered that Deese had published it 10 years prior. 1430 01:11:27,600 --> 01:11:30,460 And so it now goes by all of their names. 1431 01:11:30,460 --> 01:11:34,580 In any case, what they did, they were deliberately looking 1432 01:11:34,580 --> 01:11:36,930 for this, what they did was they -- where did my semantic 1433 01:11:36,930 --> 01:11:37,630 network go? 1434 01:11:37,630 --> 01:11:39,920 Oh here's my semantic network. 1435 01:11:39,920 --> 01:11:44,420 What they did was they took a word and they asked people to 1436 01:11:44,420 --> 01:11:46,260 give the associates to it. 1437 01:11:46,260 --> 01:11:49,910 So let's take the word anger. 1438 01:11:49,910 --> 01:11:52,830 What words come to mind when you think of the word anger? 1439 01:11:52,830 --> 01:11:57,150 Well fight, rage, enrage, emotion, ire. 1440 01:11:57,150 --> 01:11:58,200 This collection of words. 1441 01:11:58,200 --> 01:12:03,060 You then use a training list of words that has the 1442 01:12:03,060 --> 01:12:06,660 associates, but not the target word. 1443 01:12:06,660 --> 01:12:12,850 And what you're in effect doing is saying, animal, meow, 1444 01:12:12,850 --> 01:12:16,470 mouse, dog, fur. 1445 01:12:16,470 --> 01:12:20,310 And each time you do that, the spreading activation gets all 1446 01:12:20,310 --> 01:12:21,740 sorts of stuff, but all of these are 1447 01:12:21,740 --> 01:12:24,230 pointing towards cat. 1448 01:12:24,230 --> 01:12:27,480 And then later I say well was cats on the list? 1449 01:12:27,480 --> 01:12:30,360 And you go and look at that note and lo and behold it's 1450 01:12:30,360 --> 01:12:32,710 glowing softly in your mind. 1451 01:12:32,710 --> 01:12:36,930 And you say, yeah, yeah. 1452 01:12:36,930 --> 01:12:38,440 That was there. 1453 01:12:38,440 --> 01:12:42,160 And if you take one of other, it didn't particularly work in 1454 01:12:42,160 --> 01:12:44,620 this case, ire sometimes works for instance. 1455 01:12:44,620 --> 01:12:47,540 You take a word like ire people are no more sure that 1456 01:12:47,540 --> 01:12:51,120 ire was on the list than they are that 1457 01:12:51,120 --> 01:12:53,580 anger was on the list. 1458 01:12:53,580 --> 01:12:57,640 People are thoroughly in these experiments confused about 1459 01:12:57,640 --> 01:13:01,410 whether or not the target word was on the list. 1460 01:13:01,410 --> 01:13:05,770 Because you don't have access to the 1461 01:13:05,770 --> 01:13:07,130 truth about your memore. 1462 01:13:07,130 --> 01:13:08,850 You have the access to what you can manage to 1463 01:13:08,850 --> 01:13:10,200 reconstruct about it. 1464 01:13:10,200 --> 01:13:11,720 Now that's fine. 1465 01:13:11,720 --> 01:13:17,630 That sounds benign enough as a lab demo, but let's suppose 1466 01:13:17,630 --> 01:13:21,820 that you're not just sitting in some nice intro psych 1467 01:13:21,820 --> 01:13:25,760 class, but you're sitting in the witness box. 1468 01:13:25,760 --> 01:13:27,550 And somebody is saying, where were you on the 1469 01:13:27,550 --> 01:13:30,460 night of May 5th? 1470 01:13:30,460 --> 01:13:32,530 And expecting you to tell the truth. 1471 01:13:32,530 --> 01:13:34,880 Well of course you're going to try to tell the truth, but the 1472 01:13:34,880 --> 01:13:40,350 truth in this case is going to get shaped by the nature of 1473 01:13:40,350 --> 01:13:44,470 your particular semantic network, among other things. 