1 00:00:00,080 --> 00:00:01,800 The following content is provided 2 00:00:01,800 --> 00:00:04,030 under a Creative Commons license. 3 00:00:04,030 --> 00:00:06,880 Your support will help MIT OpenCourseWare continue 4 00:00:06,880 --> 00:00:10,740 to offer high-quality educational resources for free. 5 00:00:10,740 --> 00:00:13,350 To make a donation, or view additional materials 6 00:00:13,350 --> 00:00:17,237 from hundreds of MIT courses, visit MIT OpenCourseWare 7 00:00:17,237 --> 00:00:17,862 at ocw.mit.edu. 8 00:00:22,880 --> 00:00:26,390 PROFESSOR: We had started to talk about learning, 9 00:00:26,390 --> 00:00:29,510 beginning with habituation, which 10 00:00:29,510 --> 00:00:34,390 led to a discussion of the effects of novel stimuli. 11 00:00:34,390 --> 00:00:37,770 And I had mentioned the autonomic as well as 12 00:00:37,770 --> 00:00:40,480 behavioral effects of novel stimuli. 13 00:00:45,500 --> 00:00:47,280 At the end of the session, I said 14 00:00:47,280 --> 00:00:52,260 that autonomic responses to stimulus novelty 15 00:00:52,260 --> 00:00:55,500 are the same as the responses measured by a lie detector. 16 00:00:55,500 --> 00:00:58,470 So what can explain that? 17 00:01:07,660 --> 00:01:10,360 This is the way I explain it. 18 00:01:10,360 --> 00:01:14,980 That the autonomic nervous system 19 00:01:14,980 --> 00:01:17,340 and the orienting mechanisms of the midbrain 20 00:01:17,340 --> 00:01:19,145 respond to any conflict. 21 00:01:24,900 --> 00:01:28,290 Usually, when you're dealing with stimulus novelty, 22 00:01:28,290 --> 00:01:33,120 the conflict you're dealing with is between the expected 23 00:01:33,120 --> 00:01:36,025 stimulus and what is actually perceived. 24 00:01:39,560 --> 00:01:43,920 We expect things all the time, because we 25 00:01:43,920 --> 00:01:46,700 have an internal model of the world. 26 00:01:46,700 --> 00:01:48,200 We have a visual model of the world. 27 00:01:48,200 --> 00:01:51,640 So every time you flick your eyes-- 28 00:01:51,640 --> 00:01:56,020 if I suddenly move my eyes over here, there's nothing novel, 29 00:01:56,020 --> 00:01:57,920 I expect. 30 00:01:57,920 --> 00:02:01,550 I have a pretty good model of this room. 31 00:02:01,550 --> 00:02:04,580 And so I always expect what I'm going to see. 32 00:02:04,580 --> 00:02:12,990 And if I don't see what I expect-- if suddenly there's 33 00:02:12,990 --> 00:02:16,390 somebody different over there I didn't expect-- now 34 00:02:16,390 --> 00:02:17,290 there's nobody there. 35 00:02:17,290 --> 00:02:19,240 But if there suddenly was somebody, 36 00:02:19,240 --> 00:02:23,590 I would have an autonomic response. 37 00:02:23,590 --> 00:02:27,190 Of course, my behavior would change too, in various ways. 38 00:02:31,190 --> 00:02:34,190 So that's the kind of conflict we're talking about. 39 00:02:34,190 --> 00:02:36,520 Not the usual meaning, perhaps, of conflict. 40 00:02:36,520 --> 00:02:37,680 But it's a difference. 41 00:02:37,680 --> 00:02:40,610 We're always making this comparison between what we see 42 00:02:40,610 --> 00:02:44,000 and what we have expected to see. 43 00:02:44,000 --> 00:02:47,200 But the conflict can also be between 44 00:02:47,200 --> 00:02:50,270 an internal representation of what 45 00:02:50,270 --> 00:02:57,500 is known to be the case-- what's true-- and a cognitive system. 46 00:02:57,500 --> 00:03:00,082 And it could be a system that generates output, 47 00:03:00,082 --> 00:03:01,040 like the speech system. 48 00:03:04,540 --> 00:03:06,540 And they can differ. 49 00:03:06,540 --> 00:03:08,220 Why would the difference arise? 50 00:03:08,220 --> 00:03:14,690 Well, because these memories and expectancies 51 00:03:14,690 --> 00:03:17,770 are cognitive systems. 52 00:03:17,770 --> 00:03:21,750 And they're closely tied with motivational systems. 53 00:03:21,750 --> 00:03:24,280 So you can have two different systems active 54 00:03:24,280 --> 00:03:26,250 at the same time. 55 00:03:26,250 --> 00:03:28,840 And when they're in conflict, they're 56 00:03:28,840 --> 00:03:30,780 going to generate autonomic responses. 57 00:03:36,200 --> 00:03:39,060 You should ask right away, is that always true? 58 00:03:39,060 --> 00:03:44,220 And I point out here that for some individuals, 59 00:03:44,220 --> 00:03:47,960 these systems can be pretty isolated from each other, 60 00:03:47,960 --> 00:03:51,390 and not cause any actual conflict. 61 00:03:51,390 --> 00:03:57,090 And we call such individuals psychopathic or sociopathic. 62 00:03:57,090 --> 00:03:59,790 That is, they don't feel any conflict. 63 00:04:02,810 --> 00:04:05,410 And it's possible that people can also 64 00:04:05,410 --> 00:04:10,580 learn to suppress autonomic responses-- although I can tell 65 00:04:10,580 --> 00:04:12,015 you, it's exceedingly difficult. 66 00:04:18,589 --> 00:04:20,420 Let's look at some other learning topics. 67 00:04:23,300 --> 00:04:26,380 We know that scientists have often 68 00:04:26,380 --> 00:04:34,180 chosen to study learning in very simple animals like the sea 69 00:04:34,180 --> 00:04:37,275 slug, and snails, and the fruit fly. 70 00:04:40,180 --> 00:04:42,950 And there are other insects that have been used. 71 00:04:42,950 --> 00:04:49,870 They've been studied in some crustaceans as well. 72 00:04:49,870 --> 00:04:52,440 What are the reasons for that? 73 00:04:52,440 --> 00:04:55,890 Now we know that the animal that dominates as a model 74 00:04:55,890 --> 00:04:58,680 is the mouse. 75 00:04:58,680 --> 00:05:04,360 But that hasn't always been, and it 76 00:05:04,360 --> 00:05:08,090 was exceedingly difficult to get at the cell 77 00:05:08,090 --> 00:05:12,430 on the molecular level in studying learning, 78 00:05:12,430 --> 00:05:14,530 before relatively recently. 79 00:05:14,530 --> 00:05:18,470 And for that reason, many people have studied other animals, 80 00:05:18,470 --> 00:05:20,140 like this animal. 81 00:05:20,140 --> 00:05:26,150 This is the sea slug-- the California sea slug. 82 00:05:26,150 --> 00:05:26,650 Aplysia. 83 00:05:29,440 --> 00:05:32,800 So why do they choose such an animal? 84 00:05:32,800 --> 00:05:36,450 Why do they choose certain insects? 85 00:05:36,450 --> 00:05:38,370 There's several reasons. 86 00:05:38,370 --> 00:05:40,670 One is, the nervous system is simpler. 87 00:05:43,200 --> 00:05:48,900 There are fewer neurons, and it's especially important 88 00:05:48,900 --> 00:05:51,190 that many times, these nervous systems 89 00:05:51,190 --> 00:05:54,190 have individually identified cells. 90 00:05:54,190 --> 00:05:58,270 So that you can identify the very same cells from one animal 91 00:05:58,270 --> 00:05:58,800 to the next. 92 00:06:01,790 --> 00:06:05,140 And in many cases, they formed very similar 93 00:06:05,140 --> 00:06:07,210 synaptic connections when those are studied. 94 00:06:07,210 --> 00:06:13,730 So you can look at the very same connections 95 00:06:13,730 --> 00:06:16,430 in different members of the same species. 96 00:06:16,430 --> 00:06:20,510 And that's very important in science, to be able to repeat. 97 00:06:20,510 --> 00:06:23,750 If you're studying the details of what happens at the synapse, 98 00:06:23,750 --> 00:06:27,070 you're looking at molecular changes-- 99 00:06:27,070 --> 00:06:28,700 that's very important. 100 00:06:28,700 --> 00:06:32,280 And most of these animals, in simple learning situations, 101 00:06:32,280 --> 00:06:33,880 are very predictable. 102 00:06:33,880 --> 00:06:39,260 Much more predictable than the higher vertebrates. 103 00:06:39,260 --> 00:06:44,230 So those are the reasons people choose those animals. 104 00:06:44,230 --> 00:06:49,820 And people also have chosen to study some simpler 105 00:06:49,820 --> 00:06:54,480 forms of learning, like classical conditioning. 