1
00:00:00,120 --> 00:00:01,800
The following
content is provided

2
00:00:01,800 --> 00:00:04,040
under a Creative
Commons license.

3
00:00:04,040 --> 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 --> 00:00:13,500
To make a donation or view
additional materials from

6
00:00:13,500 --> 00:00:17,270
hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:17,270 --> 00:00:21,318
at ocw.mit.edu .

8
00:00:21,318 --> 00:00:22,550
PROFESSOR: OK.

9
00:00:22,550 --> 00:00:28,840
So we've started talking
about many of the key concepts

10
00:00:28,840 --> 00:00:32,930
in behavioral
studies of learning.

11
00:00:32,930 --> 00:00:36,750
And I'm basing this
on Konrad Lorenz.

12
00:00:36,750 --> 00:00:47,360
And this is where we ended.

13
00:00:47,360 --> 00:00:50,100
We were about to talk about
avoidance responses acquired

14
00:00:50,100 --> 00:00:52,120
through trauma.

15
00:00:52,120 --> 00:00:56,200
We had just talked about
conditioned reflexes.

16
00:00:56,200 --> 00:01:00,180
Remember, he sees the
conditioned reflex,

17
00:01:00,180 --> 00:01:04,860
as the way it has been studied
in experimental situations,

18
00:01:04,860 --> 00:01:06,580
as a kind of stimulus selection.

19
00:01:09,100 --> 00:01:10,990
And that's always the
way it's interpreted.

20
00:01:10,990 --> 00:01:13,410
And there are many,
many experiments

21
00:01:13,410 --> 00:01:19,480
that perfectly fit the
Lorenz definition, even

22
00:01:19,480 --> 00:01:22,210
though Pavlov's
original experiment

23
00:01:22,210 --> 00:01:28,650
that so popularly
described as an example

24
00:01:28,650 --> 00:01:31,360
of classical conditioning.

25
00:01:31,360 --> 00:01:33,860
So we'll bring that up again
a little later because you'll

26
00:01:33,860 --> 00:01:36,780
see, then, what Lorenz calls it.

27
00:01:36,780 --> 00:01:39,200
But first let's finish
talking about learning

28
00:01:39,200 --> 00:01:43,480
through association without
feedback, without feedback

29
00:01:43,480 --> 00:01:48,230
reporting the success
of the movement.

30
00:01:48,230 --> 00:01:52,380
So we'll talk about
avoidance response

31
00:01:52,380 --> 00:01:53,750
acquired through trauma.

32
00:01:53,750 --> 00:01:55,950
It's very interesting
kind of learning,

33
00:01:55,950 --> 00:01:59,990
because it's pretty
irreversible.

34
00:01:59,990 --> 00:02:03,750
It's another type of
stimulus selection.

35
00:02:03,750 --> 00:02:06,590
It can occur after
just one experience,

36
00:02:06,590 --> 00:02:10,740
and then be
considerably permanent.

37
00:02:10,740 --> 00:02:15,750
He talks about how trainers
of dogs and horses--

38
00:02:15,750 --> 00:02:24,140
just like psychoanalysts-- they
know how, once an animal learns

39
00:02:24,140 --> 00:02:27,720
to avoid something, some
stimulus in his environment

40
00:02:27,720 --> 00:02:33,750
becomes associated with
avoidance, something that's

41
00:02:33,750 --> 00:02:37,540
frightened the animal a lot,
or frightened the person,

42
00:02:37,540 --> 00:02:40,170
becomes associated
with certain stimuli,

43
00:02:40,170 --> 00:02:42,345
it's very difficult
to change that.

44
00:02:45,550 --> 00:02:47,730
In the laboratory
usually it's done

45
00:02:47,730 --> 00:02:51,830
by pairing sound
with electric shock.

46
00:02:51,830 --> 00:02:56,430
Sometimes it's not sound,
it's another stimulus.

47
00:02:56,430 --> 00:03:00,280
And then later, with later
presentations of the sound,

48
00:03:00,280 --> 00:03:04,785
you will get autonomic
responses, increased heart rate

49
00:03:04,785 --> 00:03:06,500
and blood pressure.

50
00:03:06,500 --> 00:03:10,280
And at least some
forms of it are just

51
00:03:10,280 --> 00:03:12,950
one trial learning,
one experience.

52
00:03:12,950 --> 00:03:15,870
The most common type in the
laboratory, as many of you

53
00:03:15,870 --> 00:03:18,250
know, is what we call
the step down response,

54
00:03:18,250 --> 00:03:21,560
because a little rodent
on a raised platform

55
00:03:21,560 --> 00:03:26,920
will tend to step down from
a platform, unless his vision

56
00:03:26,920 --> 00:03:30,190
indicates that
he's very high up,

57
00:03:30,190 --> 00:03:33,220
then he's much less
likely to step down.

58
00:03:33,220 --> 00:03:37,210
But if he can step down fairly
comfortably he will do it.

59
00:03:37,210 --> 00:03:40,500
But when he steps
down, the grid--

60
00:03:40,500 --> 00:03:44,060
he steps on an
electrified grid-- then,

61
00:03:44,060 --> 00:03:46,670
just in that one trial,
you bring him back later,

62
00:03:46,670 --> 00:03:51,700
and he'll just stay and stay
and stay on the platform.

63
00:03:51,700 --> 00:03:55,790
So that's an example of
avoidance responses acquired

64
00:03:55,790 --> 00:03:56,415
through trauma.

65
00:03:58,940 --> 00:04:05,920
And I just point out here that
most psychologists-- anything

66
00:04:05,920 --> 00:04:07,730
I've added in red
on these slides

67
00:04:07,730 --> 00:04:10,990
is not in the thing
posted online,

68
00:04:10,990 --> 00:04:15,030
but I'll change that to the
annotated version later on.

69
00:04:15,030 --> 00:04:17,620
So most psychologists
include this kind

70
00:04:17,620 --> 00:04:19,940
of learning in the classical
conditioning category.

71
00:04:19,940 --> 00:04:24,370
But it's got enough differences,
as Lorenz points out,

72
00:04:24,370 --> 00:04:25,250
to separate it.

73
00:04:28,710 --> 00:04:32,380
In other words, animals appear
to have evolved differences.

74
00:04:32,380 --> 00:04:35,770
That's why this type is
so irreversible compared

75
00:04:35,770 --> 00:04:40,984
with other types of
stimulus selection learning.

76
00:04:40,984 --> 00:04:43,275
And then imprinting, which
we have talked about before.

77
00:04:47,310 --> 00:04:50,170
I said here-- this is
on the posted slides,

78
00:04:50,170 --> 00:04:54,320
both of the previous two--
that is, both avoidance

79
00:04:54,320 --> 00:04:57,050
response acquired through
trauma and conditioned reflexes

80
00:04:57,050 --> 00:04:59,320
are irreversible.

81
00:04:59,320 --> 00:05:02,090
It's certainly true of the
avoidance responses acquired

82
00:05:02,090 --> 00:05:02,780
through trauma.

83
00:05:02,780 --> 00:05:04,560
Relatively speaking,
it's irreversible.

84
00:05:07,490 --> 00:05:09,470
Other types of
classical conditioning

85
00:05:09,470 --> 00:05:11,010
are perhaps less irreversible.

86
00:05:14,600 --> 00:05:17,690
But we know that
imprinting will cause

87
00:05:17,690 --> 00:05:21,390
a fixation of
specific responses.

88
00:05:21,390 --> 00:05:22,680
They're very specific.

89
00:05:22,680 --> 00:05:27,610
We've learned that birds don't
become imprinted on humans.