1474 01:13:44,470 --> 01:13:46,800 Let's talk about that part a little bit first. 1475 01:13:46,800 --> 01:13:51,170 Famous experiment done first in the 40's replicated many 1476 01:13:51,170 --> 01:13:54,830 times since in many different ways, but let me describe the 1477 01:13:54,830 --> 01:13:56,640 original version. 1478 01:13:56,640 --> 01:14:01,920 You, and in this case you are a white male, are shown a 1479 01:14:01,920 --> 01:14:05,350 depiction of an altercation on a subway. 1480 01:14:05,350 --> 01:14:07,650 Two guys get into a fight. 1481 01:14:07,650 --> 01:14:10,900 In the 40's it was done as a hand drawing cartoon. 1482 01:14:10,900 --> 01:14:12,950 It's been replicated with, you know, snazzy video. 1483 01:14:12,950 --> 01:14:14,160 It doesn't matter. 1484 01:14:14,160 --> 01:14:18,650 Two guys get into a fight, or get into a sort of shouting 1485 01:14:18,650 --> 01:14:19,330 kind of argument. 1486 01:14:19,330 --> 01:14:24,770 At some point one guy pulls a knife, and then sometimes 1487 01:14:24,770 --> 01:14:27,810 subsequent we'll ask you about this, OK. 1488 01:14:34,580 --> 01:14:35,930 Well the question you're going to be asked is 1489 01:14:35,930 --> 01:14:37,220 who pulled the knife. 1490 01:14:37,220 --> 01:14:40,900 And the variable of interest is the race of the 1491 01:14:40,900 --> 01:14:41,640 participants. 1492 01:14:41,640 --> 01:14:44,110 So you got a nice little 2 by 2 here. 1493 01:14:44,110 --> 01:14:48,270 They can both be caucasian, they can both be African 1494 01:14:48,270 --> 01:14:51,550 American, or it can be crossed. 1495 01:14:51,550 --> 01:14:55,970 The interesting and disturbing finding is the number of times 1496 01:14:55,970 --> 01:15:00,390 when the knife is pulled by the white guy, it ends up in 1497 01:15:00,390 --> 01:15:04,130 memory in the hand of the black guy. 1498 01:15:04,130 --> 01:15:07,850 Now as I say, this has been replicated all sorts of 1499 01:15:07,850 --> 01:15:08,420 different ways. 1500 01:15:08,420 --> 01:15:12,130 This is not a, you know, white guys are all biggits 1501 01:15:12,130 --> 01:15:13,030 experiment. 1502 01:15:13,030 --> 01:15:16,860 This is we all have biases, and the interesting and 1503 01:15:16,860 --> 01:15:21,600 disturbing thing is that they can actually interfere with 1504 01:15:21,600 --> 01:15:25,060 what we think of as, you know, our nice clear memory. 1505 01:15:25,060 --> 01:15:27,390 These people weren't sitting there saying, you know, oh 1506 01:15:27,390 --> 01:15:32,190 yeah I'm being paid, you know, 1940 $0.10 an hour to be in 1507 01:15:32,190 --> 01:15:33,120 this psych experiment. 1508 01:15:33,120 --> 01:15:37,920 Why don't I see if I can stick it to an imaginary black guy. 1509 01:15:37,920 --> 01:15:39,400 No. 1510 01:15:39,400 --> 01:15:44,550 They were presumably being as honest as they could be. 1511 01:15:44,550 --> 01:15:48,070 But the structure of their memory. 1512 01:15:48,070 --> 01:15:52,690 Which included things like who do I think might pull a knife, 1513 01:15:52,690 --> 01:15:54,660 influenced what they recalled, what they 1514 01:15:54,660 --> 01:15:56,540 pulled out of memory. 