106 00:06:54,480 --> 00:06:56,090 And Scott discusses this. 107 00:07:01,610 --> 00:07:05,750 It's often called Pavlovian conditioning, or Type I 108 00:07:05,750 --> 00:07:10,400 conditioning-- as opposed to instrumental conditioning, 109 00:07:10,400 --> 00:07:17,260 or problem solving involved in, for example, trial 110 00:07:17,260 --> 00:07:19,880 and error learning. 111 00:07:19,880 --> 00:07:21,940 Now, you should know the paradigm 112 00:07:21,940 --> 00:07:23,660 for classical conditioning. 113 00:07:23,660 --> 00:07:24,370 It's very simple. 114 00:07:29,870 --> 00:07:32,700 I just want to point out the historical point here 115 00:07:32,700 --> 00:07:36,760 that Ivan Pavlov and his followers in Russia 116 00:07:36,760 --> 00:07:43,070 interpreted all learning in terms of conditioned reflexes. 117 00:07:43,070 --> 00:07:49,180 And they wrote books about it, books explaining 118 00:07:49,180 --> 00:07:52,550 human psychology in these terms. 119 00:07:52,550 --> 00:07:55,610 Their teaching-- the Pavlovian teaching-- 120 00:07:55,610 --> 00:08:00,030 became state policy in the Soviet Union. 121 00:08:00,030 --> 00:08:03,060 And scientists who didn't go along 122 00:08:03,060 --> 00:08:05,720 were in danger of persecution. 123 00:08:05,720 --> 00:08:10,160 I had the good fortune of winning 124 00:08:10,160 --> 00:08:14,330 a prize when I was young student, to go over 125 00:08:14,330 --> 00:08:16,220 to Moscow to a meeting. 126 00:08:18,870 --> 00:08:23,550 I'm also hesitant to tell you how long ago it was. 127 00:08:23,550 --> 00:08:26,430 It was actually in 1966. 128 00:08:26,430 --> 00:08:28,500 I remember I was pretty young at the time. 129 00:08:28,500 --> 00:08:32,500 And I felt there weren't other people my age 130 00:08:32,500 --> 00:08:34,350 going, except one other American, 131 00:08:34,350 --> 00:08:36,120 because we had both won the same prize. 132 00:08:36,120 --> 00:08:38,440 A guy from California. 133 00:08:38,440 --> 00:08:43,700 And I still remember going to these meetings. 134 00:08:43,700 --> 00:08:46,470 And there would be some American and European speakers, 135 00:08:46,470 --> 00:08:50,330 and some Russian speakers. 136 00:08:50,330 --> 00:08:55,230 And the Russian speakers, invariably, 137 00:08:55,230 --> 00:09:02,970 would, say, introduce their talk with something about Pavlov. 138 00:09:02,970 --> 00:09:07,140 I asked them about it, and they explained to me 139 00:09:07,140 --> 00:09:12,990 that, yes, we have to do that. 140 00:09:12,990 --> 00:09:16,830 But then they would look around a little bit, 141 00:09:16,830 --> 00:09:22,020 and then say, in a lower voice, we don't actually 142 00:09:22,020 --> 00:09:26,025 believe it all anymore, but this is what was required. 143 00:09:32,660 --> 00:09:37,250 Anyway, Scott, when he discusses it, makes a mistake. 144 00:09:37,250 --> 00:09:40,290 And I think it's good for you to be 145 00:09:40,290 --> 00:09:42,580 able to read your textbooks critically. 146 00:09:42,580 --> 00:09:48,030 No textbooks are written totally free of mistakes. 147 00:09:48,030 --> 00:09:52,300 And I actually quote the statement here. 148 00:09:52,300 --> 00:09:54,795 So look at that, and tell me what's wrong with it. 149 00:10:01,550 --> 00:10:06,340 If you know the paradigm for classical conditioning, 150 00:10:06,340 --> 00:10:10,340 you would be able to see this. 151 00:10:10,340 --> 00:10:13,340 So what's the basis for classical conditioning? 152 00:10:15,890 --> 00:10:19,680 You have a response that occurs automatically. 153 00:10:19,680 --> 00:10:22,330 It's a reflex response. 154 00:10:22,330 --> 00:10:26,900 Take, for example, leg withdrawal. 155 00:10:26,900 --> 00:10:33,480 What causes leg withdrawal because of an inherited, 156 00:10:33,480 --> 00:10:35,940 innate reflex? 157 00:10:35,940 --> 00:10:38,780 Well, something painful. 158 00:10:38,780 --> 00:10:42,500 You step on something painful. 159 00:10:42,500 --> 00:10:43,720 You pull your foot back. 160 00:10:43,720 --> 00:10:45,610 And that's automatic. 161 00:10:45,610 --> 00:10:48,180 You become aware that you did that, 162 00:10:48,180 --> 00:10:51,380 and you become aware of the pain after you're already 163 00:10:51,380 --> 00:10:53,760 pulling your leg back. 164 00:10:53,760 --> 00:10:57,260 That's the withdrawal reflex. 165 00:10:57,260 --> 00:11:02,310 And you can elicit in your friend-- if you hold the hand 166 00:11:02,310 --> 00:11:06,680 and pinch him at an unexpected time, 167 00:11:06,680 --> 00:11:09,200 he will pull his hand back. 168 00:11:09,200 --> 00:11:14,325 And then he will yell out, what did you do that for? 169 00:11:14,325 --> 00:11:15,450 But he's already pulled it. 170 00:11:15,450 --> 00:11:21,402 So this is an unconditional, or unconditioned, reflex. 171 00:11:21,402 --> 00:11:23,360 The original meaning-- the better translation-- 172 00:11:23,360 --> 00:11:26,090 is actually unconditional. 173 00:11:26,090 --> 00:11:31,225 And the unconditioned stimulus is that painful stimulus. 174 00:11:31,225 --> 00:11:33,866 It can be heat. 175 00:11:33,866 --> 00:11:39,830 It can be other kinds of sensory induced pain. 176 00:11:39,830 --> 00:11:41,260 Basically, an intense stimulus. 177 00:11:44,590 --> 00:11:50,570 So then, how do you condition that in the classical manner? 178 00:11:50,570 --> 00:11:55,910 You cause a stimulus to occur just 179 00:11:55,910 --> 00:11:59,940 before the unconditional stimulus. 180 00:11:59,940 --> 00:12:03,640 So the unconditional stimulus, or unconditioned stimulus, 181 00:12:03,640 --> 00:12:05,450 is the UCS. 182 00:12:05,450 --> 00:12:07,830 The conditioned stimulus is the CS. 183 00:12:07,830 --> 00:12:11,230 The conditioned stimulus has to occur first. 184 00:12:11,230 --> 00:12:14,130 Because then, the animal or the person 185 00:12:14,130 --> 00:12:19,170 learns that that conditioned stimulus is always 186 00:12:19,170 --> 00:12:22,340 going to be followed by-- it's not a cognitive thing. 187 00:12:22,340 --> 00:12:24,073 It happens pretty automatically. 188 00:12:28,430 --> 00:12:29,670 So what will happen? 189 00:12:29,670 --> 00:12:33,712 Well, after a number of pairings of the conditioned stimulus 190 00:12:33,712 --> 00:12:35,170 and unconditioned stimulus, they'll 191 00:12:35,170 --> 00:12:38,630 withdraw the leg just in response 192 00:12:38,630 --> 00:12:41,710 to the conditioned stimulus. 193 00:12:41,710 --> 00:12:42,485 Could be a tone. 194 00:12:45,450 --> 00:12:48,360 All stimuli don't work equally well, 195 00:12:48,360 --> 00:12:52,200 although the early writers-- the early students 196 00:12:52,200 --> 00:12:55,025 of classical conditioning-- thought just any stimulus 197 00:12:55,025 --> 00:12:56,640 would work. 198 00:12:56,640 --> 00:13:00,600 It's been the ecologists who've shown that's not actually true. 199 00:13:00,600 --> 00:13:06,380 There are specific differences in different animals. 200 00:13:06,380 --> 00:13:09,390 Anyway, he mixes up the direction. 201 00:13:09,390 --> 00:13:13,860 He says the application of the US, and then the CS immediately 202 00:13:13,860 --> 00:13:15,240 after it, every five minutes. 203 00:13:15,240 --> 00:13:17,500 He's got it reversed. 204 00:13:17,500 --> 00:13:19,790 That wouldn't make any sense. 205 00:13:19,790 --> 00:13:21,280 You see, the conditioned stimulus-- 206 00:13:21,280 --> 00:13:26,330 it's something that normally predicts the stimulus. 207 00:13:26,330 --> 00:13:29,150 What happens if you always present the stimulus just 208 00:13:29,150 --> 00:13:30,536 afterwards? 209 00:13:30,536 --> 00:13:31,910 I won't lead to any conditioning. 210 00:13:35,120 --> 00:13:37,810 And he says it leads to a much stronger response 211 00:13:37,810 --> 00:13:39,060 to the unconditioned stimulus. 212 00:13:39,060 --> 00:13:39,559 No. 213 00:13:39,559 --> 00:13:40,170 Not at all. 214 00:13:40,170 --> 00:13:42,695 It leads to a response to the conditioned stimulus. 