90
00:05:27,610 --> 00:05:28,700
There's no such thing.

91
00:05:28,700 --> 00:05:32,400
They become imprinted for
particular situations,

92
00:05:32,400 --> 00:05:38,610
particular kind of
responses, like his jackdaws.

93
00:05:38,610 --> 00:05:40,940
Because of early
imprinting, they

94
00:05:40,940 --> 00:05:47,060
can become fixated on
companions for sex,

95
00:05:47,060 --> 00:05:50,590
or they can become separately
imprinted on companions

96
00:05:50,590 --> 00:05:58,440
for flocking and flying, and
separately for parental care.

97
00:05:58,440 --> 00:06:01,290
And these things
are quite separable.

98
00:06:01,290 --> 00:06:04,260
And all that became clear
in his studies of jackdaws.

99
00:06:04,260 --> 00:06:07,590
And it's also clear in geese.

100
00:06:10,660 --> 00:06:13,270
Geese that are
raised with humans

101
00:06:13,270 --> 00:06:17,490
can become imprinted
for what he calls

102
00:06:17,490 --> 00:06:19,760
filial and social responses.

103
00:06:19,760 --> 00:06:23,310
And yet in their sexual
behavior they're not affected.

104
00:06:23,310 --> 00:06:26,720
They will have sexual
behavior normally

105
00:06:26,720 --> 00:06:29,030
with other members
of their species.

106
00:06:29,030 --> 00:06:32,097
And so there are real
species differences in this.

107
00:06:36,530 --> 00:06:42,240
You can separate imprinting
from conditioning with reward

108
00:06:42,240 --> 00:06:44,940
by simply pitting them
against each other.

109
00:06:44,940 --> 00:06:48,910
For example, you
can have animals

110
00:06:48,910 --> 00:06:57,090
that are imprinted on, say,
humans as sexual partners.

111
00:06:57,090 --> 00:07:01,570
And we've had examples of that,
especially in the jackdaws.

112
00:07:04,190 --> 00:07:06,390
But it's been done with a
number of other species.

113
00:07:06,390 --> 00:07:10,390
And in his book, he
describes specific species

114
00:07:10,390 --> 00:07:16,015
where a species was imprinted
for mating on another species,

115
00:07:16,015 --> 00:07:19,060
a similar species but different.

116
00:07:19,060 --> 00:07:22,110
Well, if you deprive the animal
for a long period of time,

117
00:07:22,110 --> 00:07:24,880
you never expose him
anymore to that species

118
00:07:24,880 --> 00:07:29,520
he's imprinted on,
he eventually will--

119
00:07:29,520 --> 00:07:32,340
because his action-specific--
what happens?

120
00:07:32,340 --> 00:07:35,020
You deprive an animal
for a very long time,

121
00:07:35,020 --> 00:07:36,640
the action-specific
potential, that

122
00:07:36,640 --> 00:07:38,800
is, the internal
drive becomes so high,

123
00:07:38,800 --> 00:07:42,710
thresholds are lowered,
he's likely to start mating

124
00:07:42,710 --> 00:07:43,980
with its own species.

125
00:07:43,980 --> 00:07:44,480
OK.

126
00:07:44,480 --> 00:07:46,650
So you let him do
that for a while,

127
00:07:46,650 --> 00:07:49,710
taking care of
young, everything.

128
00:07:49,710 --> 00:07:54,710
And now you bring
back the other species

129
00:07:54,710 --> 00:07:56,810
that he was imprinted on.

130
00:07:56,810 --> 00:07:58,390
What happens?

131
00:07:58,390 --> 00:08:03,670
Either immediately or
gradually, he switches over,

132
00:08:03,670 --> 00:08:05,170
he switches back.

133
00:08:05,170 --> 00:08:07,630
With all the reward and
everything associated

134
00:08:07,630 --> 00:08:11,540
with mating with
his own species,

135
00:08:11,540 --> 00:08:17,710
he still will switch back to
the imprint of species, which

136
00:08:17,710 --> 00:08:21,840
shows the independence
of these two processes.

137
00:08:21,840 --> 00:08:25,090
Then he gives the example of
innate releasing mechanisms

138
00:08:25,090 --> 00:08:29,150
in mallard males and females.

139
00:08:29,150 --> 00:08:31,255
The males can be
sexually imprinted

140
00:08:31,255 --> 00:08:35,330
to other species,
but not the females.

141
00:08:35,330 --> 00:08:37,860
The reason is
they're so dominated

142
00:08:37,860 --> 00:08:42,630
by innate releasing
mechanisms that

143
00:08:42,630 --> 00:08:46,740
respond to the mallard
drake's colors,

144
00:08:46,740 --> 00:08:51,310
his mating colors,
nuptial colors.

145
00:08:51,310 --> 00:08:56,240
So it's a very different
situation for male and female.

146
00:08:56,240 --> 00:08:59,710
Just like there are differences
from one species to the other,

147
00:08:59,710 --> 00:09:03,990
there can be differences
between the two sexes.

148
00:09:03,990 --> 00:09:07,250
So and then finally what he
calls conditioned inhibition.

149
00:09:09,760 --> 00:09:16,000
And we train domestic
animals and circus animals.

150
00:09:16,000 --> 00:09:23,640
We even train tigers and
bears for circus performance.

151
00:09:23,640 --> 00:09:30,870
And they learn, they are taught,
to inhibit natural responses.

152
00:09:30,870 --> 00:09:33,710
Obviously you're going
to be in a circus,

153
00:09:33,710 --> 00:09:37,990
you don't want them to be
showing prey-catching responses

154
00:09:37,990 --> 00:09:41,290
and so forth to a
little child running

155
00:09:41,290 --> 00:09:43,330
in front of the audience.

156
00:09:43,330 --> 00:09:49,400
So you can suppress,
you can condition them

157
00:09:49,400 --> 00:09:52,990
so they're inhibiting
their normal responses.

158
00:09:52,990 --> 00:09:56,490
The problem is, if the
action-specific potentials

159
00:09:56,490 --> 00:10:02,380
build up and build up and build
up because of long suppression,

160
00:10:02,380 --> 00:10:06,410
you can sometimes have very
dangerous consequences.

161
00:10:06,410 --> 00:10:11,720
And his examples are pack
animals, like wolves.

162
00:10:11,720 --> 00:10:17,690
A dominant pack member can
totally suppress resistance

163
00:10:17,690 --> 00:10:19,520
by other members of the pack.

164
00:10:19,520 --> 00:10:21,810
And so you'll have passive
members of the pack.

165
00:10:21,810 --> 00:10:26,130
But they're really-- if he's
at the bottom of the totem

166
00:10:26,130 --> 00:10:28,980
pole he's really repressed.

167
00:10:28,980 --> 00:10:36,890
He can suddenly-- his
desire to respond more

168
00:10:36,890 --> 00:10:41,590
normally and oppose that
leader can suddenly break out,

169
00:10:41,590 --> 00:10:43,380
and basically all
hell breaks loose

170
00:10:43,380 --> 00:10:46,729
and the animal becomes
very dangerous,

171
00:10:46,729 --> 00:10:48,770
because the normal
inhibitions can suddenly fail.

172
00:10:48,770 --> 00:10:52,910
And there are specific
examples Lorenz goes through.

173
00:10:52,910 --> 00:10:56,710
Hugo Van Lawick studied this
in African hunting dogs.

174
00:10:56,710 --> 00:10:58,770
And he showed examples of it.

175
00:10:58,770 --> 00:11:01,050
They're another
group-living animal.