1515 01:15:56,540 --> 01:15:58,200 And we can impose this structure from 1516 01:15:58,200 --> 01:16:00,780 the outside as well. 1517 01:16:00,780 --> 01:16:03,090 So suppose we do the following experiment. 1518 01:16:03,090 --> 01:16:07,250 This is an experiment also done many different ways at 1519 01:16:07,250 --> 01:16:10,020 this point. 1520 01:16:10,020 --> 01:16:13,170 Elizabeth Loftus is the name associated with 1521 01:16:13,170 --> 01:16:16,300 this line of work. 1522 01:16:16,300 --> 01:16:18,110 You're going to see another film. 1523 01:16:18,110 --> 01:16:20,420 This time you're going to watch a car crash. 1524 01:16:20,420 --> 01:16:25,880 The red car is going to get into an accident 1525 01:16:25,880 --> 01:16:27,160 with the blue car. 1526 01:16:27,160 --> 01:16:29,620 All right. 1527 01:16:29,620 --> 01:16:29,870 OK. 1528 01:16:29,870 --> 01:16:30,510 You see this. 1529 01:16:30,510 --> 01:16:31,350 You're going to be asked about it. 1530 01:16:31,350 --> 01:16:33,570 We'll show all of you guys the same film. 1531 01:16:33,570 --> 01:16:37,740 And now we'll ask the question three different ways. 1532 01:16:37,740 --> 01:16:46,890 At what speed did the red car bump into the blue car? 1533 01:16:46,890 --> 01:16:51,200 At what speed did the red car hit the blue car? 1534 01:16:51,200 --> 01:16:55,990 At what speed did the red car smashed into the blue car? 1535 01:16:55,990 --> 01:16:57,510 That's the only manipulation here. 1536 01:16:57,510 --> 01:17:01,130 You all saw the same video. 1537 01:17:01,130 --> 01:17:04,600 So what do you figure the difference is? 1538 01:17:04,600 --> 01:17:07,710 What's the result look like? 1539 01:17:07,710 --> 01:17:12,490 So who gives the higher speed responses? 1540 01:17:12,490 --> 01:17:12,580 AUDIENCE: The smashed guy. 1541 01:17:12,580 --> 01:17:14,600 PROFESSOR: The smashed guys Lawyers know this. 1542 01:17:14,600 --> 01:17:20,170 This is why courts attempts to avoid leading question. 1543 01:17:20,170 --> 01:17:22,910 But it's very hard to avoid. 1544 01:17:22,910 --> 01:17:27,160 But, you know, that's a pretty subtle kind of lead. 1545 01:17:27,160 --> 01:17:28,930 But it's not a subtle kind of effect. 1546 01:17:28,930 --> 01:17:32,630 It's about as I recall a 15 MPH kind of effect, which is a 1547 01:17:32,630 --> 01:17:37,080 sort of thing that has an impact, you should pardon the 1548 01:17:37,080 --> 01:17:40,030 expression, on what a jury is going to think about whether 1549 01:17:40,030 --> 01:17:44,600 or not you, the driver of that red car, was at fault or not. 1550 01:17:44,600 --> 01:17:49,200 If you were screaming through the intersection at 45 MPH 1551 01:17:49,200 --> 01:17:52,280 that's worse than if you're screaming through the 1552 01:17:52,280 --> 01:17:55,920 intersection at 30 MPH or something like that. 1553 01:17:55,920 --> 01:18:02,650 So your memory can be influenced by the way that you 1554 01:18:02,650 --> 01:18:04,970 are asked about that memory. 1555 01:18:04,970 --> 01:18:08,430 This shows up not just on the stand. 1556 01:18:08,430 --> 01:18:12,350 Well actually a version of this shows up beautifully in 1557 01:18:12,350 --> 01:18:18,140 the experience of most people at some point who come to MIT. 1558 01:18:18,140 --> 01:18:23,170 Most people who get to MIT are smart people, and they would 1559 01:18:23,170 --> 01:18:25,010 do well to remember that. 1560 01:18:25,010 --> 01:18:28,080 Because most people sometime, like within the first few 1561 01:18:28,080 --> 01:18:32,780 weeks of being at MIT, receive a data point of some sort that 1562 01:18:32,780 --> 01:18:35,900 suggests that this might not be true. 1563 01:18:35,900 --> 01:18:36,160 Right? 