215 00:13:45,340 --> 00:13:47,000 And this is the paradigm, then. 216 00:13:51,800 --> 00:13:53,560 You can say, US. 217 00:13:53,560 --> 00:13:55,830 In Scott, it's the unconditioned stimulus, 218 00:13:55,830 --> 00:13:59,790 but UCS is more common. 219 00:13:59,790 --> 00:14:02,740 So originally, before any learning, 220 00:14:02,740 --> 00:14:06,010 the UCS leads to the unconditioned response, 221 00:14:06,010 --> 00:14:07,000 the UCR. 222 00:14:07,000 --> 00:14:09,340 And after learning, the conditional stimulus, 223 00:14:09,340 --> 00:14:14,440 or conditioned stimulus, leads to the conditioned response. 224 00:14:14,440 --> 00:14:16,650 Which is usually almost identical 225 00:14:16,650 --> 00:14:19,540 to the unconditioned response. 226 00:14:19,540 --> 00:14:21,730 If you measure the details, you might find out 227 00:14:21,730 --> 00:14:24,605 it's not exactly, in every way, identical. 228 00:14:27,510 --> 00:14:30,030 And I'm just giving some examples 229 00:14:30,030 --> 00:14:33,250 here of laboratory studies that conditioned 230 00:14:33,250 --> 00:14:35,610 leg withdrawal that I was talking about. 231 00:14:35,610 --> 00:14:39,110 The innate flexion reflex in response to painful stimulus. 232 00:14:39,110 --> 00:14:40,840 They've done condition eye closure. 233 00:14:40,840 --> 00:14:41,965 That's very common. 234 00:14:41,965 --> 00:14:43,506 They usually use a little puff of air 235 00:14:43,506 --> 00:14:45,290 as the unconditional stimulus. 236 00:14:45,290 --> 00:14:48,060 And you can get conditioned eye blink. 237 00:14:51,370 --> 00:14:54,940 And you can train someone, or an animal, 238 00:14:54,940 --> 00:14:58,390 to blink their eyes every time a little tone comes on, 239 00:14:58,390 --> 00:14:59,360 or some other stimulus. 240 00:15:03,490 --> 00:15:08,480 And of course, we are conditioned just 241 00:15:08,480 --> 00:15:10,470 by our experience. 242 00:15:10,470 --> 00:15:13,550 Anything that predicts anything coming at our eye, 243 00:15:13,550 --> 00:15:14,340 we will blink. 244 00:15:24,300 --> 00:15:26,110 Now, there's other kinds of learning. 245 00:15:26,110 --> 00:15:28,380 This is the topic of social learning. 246 00:15:28,380 --> 00:15:29,880 When I say, in your own development, 247 00:15:29,880 --> 00:15:35,370 how have you shown learning by imitation or by mimicry? 248 00:15:35,370 --> 00:15:39,310 And what's the difference between learning by imitation 249 00:15:39,310 --> 00:15:41,520 and learning by mimicry? 250 00:15:41,520 --> 00:15:43,980 It's actually pretty hard, but there 251 00:15:43,980 --> 00:15:48,960 is a difference that's spelled out in Scott. 252 00:15:48,960 --> 00:15:50,550 Does anybody know the difference? 253 00:15:55,440 --> 00:15:58,540 How do birds learn to sing? 254 00:15:58,540 --> 00:16:04,790 Male songbirds will pick up the songs of adult males 255 00:16:04,790 --> 00:16:06,720 that are singing. 256 00:16:06,720 --> 00:16:09,310 And they learn those songs before they actually 257 00:16:09,310 --> 00:16:10,150 start signing. 258 00:16:13,080 --> 00:16:15,360 Now, when they're learning, there's no reward. 259 00:16:15,360 --> 00:16:17,450 So we call it mimicry. 260 00:16:17,450 --> 00:16:19,440 They learn by mimicry, but they don't even 261 00:16:19,440 --> 00:16:22,130 have to start mimicking initially, 262 00:16:22,130 --> 00:16:24,390 while it's happening. 263 00:16:24,390 --> 00:16:27,820 In many cases, they do. 264 00:16:27,820 --> 00:16:33,590 But with imitation-- like, a young cat 265 00:16:33,590 --> 00:16:36,830 will imitate its mother in hunting. 266 00:16:36,830 --> 00:16:40,630 It's got the fixed action patterns to hunt, 267 00:16:40,630 --> 00:16:42,850 but without the imitation of its mother, 268 00:16:42,850 --> 00:16:45,410 it doesn't acquire normal motivation to hunt, 269 00:16:45,410 --> 00:16:48,750 as we've learned. 270 00:16:48,750 --> 00:16:52,930 So with imitation, there's immediate reward; 271 00:16:52,930 --> 00:16:57,550 with mimicry, there's no tangible immediate reward. 272 00:16:57,550 --> 00:16:59,930 That's the way we discriminate them. 273 00:16:59,930 --> 00:17:05,329 And they've been observed in many non-human primates. 274 00:17:05,329 --> 00:17:06,310 Not just monkeys. 275 00:17:06,310 --> 00:17:09,470 We think monkeys are the ones that imitate and mimic 276 00:17:09,470 --> 00:17:10,930 the most. 277 00:17:10,930 --> 00:17:13,400 But in fact it occurs in many other animals. 278 00:17:13,400 --> 00:17:16,270 How have we brought that up in the class already? 279 00:17:16,270 --> 00:17:21,089 Well I mentioned kittens imitating their mother. 280 00:17:21,089 --> 00:17:25,740 That's certainly a non-human animal. 281 00:17:25,740 --> 00:17:28,180 But it occurred then in jackdaws. 282 00:17:28,180 --> 00:17:30,290 Remember the Konrad Lorenz story of how 283 00:17:30,290 --> 00:17:33,590 they learn to identify enemies? 284 00:17:36,100 --> 00:17:36,990 How do they do it? 285 00:17:44,195 --> 00:17:46,070 They imitate-- or mimic. 286 00:17:46,070 --> 00:17:49,980 It's a little hard to know; there is some social reward 287 00:17:49,980 --> 00:17:53,070 here, so I suppose we can call it imitation. 288 00:17:53,070 --> 00:17:57,780 But the adults, when they detect a dangerous animal that 289 00:17:57,780 --> 00:18:03,880 has been seen dangling something black that could be a jackdaw, 290 00:18:03,880 --> 00:18:05,570 the young learn from them. 291 00:18:05,570 --> 00:18:09,680 And you know, from the adult encounters with that person 292 00:18:09,680 --> 00:18:12,630 or animal-- like, it could be a hunter. 293 00:18:12,630 --> 00:18:14,910 They learn to recognize. 294 00:18:14,910 --> 00:18:16,670 They learn by their calls. 295 00:18:16,670 --> 00:18:19,580 They make a lot of noise when there are enemies around. 296 00:18:19,580 --> 00:18:23,220 And the young pick that up, and they imitate the parents. 297 00:18:23,220 --> 00:18:24,690 And they learn. 298 00:18:24,690 --> 00:18:27,300 And this has been studied in the corvids. 299 00:18:27,300 --> 00:18:28,800 Crows in particular. 300 00:18:28,800 --> 00:18:31,370 There's experimental studies of this. 301 00:18:31,370 --> 00:18:33,110 Of the specificity of their learning. 302 00:18:33,110 --> 00:18:35,020 And the interesting thing about those studies 303 00:18:35,020 --> 00:18:37,930 of crows that have been done more recently, 304 00:18:37,930 --> 00:18:42,390 since the earlier studies, was that that knowledge 305 00:18:42,390 --> 00:18:50,680 can be passed on to the young that have never 306 00:18:50,680 --> 00:18:56,256 seen the particular face of the person 307 00:18:56,256 --> 00:18:57,505 that they've learned to avoid. 308 00:19:00,310 --> 00:19:05,120 They can somehow teach their young about this. 309 00:19:05,120 --> 00:19:10,530 Songbirds-- we know they learn songs by mimicking adults. 310 00:19:10,530 --> 00:19:13,290 Scott mentions that in newly hatched chicks-- 311 00:19:13,290 --> 00:19:15,180 this was a surprise to me; I didn't 312 00:19:15,180 --> 00:19:18,080 realize that these old newborn chickens can 313 00:19:18,080 --> 00:19:22,280 learn by imitating adults. 314 00:19:22,280 --> 00:19:26,520 But the adults that they observe pecking food 315 00:19:26,520 --> 00:19:29,280 will avoid certain food because it's bitter, 316 00:19:29,280 --> 00:19:33,520 and feed on things that are not bitter. 317 00:19:33,520 --> 00:19:36,930 And the chicks learn that. 318 00:19:36,930 --> 00:19:41,820 The chicks don't have to learn everything by trial and error. 319 00:19:41,820 --> 00:19:45,010 They do learn just by watching. 320 00:19:45,010 --> 00:19:46,520 Just by watching adults. 321 00:19:46,520 --> 00:19:50,700 It didn't fit the people that believed 322 00:19:50,700 --> 00:19:52,425 everything was conditioned learning. 