176
00:11:01,050 --> 00:11:05,320
There's a dominant
animal, and the animals

177
00:11:05,320 --> 00:11:09,230
low down on the social
hierarchy can occasionally

178
00:11:09,230 --> 00:11:13,010
do this if their normal behavior
is suppressed for too long.

179
00:11:13,010 --> 00:11:14,900
We mentioned they're
already in wolves.

180
00:11:14,900 --> 00:11:17,470
It happens in humans,
too, in a situation

181
00:11:17,470 --> 00:11:23,840
that certainly has legal
consequences, namely

182
00:11:23,840 --> 00:11:25,370
spousal murder.

183
00:11:25,370 --> 00:11:29,940
So it usually happens when a
woman is so totally controlled

184
00:11:29,940 --> 00:11:35,580
by a spouse that her normal
reactions are totally

185
00:11:35,580 --> 00:11:36,280
suppressed.

186
00:11:36,280 --> 00:11:40,680
Her normal drives
she cannot express.

187
00:11:40,680 --> 00:11:44,150
And if this goes
on and on and on,

188
00:11:44,150 --> 00:11:47,440
especially if she's
cruelly treated,

189
00:11:47,440 --> 00:11:51,010
she eventually-- her
inhibitions can suddenly fail,

190
00:11:51,010 --> 00:11:54,600
and she just kills him.

191
00:11:54,600 --> 00:11:59,670
And when that comes to the
court, women in that situation

192
00:11:59,670 --> 00:12:03,240
are often acquitted.

193
00:12:03,240 --> 00:12:05,790
So this does have
some recognition

194
00:12:05,790 --> 00:12:12,480
in our legal system, this
consequences of basically

195
00:12:12,480 --> 00:12:16,110
conditioned inhibition,
although in the legal sense

196
00:12:16,110 --> 00:12:17,070
it's not called that.

197
00:12:19,740 --> 00:12:24,430
So now, learning affected by
the consequences of behavior.

198
00:12:24,430 --> 00:12:27,110
These are the common kinds.

199
00:12:27,110 --> 00:12:30,040
Conditioned appetitive
behavior we'll deal with first.

200
00:12:30,040 --> 00:12:33,500
Conditioned aversion,
conditioned action.

201
00:12:33,500 --> 00:12:35,310
Conditioned aversion
we've talked about.

202
00:12:35,310 --> 00:12:39,390
Conditioned taste aversion
is the usual phenomenon

203
00:12:39,390 --> 00:12:40,060
we talk about.

204
00:12:40,060 --> 00:12:43,300
Then conditioned action,
conditioned appetitive behavior

205
00:12:43,300 --> 00:12:44,890
directed at quiescence,
when there's

206
00:12:44,890 --> 00:12:52,700
some very strong annoyer
like hunger or thirst,

207
00:12:52,700 --> 00:12:57,400
or anything that's
extremely bothersome

208
00:12:57,400 --> 00:12:59,960
and difficult and uncomfortable.

209
00:12:59,960 --> 00:13:01,850
And finally operant
conditioning,

210
00:13:01,850 --> 00:13:04,560
which is not stimulus
conditioning at all

211
00:13:04,560 --> 00:13:08,940
but response selection, which
Lorenz claims is much less

212
00:13:08,940 --> 00:13:15,600
common than people
tend to be taught

213
00:13:15,600 --> 00:13:19,500
that learn about what they
call instrumental conditioning.

214
00:13:19,500 --> 00:13:21,960
It's usually operant
conditioning,

215
00:13:21,960 --> 00:13:25,430
which involves
response selection.

216
00:13:25,430 --> 00:13:27,770
This is just an
introduction to that.

217
00:13:27,770 --> 00:13:28,880
Lorenz.

218
00:13:28,880 --> 00:13:33,890
We've been talking mainly about
Type S or stimulus selection

219
00:13:33,890 --> 00:13:36,830
types of learning.

220
00:13:36,830 --> 00:13:40,200
He says much exploratory
behavior is that way,

221
00:13:40,200 --> 00:13:42,320
and it's the most common.

222
00:13:42,320 --> 00:13:44,700
Type R conditioning,
response selection,

223
00:13:44,700 --> 00:13:48,170
is a little more rare.

224
00:13:48,170 --> 00:13:50,620
It includes operant
conditioning,

225
00:13:50,620 --> 00:13:52,700
like with a confined
animal that's

226
00:13:52,700 --> 00:13:55,195
trying every response he
can in order to escape.

227
00:13:59,300 --> 00:14:02,540
It involves reinforcement,
some kind of reward.

228
00:14:02,540 --> 00:14:05,136
He likes the word encouraging
rather than reinforcement.

229
00:14:08,050 --> 00:14:10,540
And then he talks
about, applies this

230
00:14:10,540 --> 00:14:12,740
to the study of fixed
action patterns, where

231
00:14:12,740 --> 00:14:16,640
you have-- it starts
with the drive,

232
00:14:16,640 --> 00:14:19,540
the action-specific potential
which leads to appetitive

233
00:14:19,540 --> 00:14:25,020
behavior, which basically he's
looking for the stimulus that

234
00:14:25,020 --> 00:14:31,190
can release the response of
that response side of the fixed

235
00:14:31,190 --> 00:14:33,110
action pattern.

236
00:14:33,110 --> 00:14:35,740
So then his innate
releasing mechanism

237
00:14:35,740 --> 00:14:40,215
is activated, which causes
the consummatory behavior,

238
00:14:40,215 --> 00:14:42,410
the consummatory response.

239
00:14:42,410 --> 00:14:49,520
And then the feedback from
that behavior, a the feedback

240
00:14:49,520 --> 00:14:52,560
affects the
action-specific potential.

241
00:14:52,560 --> 00:14:53,430
It lowers it.

242
00:14:58,860 --> 00:15:02,740
It usually stops the
consummatory behavior,

243
00:15:02,740 --> 00:15:03,735
reduces the drive.

244
00:15:07,880 --> 00:15:11,620
And so it reduces
appetitive behavior.

245
00:15:11,620 --> 00:15:16,630
And it can actually shape the
original releasing mechanism,

246
00:15:16,630 --> 00:15:21,350
because there's-- learning in
these innate behavior sequences

247
00:15:21,350 --> 00:15:24,200
is mostly on the stimulus side.

248
00:15:24,200 --> 00:15:31,150
Like an animal that
has built-in responses.

249
00:15:31,150 --> 00:15:34,600
We talked about the
smiling response.

250
00:15:34,600 --> 00:15:36,570
It's a fixed action pattern.

251
00:15:36,570 --> 00:15:38,555
They will smile
spontaneously sometimes,

252
00:15:38,555 --> 00:15:41,880
but they normally respond
in almost reflex fashion

253
00:15:41,880 --> 00:15:44,650
to this specific configuration.

254
00:15:44,650 --> 00:15:48,390
And yet over time it acquires
other characteristics,

255
00:15:48,390 --> 00:15:50,420
and his smiling
becomes more limited,

256
00:15:50,420 --> 00:15:53,410
restricted to
certain situations.

257
00:15:53,410 --> 00:15:54,570
OK

258
00:15:54,570 --> 00:15:56,440
And he has this
interesting statement

259
00:15:56,440 --> 00:15:59,300
about why behaviorists
have oversimplified

260
00:15:59,300 --> 00:16:01,820
the problem of
learning and instinct.

261
00:16:01,820 --> 00:16:05,480
And he's talking about American
comparative psychology.

262
00:16:05,480 --> 00:16:09,040
He says, they hope to be able
to abstract generally valid laws

263
00:16:09,040 --> 00:16:12,780
prevailing in all learning
processes, if not all behavior.