1564 01:18:36,160 --> 01:18:38,660 Like the first calculus test or something like that. 1565 01:18:38,660 --> 01:18:42,580 You've never gotten a score below 99.8 on a math 1566 01:18:42,580 --> 01:18:43,590 test in your life. 1567 01:18:43,590 --> 01:18:48,170 You get back the first test in 18 or whatever it was, and you 1568 01:18:48,170 --> 01:18:51,290 know, 10 cool. 1569 01:18:51,290 --> 01:18:53,570 Oh man it's not 10 out of 10 is it? 1570 01:18:53,570 --> 01:18:54,180 AUDIENCE: [LAUGHTER] 1571 01:18:54,180 --> 01:18:55,380 PROFESSOR: You know this is really bad. 1572 01:18:55,380 --> 01:18:58,670 And so that's a depressing data point. 1573 01:18:58,670 --> 01:19:02,340 But a surprising number of people come to the conclusion 1574 01:19:02,340 --> 01:19:05,680 from that, that they not only are stupid 1575 01:19:05,680 --> 01:19:07,840 but always were stupid. 1576 01:19:07,840 --> 01:19:15,750 And probably unlovely and generally, you know, bad, 1577 01:19:15,750 --> 01:19:16,950 horrible individuals. 1578 01:19:16,950 --> 01:19:21,790 And you should remember this is one of these sort of 1579 01:19:21,790 --> 01:19:25,040 context effects. 1580 01:19:25,040 --> 01:19:26,440 That may be a little over stating it. 1581 01:19:26,440 --> 01:19:28,210 But look you know -- 1582 01:19:28,210 --> 01:19:30,100 and this has actually been tested with clinical 1583 01:19:30,100 --> 01:19:32,470 populations of depressive people who go from being 1584 01:19:32,470 --> 01:19:34,440 depressed to being undepressed. 1585 01:19:34,440 --> 01:19:38,130 Suppose you ask somebody who's depressed clinically or 1586 01:19:38,130 --> 01:19:40,920 otherwise, you wake up in the morning it's pouring 1587 01:19:40,920 --> 01:19:42,770 cold rain out there. 1588 01:19:42,770 --> 01:19:45,220 The problems that did you get done because you fell asleep 1589 01:19:45,220 --> 01:19:47,740 in there, and there are now drool marks on it. 1590 01:19:47,740 --> 01:19:50,790 And not only is there drook mark, there's a little heart 1591 01:19:50,790 --> 01:19:53,510 shaped note next to it saying, I'm leaving 1592 01:19:53,510 --> 01:19:54,920 you for your roommate. 1593 01:19:54,920 --> 01:19:55,690 AUDIENCE: [LAUGHTER] 1594 01:19:55,690 --> 01:19:57,510 PROFESSOR: Right? 1595 01:19:57,510 --> 01:20:05,490 So if I asked you at that point how was your childhood? 1596 01:20:05,490 --> 01:20:07,250 Not how do you feel, but what kind of 1597 01:20:07,250 --> 01:20:09,340 childhood did you have? 1598 01:20:09,340 --> 01:20:12,130 You had not maybe a miserable childhood. 1599 01:20:12,130 --> 01:20:15,530 But to exaggerate, you know, you grew up in a closet. 1600 01:20:18,940 --> 01:20:21,140 Later on if you're feeling better I ask you about the 1601 01:20:21,140 --> 01:20:22,330 same childhood. 1602 01:20:22,330 --> 01:20:25,930 The childhood has magically improved. 1603 01:20:25,930 --> 01:20:26,490 All right. 1604 01:20:26,490 --> 01:20:30,660 We'll see you on Thursday if I remember correctly.