323 00:19:55,720 --> 00:20:00,520 They learn by imitation and mimicry as well. 324 00:20:00,520 --> 00:20:02,410 And there's very good evidence for that. 325 00:20:02,410 --> 00:20:05,475 Experimental studies, for example, of the chickens. 326 00:20:05,475 --> 00:20:07,795 The little chicks that Scott mentions. 327 00:20:10,760 --> 00:20:15,270 And there's various other things you can find, too. 328 00:20:15,270 --> 00:20:15,770 OK. 329 00:20:15,770 --> 00:20:19,750 So let's move to a slightly different topic, 330 00:20:19,750 --> 00:20:22,800 involving a lot of learning-- the topic of navigation. 331 00:20:28,030 --> 00:20:32,690 And I also gave you a reading from Niko Tinbergen, 332 00:20:32,690 --> 00:20:35,020 whose little book, Curious Naturalists, 333 00:20:35,020 --> 00:20:38,890 gives very interesting descriptions of some 334 00:20:38,890 --> 00:20:42,250 of his experiments-- not just with gulls, but also 335 00:20:42,250 --> 00:20:44,280 with insects. 336 00:20:44,280 --> 00:20:48,555 He studied landmark orientation in the female digger wasp. 337 00:20:52,330 --> 00:20:55,430 He studied several different species of these, usually 338 00:20:55,430 --> 00:21:00,953 sand wasps that make underground nests. 339 00:21:04,700 --> 00:21:06,310 He studied them with his students. 340 00:21:06,310 --> 00:21:09,820 And not just by observing them. 341 00:21:09,820 --> 00:21:11,900 They observed them in great detail. 342 00:21:11,900 --> 00:21:13,520 They spend many, many hours observing 343 00:21:13,520 --> 00:21:16,170 details of how these wasps do it. 344 00:21:16,170 --> 00:21:18,276 But then they also did specific experiments. 345 00:21:21,360 --> 00:21:27,870 And this is from the chapter that I asked you to read. 346 00:21:27,870 --> 00:21:37,255 In this case, the wasp had his nest over here. 347 00:21:40,050 --> 00:21:44,170 And they had observe the walls wasp bringing larvae back 348 00:21:44,170 --> 00:21:49,510 to the nest to feed-- they brought dead insects, rather, 349 00:21:49,510 --> 00:21:53,240 back to their larvae in the next. 350 00:21:53,240 --> 00:21:58,740 And the way they did the experiment 351 00:21:58,740 --> 00:22:02,870 was, they surrounded that nest with pine cones, 352 00:22:02,870 --> 00:22:07,980 and left those pine cones there for an extended period. 353 00:22:07,980 --> 00:22:11,210 Because they wanted to see-- does the female wasp 354 00:22:11,210 --> 00:22:13,535 use landmarks like that, like the position 355 00:22:13,535 --> 00:22:16,175 of those pine cones, to actually find her nest? 356 00:22:16,175 --> 00:22:18,590 Or is she using other means? 357 00:22:18,590 --> 00:22:20,470 So then, of course, the test was simply 358 00:22:20,470 --> 00:22:23,070 to move the pine cones as you see here. 359 00:22:23,070 --> 00:22:27,800 And this is what happened, as he shows in his drawing. 360 00:22:27,800 --> 00:22:31,390 The initial place the wasp looks is 361 00:22:31,390 --> 00:22:34,050 right in the middle of those pine cones, where the nest used 362 00:22:34,050 --> 00:22:36,780 to be-- even though it's been displaced. 363 00:22:36,780 --> 00:22:40,025 Now, of course, you could use other landmarks, too, 364 00:22:40,025 --> 00:22:41,070 to position it. 365 00:22:41,070 --> 00:22:46,860 But prominent landmarks are usually the ones being used. 366 00:22:46,860 --> 00:22:48,120 This is from another study. 367 00:22:50,680 --> 00:22:52,380 I think we talked about this before. 368 00:22:57,110 --> 00:22:59,585 The colors aren't showing up over here. 369 00:22:59,585 --> 00:23:01,860 They're showing up a little better there. 370 00:23:01,860 --> 00:23:04,510 But anyway, here, in this particular wasp, 371 00:23:04,510 --> 00:23:07,290 he's got five different nests. 372 00:23:07,290 --> 00:23:12,610 At least, by the fourth day, he had five nests. 373 00:23:12,610 --> 00:23:16,030 These are the days here. 374 00:23:16,030 --> 00:23:18,080 And when they first observed this wasp, 375 00:23:18,080 --> 00:23:20,865 he was bringing food to one nest. 376 00:23:23,470 --> 00:23:25,860 Then she came back to the nest. 377 00:23:25,860 --> 00:23:28,610 And the open circle there indicates that all she did 378 00:23:28,610 --> 00:23:29,650 is open it and look. 379 00:23:32,240 --> 00:23:36,150 And the next step was she closed it up. 380 00:23:36,150 --> 00:23:37,580 Close it up and left. 381 00:23:37,580 --> 00:23:40,290 She never came back again to that nest. 382 00:23:40,290 --> 00:23:43,540 Instead, she went to another nest here. 383 00:23:43,540 --> 00:23:44,950 This was all on the first day. 384 00:23:44,950 --> 00:23:47,050 And just opened it and observed. 385 00:23:47,050 --> 00:23:49,490 Then she flew away. 386 00:23:49,490 --> 00:23:52,920 But then, rather than go get a larvae and bring it back, 387 00:23:52,920 --> 00:23:54,170 she didn't do that right away. 388 00:23:54,170 --> 00:23:56,030 She dug another nest. 389 00:23:56,030 --> 00:23:59,710 The third nest, here. 390 00:23:59,710 --> 00:24:05,050 She dug it, and then she got a dead caterpillar, 391 00:24:05,050 --> 00:24:10,630 and she brought it there and laid an egg, 392 00:24:10,630 --> 00:24:13,738 and left the caterpillar next to the egg. 393 00:24:20,800 --> 00:24:22,840 It was at the beginning of the second day 394 00:24:22,840 --> 00:24:25,590 that she came with the caterpillar and laid the egg. 395 00:24:25,590 --> 00:24:28,180 She had dug the nest at the end of the first day. 396 00:24:28,180 --> 00:24:32,980 And then, later in that day, she came back to that nest two, 397 00:24:32,980 --> 00:24:39,220 observed it, and then she started bringing dead insects-- 398 00:24:39,220 --> 00:24:42,460 worms, or caterpillars-- to that nest. 399 00:24:42,460 --> 00:24:45,940 And that's what the green dots here-- and if it's open, 400 00:24:45,940 --> 00:24:49,700 all she did is observe what is she observing? 401 00:24:49,700 --> 00:24:54,270 Well, she's observing if the caterpillar has been eaten. 402 00:24:54,270 --> 00:24:58,130 And also, she observes the state of the larvae. 403 00:24:58,130 --> 00:24:59,250 How developed is it? 404 00:24:59,250 --> 00:25:09,295 It is ready for metamorphosis, so she can close the nest? 405 00:25:12,930 --> 00:25:14,930 And so here, she just observes it. 406 00:25:14,930 --> 00:25:19,690 Then she deposits three more dead animals. 407 00:25:19,690 --> 00:25:20,910 And then she closes it. 408 00:25:20,910 --> 00:25:22,850 Now there's enough. 409 00:25:22,850 --> 00:25:24,970 She's obviously able to make that judgment. 410 00:25:24,970 --> 00:25:28,710 And then she goes back to nest three, which she had only 411 00:25:28,710 --> 00:25:31,880 been to after digging-- she had only been two once, 412 00:25:31,880 --> 00:25:33,420 when she laid the egg. 413 00:25:33,420 --> 00:25:34,600 She just observes it. 414 00:25:34,600 --> 00:25:37,760 And then she went and dug a fourth nest. 415 00:25:37,760 --> 00:25:43,080 Then she came back, observed that second one, and then, 416 00:25:43,080 --> 00:25:47,440 soon, came back with food and left it there. 417 00:25:47,440 --> 00:25:53,250 Then she went and dug a fifth nest, and laid an egg, 418 00:25:53,250 --> 00:25:55,090 and provisioned it. 419 00:25:55,090 --> 00:25:58,620 And then, she flew off to observe nest three. 420 00:26:02,820 --> 00:26:05,660 You'd say, why did she have to come back? 421 00:26:05,660 --> 00:26:09,590 Well, she's obviously having to monitor. 422 00:26:09,590 --> 00:26:10,610 Is the larvae alive? 423 00:26:13,140 --> 00:26:19,080 Has it eaten the initial caterpillar 424 00:26:19,080 --> 00:26:20,710 that she left there? 425 00:26:20,710 --> 00:26:25,630 And if it needs more food, she starts bringing food to it. 426 00:26:25,630 --> 00:26:28,500 And that extended into the sixth day, where she brought food 427 00:26:28,500 --> 00:26:30,900 to it twice, and then she closed it up. 428 00:26:30,900 --> 00:26:34,120 And then she went to nest five, which 429 00:26:34,120 --> 00:26:36,210 had laid an egg at before. 