264
00:16:12,780 --> 00:16:16,420
And there have been many books
written like this, almost all

265
00:16:16,420 --> 00:16:17,935
over here in America.

266
00:16:22,280 --> 00:16:24,650
In this way they hope
to find a shortcut

267
00:16:24,650 --> 00:16:27,110
to an understanding of
animal and human behavior

268
00:16:27,110 --> 00:16:30,740
without going to the trouble
of analyzing the immensely

269
00:16:30,740 --> 00:16:33,090
complicated
physiological machinery

270
00:16:33,090 --> 00:16:35,870
whose function is behavior.

271
00:16:35,870 --> 00:16:37,440
And I couldn't
agree with him more.

272
00:16:41,370 --> 00:16:43,820
So these are the different
kinds of learning affected

273
00:16:43,820 --> 00:16:47,260
by the consequences of
behavior, the major categories.

274
00:16:47,260 --> 00:16:50,100
Start with conditioned
appetitive behavior.

275
00:16:50,100 --> 00:16:54,720
And he says that that and
conditioned appetitive behavior

276
00:16:54,720 --> 00:16:58,440
directed at quiescence
are the most common types.

277
00:16:58,440 --> 00:17:02,020
And I also point out that
usually operant conditioning

278
00:17:02,020 --> 00:17:04,079
is lumped together with
conditioned appetitive

279
00:17:04,079 --> 00:17:06,050
behavior, even
though they really

280
00:17:06,050 --> 00:17:09,250
are quite separable,
as Lorenz points out.

281
00:17:11,880 --> 00:17:15,180
And as you know, sometimes--
like the Russians

282
00:17:15,180 --> 00:17:19,089
interpreted all behavior in
terms of conditioned reflexes

283
00:17:19,089 --> 00:17:20,400
for many, many, many years.

284
00:17:23,680 --> 00:17:26,585
I don't know, people are
natural lumpers, maybe.

285
00:17:26,585 --> 00:17:28,990
But Lorenz is more
of a splitter here.

286
00:17:28,990 --> 00:17:30,980
He likes to divide
it up because--

287
00:17:30,980 --> 00:17:34,400
and I think this is much more
relevant to any brain studies,

288
00:17:34,400 --> 00:17:37,190
because if you
can't separate them,

289
00:17:37,190 --> 00:17:39,890
you won't be looking for
different brain mechanisms.

290
00:17:39,890 --> 00:17:42,980
And in fact there are many
distinct brain mechanisms

291
00:17:42,980 --> 00:17:47,480
going on, which we'll talk
about at the end here.

292
00:17:47,480 --> 00:17:49,810
So these are examples of
conditioned appetitive

293
00:17:49,810 --> 00:17:50,830
behavior.

294
00:17:50,830 --> 00:17:55,560
Von Frisch, Karl Von
Frisch, had kept a pet fish.

295
00:17:55,560 --> 00:18:00,880
And he always whistled
for it before he fed it.

296
00:18:00,880 --> 00:18:05,280
So the fish learned to
begin searching for food

297
00:18:05,280 --> 00:18:06,760
as soon as he
heard that whistle.

298
00:18:09,760 --> 00:18:13,650
The honey bees
studied by Von Frisch.

299
00:18:13,650 --> 00:18:16,820
The insect searches for
the stimulus situation

300
00:18:16,820 --> 00:18:18,690
which proved
rewarding in the past.

301
00:18:18,690 --> 00:18:20,830
For example, you
could teach it to fly

302
00:18:20,830 --> 00:18:24,190
to specifically colored patches.

303
00:18:24,190 --> 00:18:29,220
It's a type of
learning that he's

304
00:18:29,220 --> 00:18:33,840
rewarded if he goes
to those colors.

305
00:18:33,840 --> 00:18:37,350
Now the study of Pavlov's dogs.

306
00:18:37,350 --> 00:18:42,255
There's an interesting story
about that from Howard Liddell.

307
00:18:48,580 --> 00:18:51,777
I can actually read
what he said about that,

308
00:18:51,777 --> 00:18:53,110
because it's pretty interesting.

309
00:18:53,110 --> 00:18:53,985
He's quoting Liddell.

310
00:18:58,570 --> 00:19:03,172
In most of Pavlov's experiments,
active appetitive behavior

311
00:19:03,172 --> 00:19:06,390
is made invisible
by shackling the dog

312
00:19:06,390 --> 00:19:10,200
to a framework, the
experimental apparatus,

313
00:19:10,200 --> 00:19:14,840
so that salivation is just about
the only response which is not

314
00:19:14,840 --> 00:19:15,590
precluded.

315
00:19:15,590 --> 00:19:17,070
He can't do anything else.

316
00:19:17,070 --> 00:19:20,360
He's locked in the apparatus.

317
00:19:20,360 --> 00:19:24,500
And then Howard Liddell told him
how he once conditioned the dog

318
00:19:24,500 --> 00:19:28,130
to salivate, using the
conventional Pavlovian

319
00:19:28,130 --> 00:19:34,370
method, whenever a constantly
ticking metronome was

320
00:19:34,370 --> 00:19:37,180
made to accelerate its beat.

321
00:19:37,180 --> 00:19:39,660
So he trained it, and
he got the salivation

322
00:19:39,660 --> 00:19:42,190
in the typical Pavlovian sense.

323
00:19:45,910 --> 00:19:49,780
Then he untied the
dog, and as soon

324
00:19:49,780 --> 00:19:52,480
as the dog was released
from the apparatus,

325
00:19:52,480 --> 00:19:55,920
it ran up to the
metronome at once,

326
00:19:55,920 --> 00:19:58,480
whined, wagged its
tail violently,

327
00:19:58,480 --> 00:20:00,750
and pushed against the
metronome with its nose,

328
00:20:00,750 --> 00:20:03,400
salivating intensely
all the while,

329
00:20:03,400 --> 00:20:07,140
even though the metronome
had not changed its rhythm.

330
00:20:07,140 --> 00:20:09,013
What had really
been conditioned is

331
00:20:09,013 --> 00:20:12,010
that the dog was-- and
the dog wasn't a reflex.

332
00:20:12,010 --> 00:20:15,060
There's no conditioned
reflex here.

333
00:20:15,060 --> 00:20:18,470
But rather it was a
complex and specific system

334
00:20:18,470 --> 00:20:22,729
of appetitive behavior by
which a dog begs for food.

335
00:20:22,729 --> 00:20:23,895
That's what was conditioned.

336
00:20:27,610 --> 00:20:32,810
It's an interesting
contrast, because what

337
00:20:32,810 --> 00:20:36,410
we call classical conditioning
now in the laboratory,

338
00:20:36,410 --> 00:20:41,470
like conditioned eye blink in
response to specific stimuli,

339
00:20:41,470 --> 00:20:43,360
is really very
different from this.

340
00:20:46,100 --> 00:20:48,215
Anyway, this is conditioned
appetitive behavior.

341
00:20:51,590 --> 00:20:54,680
He talks then about nest
building in social weaver

342
00:20:54,680 --> 00:20:57,110
birds, or corvine birds.

343
00:20:57,110 --> 00:20:59,900
like crows and ravens.

344
00:20:59,900 --> 00:21:02,130
The weavers have an
innate preference

345
00:21:02,130 --> 00:21:04,180
for a particular kind of grass.

346
00:21:04,180 --> 00:21:06,310
So if they get
that grass, that's

347
00:21:06,310 --> 00:21:07,860
what they bring to the nest.

348
00:21:07,860 --> 00:21:10,540
But the corvines, like the
jackdaws and the ravens,

349
00:21:10,540 --> 00:21:13,790
they use whatever nest
material gives good feedback

350
00:21:13,790 --> 00:21:16,850
during nest-building actions.

351
00:21:16,850 --> 00:21:21,420
So they get twigs of the right
shape and size, little twigs.

352
00:21:21,420 --> 00:21:24,700
He says it causes an
orgastic climax when they're

353
00:21:24,700 --> 00:21:27,760
stuck in the nest because it
gives them the right feedback.

354
00:21:27,760 --> 00:21:34,060
It gives them a real
joy, and it rewards them.

355
00:21:34,060 --> 00:21:38,830
In fact, they can get
supernormal feedback

356
00:21:38,830 --> 00:21:41,670
if they use soft metal
wire because it gives

357
00:21:41,670 --> 00:21:43,885
such strong feedback,
it works so well

358
00:21:43,885 --> 00:21:46,250
on that particular response.

359
00:21:46,250 --> 00:21:50,210
They become addicted
to it, just like what

360
00:21:50,210 --> 00:21:53,800
happens in human addictions,
even though the wire doesn't

361
00:21:53,800 --> 00:21:56,380
make a very good nest.

362
00:21:56,380 --> 00:21:59,890
But it gives the right feedback.

363
00:21:59,890 --> 00:22:02,800
Now when they
evolved, of course,

364
00:22:02,800 --> 00:22:04,525
there wasn't wire mesh around.

365
00:22:08,010 --> 00:22:09,590
So there wasn't
anything like that

366
00:22:09,590 --> 00:22:11,600
that could give that
super normal feedback

367
00:22:11,600 --> 00:22:15,190
and result in
maladaptive behavior.

368
00:22:15,190 --> 00:22:19,520
But because of
humans, now they can

369
00:22:19,520 --> 00:22:26,890
acquire this kind of abnormal
addiction to the wire.

370
00:22:26,890 --> 00:22:27,400
All right.

371
00:22:27,400 --> 00:22:32,100
So that's conditioned
appetitive behavior.

372
00:22:32,100 --> 00:22:36,610
Now a little bit about
conditioned aversion,

373
00:22:36,610 --> 00:22:38,530
the second kind of
learning affected

374
00:22:38,530 --> 00:22:41,900
by the consequences of behavior.

375
00:22:41,900 --> 00:22:45,490
This is John Garcia's discovery
that we've talked about,

376
00:22:45,490 --> 00:22:47,420
the poison bait effect.

377
00:22:47,420 --> 00:22:53,070
I just want to point out that
the sickness, the response

378
00:22:53,070 --> 00:23:01,040
here, the nausea, the
illness, that becomes

379
00:23:01,040 --> 00:23:04,650
associated with a stimulus,
not the most recently

380
00:23:04,650 --> 00:23:08,050
encountered stimulus.

381
00:23:08,050 --> 00:23:12,550
It associates the most
novel previous stimulus.

382
00:23:12,550 --> 00:23:16,780
Now I'd like to know,
are there other cases

383
00:23:16,780 --> 00:23:19,280
of conditioned
aversion other than

384
00:23:19,280 --> 00:23:22,810
ingestion-related behaviors?

385
00:23:22,810 --> 00:23:24,840
It's likely that
there are, but this

386
00:23:24,840 --> 00:23:29,000
is the one that's
been most studied.

387
00:23:29,000 --> 00:23:30,790
Lorenz believes that
conditioned aversion

388
00:23:30,790 --> 00:23:33,990
is more common in
many animals in nature

389
00:23:33,990 --> 00:23:36,870
than is something like
operant conditioning, which

390
00:23:36,870 --> 00:23:38,290
we will talk about in a minute.

391
00:23:41,890 --> 00:23:45,500
And I just remind you that
Garcia's discovery of this

392
00:23:45,500 --> 00:23:49,190
was initially suppressed by the
editors of American behavior

393
00:23:49,190 --> 00:23:50,260
magazines.

394
00:23:50,260 --> 00:23:56,960
It didn't fit the beliefs about
conditioning and learning.

395
00:24:00,100 --> 00:24:02,920
They're not supposed to become
conditioned to something that

396
00:24:02,920 --> 00:24:05,990
occurred a lot earlier, but
only the most recent stimuli.

397
00:24:08,930 --> 00:24:13,130
But actually they become
conditioned to the more novel

398
00:24:13,130 --> 00:24:21,620
stimulus, even though it might
have occurred much earlier.

399
00:24:21,620 --> 00:24:25,620
And then conditioned action.

400
00:24:25,620 --> 00:24:28,970
This is usually an artificial
kind of motor learning,

401
00:24:28,970 --> 00:24:32,830
like you shape actions
by circus animals.

402
00:24:32,830 --> 00:24:37,090
You-- I said initially
elicit a fixed action

403
00:24:37,090 --> 00:24:39,110
pattern in response
to a command.

404
00:24:39,110 --> 00:24:44,230
You're actually eliciting a
fixed motor pattern in response

405
00:24:44,230 --> 00:24:45,000
to a command.

406
00:24:45,000 --> 00:24:50,290
So the horse's capriole,
that is the leap and kicking

407
00:24:50,290 --> 00:24:56,130
that a horse does to try to
get rid of a predator that's

408
00:24:56,130 --> 00:25:00,570
attacking it-- they will
learn to perform that just

409
00:25:00,570 --> 00:25:01,370
to get sugar.

410
00:25:01,370 --> 00:25:03,105
So that's a conditioned action.

411
00:25:05,980 --> 00:25:08,360
Von Frisch
conditioned a parakeet

412
00:25:08,360 --> 00:25:11,620
to defecate in order to be
released from his stage.

413
00:25:11,620 --> 00:25:14,590
The reward was getting
released from his cage.

414
00:25:14,590 --> 00:25:16,985
So that was certainly
a conditioned action,

415
00:25:16,985 --> 00:25:19,870
a kind of artificial learning.

416
00:25:19,870 --> 00:25:23,000
But then he points out that
a lot of actions you just

417
00:25:23,000 --> 00:25:23,900
can't do that with.

418
00:25:23,900 --> 00:25:25,550
You can't condition
them like that.

419
00:25:25,550 --> 00:25:30,260
Like tendon reflexes,
sexual action

420
00:25:30,260 --> 00:25:31,670
patterns in most animals.

421
00:25:31,670 --> 00:25:34,270
You can't condition like that.

422
00:25:34,270 --> 00:25:36,190
Bill shaking in mallards.

423
00:25:36,190 --> 00:25:38,910
And I did a lot of
work on hamsters.

424
00:25:38,910 --> 00:25:41,095
And I tried to condition
their orienting.

425
00:25:45,950 --> 00:25:50,281
For example, normally you
present the stimulus here,

426
00:25:50,281 --> 00:25:50,780
he turns.

427
00:25:50,780 --> 00:25:52,470
It's an innate response.

428
00:25:52,470 --> 00:25:55,080
And we know the pathway
in the brain from retina

429
00:25:55,080 --> 00:25:57,830
to the midbrain tectum and
the superior colliculus

430
00:25:57,830 --> 00:25:58,690
that controls it.

431
00:25:58,690 --> 00:26:02,950
And then we know the
output pathway as well.

432
00:26:02,950 --> 00:26:06,110
But what if you never
rewarded them when they turn?

433
00:26:06,110 --> 00:26:11,816
You reward them over here
and stimulate their whiskers

434
00:26:11,816 --> 00:26:12,315
afterwards.