430 00:26:36,210 --> 00:26:40,700 She looks at it, and then she brings a caterpillar to it. 431 00:26:40,700 --> 00:26:42,180 Here, she disappeared for a while; 432 00:26:42,180 --> 00:26:45,860 I think they don't know what she did. 433 00:26:45,860 --> 00:26:50,160 That led then to, finally, closure of that fifth nest. 434 00:26:50,160 --> 00:26:55,600 So she's aware of multiple nests at the same time. 435 00:26:55,600 --> 00:27:02,590 As you see here, she moves from two to three-- four, 436 00:27:02,590 --> 00:27:05,250 and then five here. 437 00:27:05,250 --> 00:27:08,440 Once she closes it, she seems to remember 438 00:27:08,440 --> 00:27:10,080 that's been closed, so she doesn't 439 00:27:10,080 --> 00:27:11,120 come back there anymore. 440 00:27:14,140 --> 00:27:16,750 And there's these three clear stages 441 00:27:16,750 --> 00:27:18,830 in the way she deals with. 442 00:27:18,830 --> 00:27:21,180 She has to remember the location of the nest. 443 00:27:21,180 --> 00:27:26,810 She has to remember the contents after she inspects it. 444 00:27:26,810 --> 00:27:29,460 And obviously, she has to be able to remember 445 00:27:29,460 --> 00:27:34,450 the status of her offspring there in the nest. 446 00:27:37,680 --> 00:27:44,910 So then he did more research on how she finds those nests. 447 00:27:44,910 --> 00:27:49,190 And in this case, this is a wasp-- a sand wasp-- 448 00:27:49,190 --> 00:27:50,460 that has the nest here. 449 00:27:50,460 --> 00:27:54,480 And I've colored that position red. 450 00:27:54,480 --> 00:27:56,470 They did experiments with the wasp. 451 00:27:56,470 --> 00:28:00,500 They would capture her and take her various distances 452 00:28:00,500 --> 00:28:03,520 in various directions from that nest, and release her. 453 00:28:03,520 --> 00:28:08,980 And you can see-- if they released her in this vicinity, 454 00:28:08,980 --> 00:28:13,990 except for one case here, she found her way back to the nest. 455 00:28:13,990 --> 00:28:17,790 In one case here-- they don't always 456 00:28:17,790 --> 00:28:19,400 know what landmark she's using. 457 00:28:19,400 --> 00:28:21,600 But she didn't go directly to it. 458 00:28:21,600 --> 00:28:23,590 But then she found it. 459 00:28:23,590 --> 00:28:24,670 But notice here. 460 00:28:24,670 --> 00:28:28,910 If they went a little bit further, 461 00:28:28,910 --> 00:28:30,615 she generally didn't find it. 462 00:28:33,620 --> 00:28:38,100 If they released the wasp up here, 463 00:28:38,100 --> 00:28:46,280 they found that in spite of it being a greater distance, 464 00:28:46,280 --> 00:28:48,250 she was able to find her way. 465 00:28:48,250 --> 00:28:53,460 Not directly, but by this more circuitous route. 466 00:28:53,460 --> 00:28:58,950 And this is a very large pine tree here, in this ridge. 467 00:28:58,950 --> 00:29:01,490 She's able to find the nest. 468 00:29:01,490 --> 00:29:04,590 So obviously, it depends on what they remember 469 00:29:04,590 --> 00:29:07,440 from their experience of the terrain. 470 00:29:07,440 --> 00:29:14,880 These wasps do fly up and spend time observing the landmarks. 471 00:29:14,880 --> 00:29:18,830 And in their earlier experience, if you take them to an area 472 00:29:18,830 --> 00:29:22,200 that they haven't observed, then they frequently 473 00:29:22,200 --> 00:29:23,440 won't find their way back. 474 00:29:23,440 --> 00:29:25,550 Or they'll at least wander around a lot 475 00:29:25,550 --> 00:29:26,910 before they will find it. 476 00:29:29,710 --> 00:29:34,000 So this is another one of the specific experiments 477 00:29:34,000 --> 00:29:35,560 about the use of landmarks. 478 00:29:35,560 --> 00:29:44,910 Here, in the initial period, there 479 00:29:44,910 --> 00:29:48,860 was food here they gave the animal. 480 00:29:48,860 --> 00:29:52,180 She's carrying back something towards the nest. 481 00:29:52,180 --> 00:29:55,260 And there's a row of trees here. 482 00:29:55,260 --> 00:29:58,360 They're actually artificial, so they can be moved. 483 00:29:58,360 --> 00:30:03,030 And all she had to do was follow the line of trees 484 00:30:03,030 --> 00:30:04,120 to get to her nest. 485 00:30:04,120 --> 00:30:09,480 And over a period of time, she did that regularly. 486 00:30:09,480 --> 00:30:14,460 So then they shifted the position of the trees. 487 00:30:14,460 --> 00:30:16,370 Now, what does she do? 488 00:30:16,370 --> 00:30:17,860 And you can see here. 489 00:30:17,860 --> 00:30:21,915 She pretty much follows those trees. 490 00:30:25,290 --> 00:30:27,550 You can see here-- she's wandering around. 491 00:30:27,550 --> 00:30:29,575 Here, she doesn't find the nest at all. 492 00:30:32,140 --> 00:30:36,000 And another time, they moved the trees over here. 493 00:30:36,000 --> 00:30:37,340 And you can see what she does. 494 00:30:37,340 --> 00:30:39,700 She expects to see the nest here, 495 00:30:39,700 --> 00:30:43,890 and so she starts flying back and forth. 496 00:30:43,890 --> 00:30:48,230 Here, she actually did find the nest. 497 00:30:48,230 --> 00:30:50,940 Because obviously, it's adaptive, if they 498 00:30:50,940 --> 00:30:53,820 don't find it, to start a random search 499 00:30:53,820 --> 00:30:58,840 pattern-- which is what she's doing. 500 00:30:58,840 --> 00:31:06,170 So sometimes, they do learn to adapt to this moving of trees. 501 00:31:06,170 --> 00:31:08,930 But in many cases, they don't-- like here. 502 00:31:08,930 --> 00:31:10,640 So then, what do they do? 503 00:31:10,640 --> 00:31:13,130 They would have to build a new nest and start over again. 504 00:31:16,400 --> 00:31:20,100 So this is the use of landmark orientation. 505 00:31:20,100 --> 00:31:22,910 And they found specific differences sometimes. 506 00:31:22,910 --> 00:31:27,550 In this case, Adriance and Berens-- 507 00:31:27,550 --> 00:31:29,720 they thought they were studying the same species, 508 00:31:29,720 --> 00:31:33,450 but when you compare the details of their notes, 509 00:31:33,450 --> 00:31:36,050 you see that they're actually bringing 510 00:31:36,050 --> 00:31:38,840 different kinds of food to the nest. 511 00:31:38,840 --> 00:31:46,230 One's bringing here the sawflies-- 512 00:31:46,230 --> 00:31:48,940 another type of larvae for their own larvae. 513 00:31:48,940 --> 00:31:53,930 And this guy observed only one nest at a time, 514 00:31:53,930 --> 00:31:58,160 whereas this one, like the one we talked about a minute ago, 515 00:31:58,160 --> 00:31:59,105 had multiple nests. 516 00:32:02,700 --> 00:32:04,690 Telescoped, indicating that they might 517 00:32:04,690 --> 00:32:06,380 be at different periods of development, 518 00:32:06,380 --> 00:32:10,990 but she took care of multiple broods at the same time. 519 00:32:10,990 --> 00:32:16,180 They had slightly different breeding periods, 520 00:32:16,180 --> 00:32:19,700 and they had different details in the way they closed up 521 00:32:19,700 --> 00:32:20,320 the nest. 522 00:32:20,320 --> 00:32:22,140 The way they handled the sand. 523 00:32:22,140 --> 00:32:26,410 And what they did with sand when they opened it. 524 00:32:26,410 --> 00:32:27,244 Did they scatter it? 525 00:32:27,244 --> 00:32:29,660 Did they take it back to where they were getting the sand? 526 00:32:29,660 --> 00:32:31,930 And so forth. 527 00:32:31,930 --> 00:32:36,070 So these details-- at least, Adriance, 528 00:32:36,070 --> 00:32:39,100 when he read the Berens result-- he 529 00:32:39,100 --> 00:32:41,510 realized that they must be different. 530 00:32:41,510 --> 00:32:46,770 And even though to Adriance and Berens, 531 00:32:46,770 --> 00:32:50,330 they looked the same, he decided, 532 00:32:50,330 --> 00:32:56,160 why not send them to a guy highly trained in morphology? 