435
00:26:12,315 --> 00:26:15,300
So you present it
visually over here.

436
00:26:15,300 --> 00:26:17,230
And then you immediately
touch the whiskers

437
00:26:17,230 --> 00:26:19,320
and they turn and
get the reward.

438
00:26:19,320 --> 00:26:20,470
What happens?

439
00:26:20,470 --> 00:26:25,180
Well, you've seen the
conditioning action.

440
00:26:25,180 --> 00:26:27,100
He does learn.

441
00:26:27,100 --> 00:26:29,340
But if you look at the
behavior in detail,

442
00:26:29,340 --> 00:26:33,050
you'll see that actually
there's a little hesitation.

443
00:26:33,050 --> 00:26:35,560
He'll usually start to
turn the correct direction,

444
00:26:35,560 --> 00:26:39,080
and then he'll inhibit that.

445
00:26:39,080 --> 00:26:41,780
He learns to
inhibit the behavior

446
00:26:41,780 --> 00:26:44,040
and then turn to get the reward.

447
00:26:44,040 --> 00:26:47,860
And if he's really
learned that and there's

448
00:26:47,860 --> 00:26:50,650
any disturbance, like a
novel stimulus, something

449
00:26:50,650 --> 00:26:56,100
that changes the
environment a little bit,

450
00:26:56,100 --> 00:26:59,940
the normal behavior
comes right back.

451
00:26:59,940 --> 00:27:02,160
You can't release press.

452
00:27:02,160 --> 00:27:07,930
You can't condition it in the
sense of the earlier examples

453
00:27:07,930 --> 00:27:09,580
here.

454
00:27:09,580 --> 00:27:11,290
The innate responses
will dominate.

455
00:27:16,180 --> 00:27:18,120
The other situation
was when I was

456
00:27:18,120 --> 00:27:21,620
able to get the optic
tract to literally go

457
00:27:21,620 --> 00:27:25,020
to the wrong side of the brain
by an early brain lesion.

458
00:27:25,020 --> 00:27:27,139
So again, you
present the stimulus,

459
00:27:27,139 --> 00:27:28,680
they always turn
the wrong direction.

460
00:27:28,680 --> 00:27:31,290
But they never get
rewarded for that

461
00:27:31,290 --> 00:27:33,230
unless you do it artificially.

462
00:27:33,230 --> 00:27:35,030
So what happens?

463
00:27:35,030 --> 00:27:40,010
I found that in that area of the
visual field that's affected,

464
00:27:40,010 --> 00:27:43,510
they never do learn, in spite
of them always being rewarded.

465
00:27:47,749 --> 00:27:49,040
Because they're never rewarded.

466
00:27:49,040 --> 00:27:52,430
They never learn to turn
in the right direction

467
00:27:52,430 --> 00:27:55,590
even though it would
be adaptive to do so,

468
00:27:55,590 --> 00:27:59,150
because the connection
won't change just

469
00:27:59,150 --> 00:28:01,580
because of learning.

470
00:28:01,580 --> 00:28:04,555
So that's what's underlying
the innate behavior.

471
00:28:07,250 --> 00:28:12,420
And even if he learns
to totally suppress,

472
00:28:12,420 --> 00:28:17,550
that you will bring out--
it's very easy to train him.

473
00:28:17,550 --> 00:28:22,350
One reward where you're getting
the wrong direction training

474
00:28:22,350 --> 00:28:24,460
and it all comes right back.

475
00:28:24,460 --> 00:28:26,722
So conditioned
appetitive behavior

476
00:28:26,722 --> 00:28:27,680
directed at quiescence.

477
00:28:27,680 --> 00:28:30,050
Now this is very common.

478
00:28:30,050 --> 00:28:35,580
We're talking here about
annoyers, tension reduction.

479
00:28:35,580 --> 00:28:40,215
The tension or the annoyer is
hunger or thirst or stress,

480
00:28:40,215 --> 00:28:44,040
a state of stress or conflict.

481
00:28:44,040 --> 00:28:48,530
You classify habitat
selection this way

482
00:28:48,530 --> 00:28:51,400
because when an animal's
searching for a habitat,

483
00:28:51,400 --> 00:28:54,290
it's another drive that's
not being satisfied.

484
00:28:54,290 --> 00:28:55,180
He's uncomfortable.

485
00:28:55,180 --> 00:28:58,020
The drive is high, and
it can only be satisfied

486
00:28:58,020 --> 00:29:01,450
if he finds a
satisfactory habitat.

487
00:29:01,450 --> 00:29:07,490
And that reduces the tension
and causes his reward.

488
00:29:07,490 --> 00:29:09,910
So that's conditioned
appetitive behavior

489
00:29:09,910 --> 00:29:11,880
directed at quiescence.

490
00:29:11,880 --> 00:29:15,560
So it usually involves a
search for a stimulus situation

491
00:29:15,560 --> 00:29:16,770
that leads to a reward.

492
00:29:16,770 --> 00:29:20,400
It's another kind of
stimulus selection.

493
00:29:20,400 --> 00:29:24,280
The reward is the reduction
in the tension, the stress,

494
00:29:24,280 --> 00:29:28,970
the anxiety, the
annoyance, whatever it is.

495
00:29:28,970 --> 00:29:34,920
So most of the kind of
learning studied in rats,

496
00:29:34,920 --> 00:29:37,980
where they're looking at what we
call instrumental conditioning,

497
00:29:37,980 --> 00:29:41,350
is actually stimulus selection.

498
00:29:41,350 --> 00:29:46,110
So we'll talk about
response selection.

499
00:29:46,110 --> 00:29:50,120
Now, operant conditioning.

500
00:29:50,120 --> 00:29:51,385
Very different from this.

501
00:29:54,550 --> 00:29:58,750
Even though a lot of scientists
group these two, conditioned

502
00:29:58,750 --> 00:30:01,360
appetitive behavior directed
at quiescence and operant

503
00:30:01,360 --> 00:30:03,640
conditioning, together.

504
00:30:03,640 --> 00:30:06,270
So this is a kind of
response selection.

505
00:30:06,270 --> 00:30:09,570
Very highly studied, like
in Skinnerian conditioning.

506
00:30:09,570 --> 00:30:12,540
Like you put a cat
in a puzzle box,

507
00:30:12,540 --> 00:30:17,960
and he will try every
response that he's able.

508
00:30:17,960 --> 00:30:19,960
He will try everything
in order to--

509
00:30:19,960 --> 00:30:23,380
and eventually,
if possible, he'll

510
00:30:23,380 --> 00:30:25,080
discover some
response that helps

511
00:30:25,080 --> 00:30:27,310
him get out of the puzzle box.

512
00:30:27,310 --> 00:30:33,050
You put a male
dog eager to mate,

513
00:30:33,050 --> 00:30:37,760
and the female dog,
the bitch, is confined,

514
00:30:37,760 --> 00:30:39,670
but there is some
way to get to her,

515
00:30:39,670 --> 00:30:44,210
he will try every response
in order to get to her.

516
00:30:44,210 --> 00:30:45,700
So that's response selection.

517
00:30:45,700 --> 00:30:49,820
He's trying to find a
response that works.

518
00:30:49,820 --> 00:30:55,040
He points out that in nature
operant conditioning is rarer

519
00:30:55,040 --> 00:30:58,190
than generally assumed.

520
00:30:58,190 --> 00:31:00,970
It occurs in situations of
appetitive behavior directed

521
00:31:00,970 --> 00:31:03,360
at quiescence, or in
exploratory behavior.

522
00:31:06,460 --> 00:31:08,885
Now we come to real
motor learning,

523
00:31:08,885 --> 00:31:13,510
where the motor responses
themselves can change.

524
00:31:16,560 --> 00:31:19,086
We know that skilled
movements can be shaped.

525
00:31:23,350 --> 00:31:27,710
We try this in
sports all the time.

526
00:31:27,710 --> 00:31:32,070
In nature you'll see
animals acquire path habits.

527
00:31:32,070 --> 00:31:34,420
When they acquire
a path habit they

528
00:31:34,420 --> 00:31:36,905
can follow that path
with much greater speed.

529
00:31:40,300 --> 00:31:42,860
It includes things like
recitation from memory.

530
00:31:42,860 --> 00:31:45,360
You can learn to
rattle off a poem.

531
00:31:45,360 --> 00:31:47,690
Even if you forget even
what the poem's about,

532
00:31:47,690 --> 00:31:51,290
you can learn all the
words and everything.

533
00:31:51,290 --> 00:31:53,200
And there are examples
in animal behavior.

534
00:31:53,200 --> 00:31:57,920
For example, a little
vole kept as a pet by,

535
00:31:57,920 --> 00:32:02,340
I think it was Lorenz,
always followed,

536
00:32:02,340 --> 00:32:06,450
ran in a certain pattern
to get back to his nest.

537
00:32:06,450 --> 00:32:08,244
And animals learn
this all the time.

538
00:32:08,244 --> 00:32:09,910
When they're in their
local environment,

539
00:32:09,910 --> 00:32:14,600
they learn actions that bring
them rapidly back to the nest.

540
00:32:14,600 --> 00:32:17,380
And if there's one action
that they are commonly

541
00:32:17,380 --> 00:32:21,170
done-- like it might
involve leaping

542
00:32:21,170 --> 00:32:24,760
onto an obstacle on
their way to the nest.

543
00:32:24,760 --> 00:32:28,210
Well, suddenly you just
take the obstacle away.

544
00:32:28,210 --> 00:32:32,610
This is how you show it's
a learned motor pattern.

545
00:32:32,610 --> 00:32:34,470
He has shaped his
motor behavior,

546
00:32:34,470 --> 00:32:36,460
because he still goes
through the whole thing.

547
00:32:36,460 --> 00:32:39,870
And suddenly that object's
not there to land on.

548
00:32:39,870 --> 00:32:40,750
And what happens?

549
00:32:40,750 --> 00:32:45,600
He leaps anyway and hits a
big surprise, because he falls

550
00:32:45,600 --> 00:32:49,140
further and lands on the ground.

551
00:32:49,140 --> 00:32:52,200
These kinds of
learned habits can

552
00:32:52,200 --> 00:32:54,650
be pretty resistant to change.

553
00:32:54,650 --> 00:32:57,330
They have a lot of similarities
to fixed motor patterns.

554
00:32:57,330 --> 00:33:01,220
In fact, they acquire their
own appetitive behavior,

555
00:33:01,220 --> 00:33:01,955
in many cases.

556
00:33:05,020 --> 00:33:08,140
People that learn particular
patterns of movement in sports

557
00:33:08,140 --> 00:33:10,440
can become very motivated
to do those things.

558
00:33:14,780 --> 00:33:17,860
And we know that different
parts of the brain

559
00:33:17,860 --> 00:33:23,370
are involved in changing
motor patterns, especially

560
00:33:23,370 --> 00:33:24,925
cerebellar mechanisms.

561
00:33:31,070 --> 00:33:34,340
Just more about shaping
of motor patterns.

562
00:33:34,340 --> 00:33:37,650
There's a lot of species
differences here.

563
00:33:37,650 --> 00:33:42,250
Let's just remind ourselves
of species differences

564
00:33:42,250 --> 00:33:47,560
in locomotion, because it gives
an example of what you can

565
00:33:47,560 --> 00:33:50,970
train and what you cannot
train an animal to do.

566
00:33:50,970 --> 00:33:54,450
So horses are very adapted
to running on flat ground.

567
00:33:54,450 --> 00:33:57,210
They can't adjust
well to uneven terrain

568
00:33:57,210 --> 00:33:59,270
no matter how much
you try to train them.

569
00:33:59,270 --> 00:34:04,320
They're very limited in the
responses available to them,

570
00:34:04,320 --> 00:34:06,530
the types of movements
they can make.

571
00:34:06,530 --> 00:34:08,500
Donkeys are more
sure-footed because they've

572
00:34:08,500 --> 00:34:10,639
got more responses available.

573
00:34:10,639 --> 00:34:15,415
Goats, the chamois, they're
adapted to hills and mountains.

574
00:34:15,415 --> 00:34:16,719
They're very sure-footed.

575
00:34:16,719 --> 00:34:19,159
So what is that sure-footedness?

576
00:34:19,159 --> 00:34:23,110
Smaller units of action.

577
00:34:23,110 --> 00:34:31,980
And with smaller units,
they can adjust direction,

578
00:34:31,980 --> 00:34:38,330
adjust for foot placement more
readily in these rapid actions.

579
00:34:38,330 --> 00:34:41,780
In other words, higher control
has more possibilities.

580
00:34:41,780 --> 00:34:45,030
There's more movement
patterns to control.

581
00:34:45,030 --> 00:34:46,960
And that leads to a
discussion of what

582
00:34:46,960 --> 00:34:49,830
we mean by voluntary action.

583
00:34:49,830 --> 00:34:53,065
He points out that
arboreal creatures,

584
00:34:53,065 --> 00:34:59,360
they need a lot
more action patterns

585
00:34:59,360 --> 00:35:00,965
to negotiate the trees.

586
00:35:06,030 --> 00:35:10,000
So it's a lot easier
to shape movements

587
00:35:10,000 --> 00:35:13,000
in primates and
squirrels and tree shrews

588
00:35:13,000 --> 00:35:14,780
and so forth that live in trees.

589
00:35:18,780 --> 00:35:22,140
We talk about will
or volition when

590
00:35:22,140 --> 00:35:25,340
it really amounts to higher
level initiation of fixed motor

591
00:35:25,340 --> 00:35:30,050
patterns, or learned
sequences, skilled movements.

592
00:35:30,050 --> 00:35:32,930
And we've already talked
about species differences

593
00:35:32,930 --> 00:35:37,050
on how small the elements
of movements are.

594
00:35:37,050 --> 00:35:40,950
And the example he gives about
voluntary movement and insight

595
00:35:40,950 --> 00:35:46,110
concerns his studies of geese.

596
00:35:46,110 --> 00:35:50,020
He says that their spatial
insight can exceed their motor

597
00:35:50,020 --> 00:35:54,050
abilities because they don't
have so many different motor

598
00:35:54,050 --> 00:35:55,310
patterns available to them.

599
00:35:55,310 --> 00:35:58,520
So you can teach them to
climb stairs fairly well.

600
00:35:58,520 --> 00:36:01,000
But when you try to teach them
how to descend a stairway,

601
00:36:01,000 --> 00:36:05,290
they can't even learn to adjust
the length of their stride

602
00:36:05,290 --> 00:36:05,790
well enough.

603
00:36:08,950 --> 00:36:10,910
And they will stumble.

604
00:36:10,910 --> 00:36:14,065
They have a lot of trouble
going down the stairs.

605
00:36:16,840 --> 00:36:20,550
So species differ
a lot in how much

606
00:36:20,550 --> 00:36:25,520
they can control
when they want to.

607
00:36:25,520 --> 00:36:27,390
Voluntary control, we call it.