533 00:32:56,160 --> 00:33:00,490 Get him to examine these wasps. 534 00:33:00,490 --> 00:33:02,450 That was done, and the conclusion 535 00:33:02,450 --> 00:33:06,020 was, they're actually different species. 536 00:33:06,020 --> 00:33:10,060 So there, they've been named separate species-- not because 537 00:33:10,060 --> 00:33:13,560 of, initially, the structure; that came later. 538 00:33:13,560 --> 00:33:16,350 Only after the behavioral differences were discovered. 539 00:33:16,350 --> 00:33:18,340 Because remember, the behavioral patterns, 540 00:33:18,340 --> 00:33:20,750 the fixed action patterns are inherited, 541 00:33:20,750 --> 00:33:23,080 just like structure of the body. 542 00:33:28,890 --> 00:33:31,970 We gave an example of navigation when 543 00:33:31,970 --> 00:33:41,750 we talked about the Lawrence question about, 544 00:33:41,750 --> 00:33:45,680 can a fixed action pattern actually be maladaptive? 545 00:33:45,680 --> 00:33:48,840 And why would they become maladaptive? 546 00:33:48,840 --> 00:33:51,340 And he brought up the problem a full 547 00:33:51,340 --> 00:33:55,080 of photopollution in hatchling sea turtles. 548 00:33:55,080 --> 00:33:57,155 And this was done by Scott here. 549 00:34:01,430 --> 00:34:08,679 And people have used knowledge of how these little turtles 550 00:34:08,679 --> 00:34:12,830 orient towards the sea to prevent 551 00:34:12,830 --> 00:34:15,170 the kind of photopollution that causes 552 00:34:15,170 --> 00:34:18,800 them to fail to reach the sea. 553 00:34:18,800 --> 00:34:25,201 And we know that the little sea turtle hatchlings-- 554 00:34:25,201 --> 00:34:27,659 I've pointed this out-- they tend to orient towards the sea 555 00:34:27,659 --> 00:34:31,449 because they're attracted to the blue light. 556 00:34:31,449 --> 00:34:34,239 And the sea obviously doesn't have 557 00:34:34,239 --> 00:34:36,670 trees impeding the view and the light, 558 00:34:36,670 --> 00:34:39,360 so that will be where the most of the blue light's 559 00:34:39,360 --> 00:34:42,759 coming from-- towards the sea, instead 560 00:34:42,759 --> 00:34:45,472 of towards inland areas. 561 00:34:45,472 --> 00:34:47,055 And they will crawl in that direction. 562 00:34:50,360 --> 00:34:52,400 Because during evolution, the strongest light 563 00:34:52,400 --> 00:34:53,199 was always seaward. 564 00:34:53,199 --> 00:34:56,469 Words 565 00:34:56,469 --> 00:35:00,915 But then electric lights show up in a town. 566 00:35:00,915 --> 00:35:03,730 And the electric lights can be blue lights. 567 00:35:03,730 --> 00:35:05,540 Neon lights are often blue. 568 00:35:08,770 --> 00:35:14,130 Hot fires can make them start crawling towards the fire, even 569 00:35:14,130 --> 00:35:17,300 at night, just because of the innate behavior 570 00:35:17,300 --> 00:35:20,440 that they've inherited. 571 00:35:20,440 --> 00:35:22,640 So, the way conservationists have 572 00:35:22,640 --> 00:35:26,550 solved the problem is, after studying it and realizing what 573 00:35:26,550 --> 00:35:37,000 very simple cues these turtles are following, 574 00:35:37,000 --> 00:35:39,590 they studied their reactions to other colors 575 00:35:39,590 --> 00:35:42,910 and found that if they used yellow lights or red lights, 576 00:35:42,910 --> 00:35:45,290 they're not as attractive as the blue lights. 577 00:35:48,570 --> 00:35:50,365 They also don't like flashing lights. 578 00:35:52,920 --> 00:35:56,730 Generally the sun doesn't flash, and the sky doesn't flash. 579 00:36:00,010 --> 00:36:06,290 So they started using flashing orange lights. 580 00:36:06,290 --> 00:36:12,610 They talked to the people at the companies 581 00:36:12,610 --> 00:36:15,670 that were using the neon and doing the advertising. 582 00:36:15,670 --> 00:36:17,850 Putting up the billboards with neon signs. 583 00:36:17,850 --> 00:36:20,520 Putting it at their stores. 584 00:36:20,520 --> 00:36:23,600 If they were near beaches, they persuaded 585 00:36:23,600 --> 00:36:26,730 them to change the color and to use more flashing 586 00:36:26,730 --> 00:36:30,350 lights, especially orange lights. 587 00:36:30,350 --> 00:36:36,330 And also, local towns made rules about when 588 00:36:36,330 --> 00:36:40,490 beach fires could not be used. 589 00:36:40,490 --> 00:36:43,740 The period of hatching of the little sea turtles. 590 00:36:43,740 --> 00:36:47,240 And it's led to a great increase in the survival of these sea 591 00:36:47,240 --> 00:36:48,400 turtles. 592 00:36:48,400 --> 00:36:51,850 Just by taking these very simple stimuli 593 00:36:51,850 --> 00:36:54,810 that they used to get to the ocean into account. 594 00:36:58,420 --> 00:37:02,240 Let's talk about dead reckoning. 595 00:37:02,240 --> 00:37:03,750 It's a funny term. 596 00:37:03,750 --> 00:37:07,220 It actually comes from the phrase deduced reckoning. 597 00:37:11,930 --> 00:37:18,410 If humans leave one place and walk a certain distance away-- 598 00:37:18,410 --> 00:37:21,430 say, in a forest-- but then they want 599 00:37:21,430 --> 00:37:24,540 to get back to where they started from, 600 00:37:24,540 --> 00:37:31,290 and they haven't left a string behind them-- 601 00:37:31,290 --> 00:37:34,660 as we know from fairy tales, that 602 00:37:34,660 --> 00:37:38,540 can be done-- how do they get back? 603 00:37:38,540 --> 00:37:44,100 Well, humans can often remember directions pretty well. 604 00:37:44,100 --> 00:37:46,100 They keep track of directions. 605 00:37:46,100 --> 00:37:49,230 So when they turn, they know they've turned. 606 00:37:49,230 --> 00:37:51,315 And they know roughly how far they've walked. 607 00:37:55,730 --> 00:37:58,710 Some people are a lot better at this than others. 608 00:37:58,710 --> 00:38:02,415 But that's a kind of deduced reckoning, using internal cues. 609 00:38:08,160 --> 00:38:10,770 Especially vestibular cues. 610 00:38:10,770 --> 00:38:15,020 But ants that have been studied are generally 611 00:38:15,020 --> 00:38:18,850 using external cues for direction. 612 00:38:18,850 --> 00:38:22,570 The position of the sun, and visual landmarks. 613 00:38:26,950 --> 00:38:30,910 Let's skip down, because we'll talk about the Sahara Desert 614 00:38:30,910 --> 00:38:32,690 and desert ants in a minute. 615 00:38:32,690 --> 00:38:35,550 Scott gives the example of wood ants 616 00:38:35,550 --> 00:38:42,160 that use skyline features as visual landmarks. 617 00:38:42,160 --> 00:38:47,290 There haven't been many studies, but there are studies actually 618 00:38:47,290 --> 00:38:50,410 more extensive than the ones I did with Syrian hamsters, 619 00:38:50,410 --> 00:38:53,990 where you showed that Syrian hamsters can 620 00:38:53,990 --> 00:38:56,670 find their way back from a foraging trip. 621 00:38:56,670 --> 00:38:59,970 They go some distance from their tunnels. 622 00:38:59,970 --> 00:39:00,980 So it was a problem. 623 00:39:00,980 --> 00:39:03,910 How do they get back? 624 00:39:03,910 --> 00:39:07,040 Are they just using olfaction? 625 00:39:07,040 --> 00:39:08,360 And it turns out-- no. 626 00:39:08,360 --> 00:39:11,840 They use visual landmarks when they can. 627 00:39:11,840 --> 00:39:14,720 They use a long horizon. 628 00:39:14,720 --> 00:39:16,680 Rabbits do the same thing. 629 00:39:16,680 --> 00:39:21,880 In fact, rabbits have denser receptors in their retina, 630 00:39:21,880 --> 00:39:24,570 in the part of the retina that represents the horizon. 631 00:39:24,570 --> 00:39:28,321 I have not noticed this in studies of the hamster, 632 00:39:28,321 --> 00:39:31,370 but hamsters definitely use these horizon cues. 633 00:39:31,370 --> 00:39:37,890 And probably, many-- some of the terrain they come from 634 00:39:37,890 --> 00:39:42,000 is mountainous, so probably they need 635 00:39:42,000 --> 00:39:49,420 to be able to use cues well above the flat horizon as well. 636 00:39:49,420 --> 00:39:51,220 But anyway, they use visual cues, 637 00:39:51,220 --> 00:39:53,822 unless the levels of light get so 638 00:39:53,822 --> 00:39:59,030 low it's really almost completely dark. 639 00:39:59,030 --> 00:40:03,360 Then they will switch and use olfactory cues. 640 00:40:03,360 --> 00:40:07,139 So they're capable of using multiple cues 641 00:40:07,139 --> 00:40:08,055 for their orientation. 642 00:40:11,730 --> 00:40:16,270 Scott also uses the example of another wood ant. 643 00:40:16,270 --> 00:40:19,270 He uses vision to judge distance, 644 00:40:19,270 --> 00:40:22,520 by using landmarks at a certain height. 645 00:40:22,520 --> 00:40:25,810 And he will adjust his distance from the landmarks. 646 00:40:25,810 --> 00:40:31,170 So the angle is at a certain point. 647 00:40:31,170 --> 00:40:33,740 And by varying the height of that landmark, 648 00:40:33,740 --> 00:40:43,070 they can change the navigation of the wood ant. 649 00:40:43,070 --> 00:40:46,470 Like, the elevation of the top of a wall on one side. 650 00:40:49,530 --> 00:40:55,540 The Sahara Desert ants are particularly interesting. 651 00:40:55,540 --> 00:41:00,850 And this is especially because there are some species of ant 652 00:41:00,850 --> 00:41:05,420 that go out in the hottest part of the day, which 653 00:41:05,420 --> 00:41:07,860 by itself is very surprising. 654 00:41:07,860 --> 00:41:14,510 But they're in an area with basically no landmarks. 655 00:41:14,510 --> 00:41:20,970 And yet, they go out from the nest. 656 00:41:20,970 --> 00:41:23,360 They wander around. 657 00:41:23,360 --> 00:41:24,810 They seem to meander around. 658 00:41:24,810 --> 00:41:26,600 What they're looking for? 659 00:41:26,600 --> 00:41:29,220 Dead insects, especially. 660 00:41:29,220 --> 00:41:34,450 Small dead animals that they can use as food. 661 00:41:34,450 --> 00:41:39,990 And if they find it, then they take it. 662 00:41:39,990 --> 00:41:43,250 And they don't follow their trail back. 663 00:41:43,250 --> 00:41:45,880 And some ants are known-- other species of ant; 664 00:41:45,880 --> 00:41:50,910 forest ants, for example-- will often even olfactory trail. 665 00:41:50,910 --> 00:41:54,235 Or they will remember by other cues how to get back, 666 00:41:54,235 --> 00:41:56,360 but then they'll leave a trail when they come back. 667 00:41:56,360 --> 00:41:57,776 And other ants will leave the nest 668 00:41:57,776 --> 00:41:59,740 and follow that olfactory trail. 669 00:41:59,740 --> 00:42:02,770 But not these desert ants. 670 00:42:02,770 --> 00:42:05,610 They also seem to know exactly how to get back. 671 00:42:11,220 --> 00:42:13,240 I made up this slide based on what 672 00:42:13,240 --> 00:42:20,420 I heard Rudiger Wehner talk about here in a colloquium 673 00:42:20,420 --> 00:42:23,900 talk he gave in October of 2011. 674 00:42:23,900 --> 00:42:25,470 Just two years ago. 675 00:42:25,470 --> 00:42:30,000 And if you search under Wehner, you 676 00:42:30,000 --> 00:42:31,640 will find a number of these studies 677 00:42:31,640 --> 00:42:36,110 that he's done with students on how these ants do it. 678 00:42:36,110 --> 00:42:39,780 He studied-- this is called path integration. 679 00:42:39,780 --> 00:42:42,140 And also, studies of landmark orientation 680 00:42:42,140 --> 00:42:43,390 that we've been talking about. 681 00:42:43,390 --> 00:42:46,490 Now let's talk about path integration. 682 00:42:46,490 --> 00:42:48,420 It's like the humans use, that are just 683 00:42:48,420 --> 00:42:52,600 using vestibular cues and their strides 684 00:42:52,600 --> 00:42:56,885 to know about how far and in what direction they've walked. 685 00:42:59,600 --> 00:43:01,920 They chose the study of this desert ant 686 00:43:01,920 --> 00:43:06,670 because there's just no landmarks in the Sahara, 687 00:43:06,670 --> 00:43:10,760 in the area they were studying, that these ants could be using. 688 00:43:10,760 --> 00:43:11,300 Just sand. 689 00:43:14,410 --> 00:43:18,760 And he found that they will walk a little ways 690 00:43:18,760 --> 00:43:20,660 and they will stop. 691 00:43:20,660 --> 00:43:22,860 Look around. 692 00:43:22,860 --> 00:43:25,730 Then they will walk further. 693 00:43:25,730 --> 00:43:32,372 So they could be stopping here, here, here, here, and so forth. 694 00:43:32,372 --> 00:43:33,860 They walk and stop. 695 00:43:33,860 --> 00:43:35,770 Walk and stop. 696 00:43:35,770 --> 00:43:39,440 What they're doing is, they're checking direction. 697 00:43:39,440 --> 00:43:42,210 How do they know direction? 698 00:43:42,210 --> 00:43:44,730 Basically, from the position of the sun. 699 00:43:44,730 --> 00:43:47,600 And they don't actually have to see the sun itself; 700 00:43:47,600 --> 00:43:51,910 they can see the pattern of polarization in the sky. 701 00:43:51,910 --> 00:43:54,880 We can't see that, but some birds can see that, 702 00:43:54,880 --> 00:43:58,005 and ants can-- many insects can see it. 703 00:43:58,005 --> 00:44:03,596 They're sensitive to polarity of the light coming 704 00:44:03,596 --> 00:44:04,845 from various parts of the sky. 705 00:44:04,845 --> 00:44:06,570 And it has a particular pattern that 706 00:44:06,570 --> 00:44:10,780 varies with how high in the sky the light is. 707 00:44:10,780 --> 00:44:15,020 OK, but what if the ant takes a long time 708 00:44:15,020 --> 00:44:16,740 before he finds an insect? 709 00:44:16,740 --> 00:44:20,140 So time has passed? 710 00:44:20,140 --> 00:44:22,170 Well, you could say, that's why he 711 00:44:22,170 --> 00:44:26,380 has to stop so often-- to recalibrate here. 712 00:44:26,380 --> 00:44:33,670 But in fact, you can transport them, 713 00:44:33,670 --> 00:44:35,770 and they'll still keep the direction 714 00:44:35,770 --> 00:44:39,150 clear from looking at the sky. 715 00:44:39,150 --> 00:44:41,650 And even if time has passed. 716 00:44:41,650 --> 00:44:43,660 You can trap them here. 717 00:44:43,660 --> 00:44:47,040 So you trap the animal here, and you keep him there. 718 00:44:47,040 --> 00:44:48,970 Let the sun move quite a bit. 719 00:44:48,970 --> 00:44:51,890 Then you let him go. 720 00:44:51,890 --> 00:44:57,230 Or you trap him just after he finds a dead insect. 721 00:44:57,230 --> 00:44:59,230 But now you keep him from going back right away. 722 00:44:59,230 --> 00:45:03,190 He still finds his way back, because he's 723 00:45:03,190 --> 00:45:07,056 got a biological clock. 724 00:45:07,056 --> 00:45:10,260 His knowledge of direction from the sun 725 00:45:10,260 --> 00:45:12,860 is always re-calibrated according to time of day. 726 00:45:16,710 --> 00:45:18,980 So anyway, the animal finds the insect, 727 00:45:18,980 --> 00:45:23,450 and he heads directly back to his nest. 728 00:45:23,450 --> 00:45:26,500 And he can do that without any landmarks. 729 00:45:26,500 --> 00:45:30,530 Now, if there are some landmarks, 730 00:45:30,530 --> 00:45:32,230 they do the same thing. 731 00:45:32,230 --> 00:45:36,950 But now they keep correcting for any cumulative error that 732 00:45:36,950 --> 00:45:37,450 occurred. 733 00:45:37,450 --> 00:45:41,430 Because they're not going to be absolutely perfect. 734 00:45:41,430 --> 00:45:47,545 But how do they figure out-- just knowing the direction back 735 00:45:47,545 --> 00:45:49,840 to the nest every time isn't enough. 736 00:45:49,840 --> 00:45:53,790 You have to know how far you've traveled, too. 737 00:45:53,790 --> 00:45:56,480 How do they know that? 738 00:45:56,480 --> 00:45:57,920 So they did experiments. 739 00:45:57,920 --> 00:45:59,360 Some amazing experiments. 740 00:45:59,360 --> 00:46:01,920 They were really quite surprising. 741 00:46:01,920 --> 00:46:05,890 They know, because of how many steps they took 742 00:46:05,890 --> 00:46:11,120 and the length of their stride. 743 00:46:11,120 --> 00:46:13,000 How did they know that? 744 00:46:13,000 --> 00:46:18,940 They did experiments where they put little stilts on the ant. 745 00:46:18,940 --> 00:46:21,830 If you put the stilts on the ant after he finds food, 746 00:46:21,830 --> 00:46:25,380 and then you release him, he's actually 747 00:46:25,380 --> 00:46:29,570 moving a lot further with the same number of steps. 748 00:46:29,570 --> 00:46:30,860 And he will make errors. 749 00:46:30,860 --> 00:46:32,812 He'll go beyond his nest. 750 00:46:35,980 --> 00:46:41,610 How can they vary the number of steps, but not the stride 751 00:46:41,610 --> 00:46:42,230 length? 752 00:46:42,230 --> 00:46:46,640 Well, they put ramps right in his path, 753 00:46:46,640 --> 00:46:49,560 so the ant has to go an additional distance. 754 00:46:54,530 --> 00:46:58,980 He has to make many more steps to get the same distance, 755 00:46:58,980 --> 00:47:01,900 because of those ramps. 756 00:47:01,900 --> 00:47:06,740 Remember, these ants are-- if it were a human, you would notice. 757 00:47:06,740 --> 00:47:09,110 You might take some of this into account. 758 00:47:09,110 --> 00:47:15,961 But the ant is limited to his use of direction and step 759 00:47:15,961 --> 00:47:16,460 length. 760 00:47:18,990 --> 00:47:20,740 And number of steps. 761 00:47:20,740 --> 00:47:23,540 So he's constantly keeping track of that, 762 00:47:23,540 --> 00:47:27,450 so when he gets to his insect here, 763 00:47:27,450 --> 00:47:30,660 he gets back pretty accurately to that nest. 764 00:47:30,660 --> 00:47:32,390 There will be some cumulative errors 765 00:47:32,390 --> 00:47:35,400 if there's no landmarks at all he's using to recalibrate. 766 00:47:39,340 --> 00:47:41,280 So now let's go to homing pigeons. 767 00:47:41,280 --> 00:47:47,550 And this is dealt with more in Scott. 768 00:47:47,550 --> 00:47:52,370 And pigeons, we know-- they know a direction. 769 00:47:52,370 --> 00:47:54,270 They've got some kind of compass. 770 00:47:54,270 --> 00:47:57,300 And at least when they're in the region 771 00:47:57,300 --> 00:48:01,980 of their nest, their loft, they do have a map of the region. 772 00:48:01,980 --> 00:48:05,990 So they use both a compass and a map. 773 00:48:05,990 --> 00:48:08,500 And we know they use multiple cues. 774 00:48:08,500 --> 00:48:13,100 There's evidence now-- and it was very controversial. 775 00:48:13,100 --> 00:48:19,090 30, 40 years ago-- even now, you see some studies of this, 776 00:48:19,090 --> 00:48:20,750 but now there's much more agreement 777 00:48:20,750 --> 00:48:25,110 that they are capable of using the earth's magnetic field 778 00:48:25,110 --> 00:48:26,290 to know direction. 779 00:48:26,290 --> 00:48:29,600 But they use the sun if it's available. 780 00:48:29,600 --> 00:48:34,230 If there's no sun, or if they're traveling at night, 781 00:48:34,230 --> 00:48:37,890 they use the magnetic field. 782 00:48:37,890 --> 00:48:39,560 Or, in the case of songbirds, they 783 00:48:39,560 --> 00:48:42,820 can actually use the stars. 784 00:48:42,820 --> 00:48:44,970 When they're in the vicinity, they 785 00:48:44,970 --> 00:48:50,220 don't have to be perfect to do that. 786 00:48:50,220 --> 00:48:52,690 They just have to get to the vicinity of their loft, 787 00:48:52,690 --> 00:48:55,140 because then they have a map. 788 00:48:55,140 --> 00:48:58,410 Because they do spend time flying around their loft, 789 00:48:58,410 --> 00:49:01,150 and they do learn all the landmarks around there. 790 00:49:01,150 --> 00:49:04,660 They develop a map. 791 00:49:04,660 --> 00:49:07,890 And so, in the vicinity of their loft-- 792 00:49:07,890 --> 00:49:12,210 because pigeons can home from very long distances. 793 00:49:12,210 --> 00:49:14,610 So they might not get that close, 794 00:49:14,610 --> 00:49:19,000 just from their compass sense. 795 00:49:19,000 --> 00:49:21,500 But they use not just visual landmarks. 796 00:49:21,500 --> 00:49:24,500 There is evidence that they use olfactory cues. 797 00:49:24,500 --> 00:49:27,990 And there's even some evidence that patterns of ultrasound 798 00:49:27,990 --> 00:49:30,290 can make a difference, and be used by at least some 799 00:49:30,290 --> 00:49:30,960 of these birds. 800 00:49:38,170 --> 00:49:41,930 We have time to talk about the Morris water maze, 801 00:49:41,930 --> 00:49:46,100 used here in the lab for studies of mice and rats. 802 00:49:48,990 --> 00:49:53,080 What is the Morris water maze? 803 00:49:56,130 --> 00:50:00,070 You do this horrible thing to these little guys. 804 00:50:00,070 --> 00:50:02,220 You throw them in water. 805 00:50:02,220 --> 00:50:03,380 And the water's too deep. 806 00:50:03,380 --> 00:50:05,120 And they have to swim to stay alive. 807 00:50:09,920 --> 00:50:12,840 It's usually a circular maze filled with water. 808 00:50:12,840 --> 00:50:15,460 In one place, there's a little platform. 809 00:50:15,460 --> 00:50:16,990 He can't see it. 810 00:50:16,990 --> 00:50:20,740 But if he finds it-- if he swims to it-- he can climb there, 811 00:50:20,740 --> 00:50:24,870 and then he can stop swimming and rest. 812 00:50:24,870 --> 00:50:27,270 So if you put them in there repeatedly, 813 00:50:27,270 --> 00:50:29,470 they learn where that platform is. 814 00:50:29,470 --> 00:50:31,820 Well, how do they learn? 815 00:50:31,820 --> 00:50:34,740 They must be using specific cues. 816 00:50:34,740 --> 00:50:37,360 How do they learn the position of the platform 817 00:50:37,360 --> 00:50:38,525 that they can't see? 818 00:50:41,450 --> 00:50:43,270 You can do studies. 819 00:50:46,550 --> 00:50:48,492 I'm summarizing there, some of the studies 820 00:50:48,492 --> 00:50:49,325 that have been done. 821 00:50:51,910 --> 00:50:54,070 We, of course, have the maze in a room. 822 00:50:54,070 --> 00:50:58,680 You can rotate the maze around. 823 00:50:58,680 --> 00:51:02,690 So all the visual cues in the room stay the same, 824 00:51:02,690 --> 00:51:04,500 but you're rotating the maze. 825 00:51:04,500 --> 00:51:05,960 What happens? 826 00:51:05,960 --> 00:51:08,350 He goes to the wrong place, because he's 827 00:51:08,350 --> 00:51:10,940 using those visual cues. 828 00:51:10,940 --> 00:51:12,540 Another way that that's been done, 829 00:51:12,540 --> 00:51:17,070 just to make sure he's not using the magnetic field 830 00:51:17,070 --> 00:51:21,090 or whatever-- you rotate the room. 831 00:51:21,090 --> 00:51:23,720 Meaning, you have sort of a room inside a room. 832 00:51:23,720 --> 00:51:26,560 You can rotate that, and leave the maze constant. 833 00:51:29,370 --> 00:51:31,340 But the mistakes he makes always seem 834 00:51:31,340 --> 00:51:35,610 to depend on the visual cues. 835 00:51:35,610 --> 00:51:38,950 The overhead cues in particular. 836 00:51:38,950 --> 00:51:42,180 Our position with respect to that platform. 837 00:51:46,440 --> 00:51:49,330 So that's how they find it. 838 00:51:49,330 --> 00:51:51,940 And we know that the hippocampal formation, 839 00:51:51,940 --> 00:51:56,610 so important in forming a cognitive map 840 00:51:56,610 --> 00:52:02,090 of the environment, is necessary for the pigeon-- not 841 00:52:02,090 --> 00:52:03,970 for his compass direction. 842 00:52:03,970 --> 00:52:05,770 They'll still go in the right direction, 843 00:52:05,770 --> 00:52:07,400 even without a hippocampus. 844 00:52:07,400 --> 00:52:10,190 But they won't have that map. 845 00:52:10,190 --> 00:52:13,220 That will be messed up by hippocampal damage. 846 00:52:13,220 --> 00:52:15,470 And similarly in the Morris water maze, 847 00:52:15,470 --> 00:52:21,540 they can't use to cues to form a map of that maze 848 00:52:21,540 --> 00:52:23,065 without the hippocampus. 849 00:52:31,280 --> 00:52:34,040 We'll come back here to just finish the last few slides 850 00:52:34,040 --> 00:52:35,590 next time.