608
00:36:27,390 --> 00:36:31,120
It's the influence of
motivational systems

609
00:36:31,120 --> 00:36:35,080
on higher level control systems.

610
00:36:35,080 --> 00:36:36,800
And then he comes
to a discussion

611
00:36:36,800 --> 00:36:38,090
of exploratory behavior.

612
00:36:48,800 --> 00:36:50,550
Exploratory behavior
or curiosity.

613
00:36:53,794 --> 00:36:55,085
And this is pretty interesting.

614
00:36:58,820 --> 00:37:01,880
When you study exploratory
behavior, of course,

615
00:37:01,880 --> 00:37:04,000
it's very influenced by novelty.

616
00:37:04,000 --> 00:37:05,690
Many animals are,
in fact, highly

617
00:37:05,690 --> 00:37:09,570
motivated to explore
novel situations,

618
00:37:09,570 --> 00:37:14,770
and they will direct multiple
different patterns of behavior,

619
00:37:14,770 --> 00:37:16,170
sometimes unrelated.

620
00:37:16,170 --> 00:37:17,290
Even the little hamster.

621
00:37:17,290 --> 00:37:20,500
You put them in a novel
room with different objects

622
00:37:20,500 --> 00:37:24,300
that they're not
accustomed to seeing.

623
00:37:24,300 --> 00:37:25,460
It's amazing.

624
00:37:25,460 --> 00:37:30,960
Their initial response, if it's
very novel, will be fright.

625
00:37:30,960 --> 00:37:33,520
Their tension levels go way up.

626
00:37:33,520 --> 00:37:37,200
But in fact, if you've
got them in a situation

627
00:37:37,200 --> 00:37:39,206
that they're used
to, you test them

628
00:37:39,206 --> 00:37:42,370
in an apparatus they're
very familiar with,

629
00:37:42,370 --> 00:37:43,830
they will explore.

630
00:37:43,830 --> 00:37:48,750
And you will see them
trying all different kinds

631
00:37:48,750 --> 00:37:49,500
of motor patterns.

632
00:37:49,500 --> 00:37:52,000
You'll see rapid switching
between different behavior

633
00:37:52,000 --> 00:37:53,395
patterns or motor patterns.

634
00:37:58,610 --> 00:38:02,510
And when there's a strong
action-specific potential

635
00:38:02,510 --> 00:38:06,130
behind that's causing
those motor patterns,

636
00:38:06,130 --> 00:38:09,580
you don't see that switching
from one to the other.

637
00:38:09,580 --> 00:38:13,050
So in exploratory
behavior, there's

638
00:38:13,050 --> 00:38:21,030
a difference in the structure
of what's behind the movement.

639
00:38:21,030 --> 00:38:24,350
And exploratory behavior
is not as well understood

640
00:38:24,350 --> 00:38:28,600
as fixed action patterns,
because now the motor part

641
00:38:28,600 --> 00:38:32,710
can become uncoupled
with the normal action.

642
00:38:32,710 --> 00:38:34,320
The motivation
here is curiosity.

643
00:38:37,020 --> 00:38:40,150
Very similar to play behavior.

644
00:38:40,150 --> 00:38:44,600
He says it occurs in a
field devoid of tension.

645
00:38:44,600 --> 00:38:46,660
And yet the
motivation to explore

646
00:38:46,660 --> 00:38:50,730
can be so strong, an
animal sometimes prefer

647
00:38:50,730 --> 00:38:53,740
that rather than eating,
even if he's hungry.

648
00:38:56,430 --> 00:39:00,340
And it does have a function,
very important functions.

649
00:39:00,340 --> 00:39:02,835
When the environment
changes, it helps the animal,

650
00:39:02,835 --> 00:39:07,890
it's adaptive for the
animal to explore it.

651
00:39:07,890 --> 00:39:10,570
He needs to learn the spatial
layout of the environment

652
00:39:10,570 --> 00:39:13,910
around him in relation
to visual landmarks

653
00:39:13,910 --> 00:39:16,490
and other sensory cues,
because if they change,

654
00:39:16,490 --> 00:39:19,200
then he's in trouble if he gets
mixed up trying to get home,

655
00:39:19,200 --> 00:39:21,830
trying to escape.

656
00:39:21,830 --> 00:39:23,735
It makes his life a
lot more efficient

657
00:39:23,735 --> 00:39:27,090
if he stays familiar
with the novel situation.

658
00:39:27,090 --> 00:39:29,930
There were experiments
on this using hamsters

659
00:39:29,930 --> 00:39:33,660
because they're so curious.

660
00:39:33,660 --> 00:39:35,480
Actually, my first
publication's called

661
00:39:35,480 --> 00:39:37,330
"Curiosity and the Hamster."

662
00:39:37,330 --> 00:39:41,130
I was using exploratory
behavior to reward them

663
00:39:41,130 --> 00:39:43,160
for doing other things.

664
00:39:43,160 --> 00:39:45,470
Reward them for pressing
a bar, reward them

665
00:39:45,470 --> 00:39:47,380
for going through doors.

666
00:39:47,380 --> 00:39:49,750
It was so novel at
the time, people still

667
00:39:49,750 --> 00:39:53,015
believed that this wasn't
supposed to be possible.

668
00:39:55,740 --> 00:39:58,005
I had to play around at
home with a pet hamster,

669
00:39:58,005 --> 00:40:01,190
and I found out I could teach
him to press a bar made out

670
00:40:01,190 --> 00:40:04,604
of a knife taped
to a baby rattle.

671
00:40:04,604 --> 00:40:06,020
And I could teach
them to do this.

672
00:40:06,020 --> 00:40:08,900
And my professor was
so intrigued by that,

673
00:40:08,900 --> 00:40:11,600
he got me to write
a paper about it.

674
00:40:11,600 --> 00:40:13,140
And that was because
of the theories

675
00:40:13,140 --> 00:40:21,030
that were dominating American
psychology at that time.

676
00:40:21,030 --> 00:40:24,680
But Catherine Blanc in Europe--
I think it was in France,

677
00:40:24,680 --> 00:40:27,370
but I don't remember
for sure anymore--

678
00:40:27,370 --> 00:40:29,550
but she studied
this experimentally.

679
00:40:29,550 --> 00:40:35,830
And she found out that in
their exploratory behavior,

680
00:40:35,830 --> 00:40:37,730
they are acquiring
knowledge that they

681
00:40:37,730 --> 00:40:40,070
can use in other situations.

682
00:40:40,070 --> 00:40:42,590
They become more efficient
in finding their way

683
00:40:42,590 --> 00:40:43,465
in another situation.

684
00:40:46,030 --> 00:40:49,040
And then Lorenz points out
that exploratory behavior

685
00:40:49,040 --> 00:40:55,550
is especially highly developed
in unspecialized species.

686
00:40:55,550 --> 00:41:01,130
He says their specialization is
being versatile, like humans,

687
00:41:01,130 --> 00:41:03,650
like rats, like ravens.

688
00:41:03,650 --> 00:41:06,350
These are animals
that aren't so limited

689
00:41:06,350 --> 00:41:08,960
to one type of environment.

690
00:41:08,960 --> 00:41:12,900
We should add crows to that.

691
00:41:12,900 --> 00:41:22,010
Mice are certainly more general
than most voles, for example,

692
00:41:22,010 --> 00:41:25,900
because they can adapt to
more different situations.

693
00:41:25,900 --> 00:41:28,880
And all of this
type of animal shows

694
00:41:28,880 --> 00:41:30,360
a lot of exploratory behavior.

695
00:41:35,080 --> 00:41:38,426
AUDIENCE: What would be
a specialized species?

696
00:41:38,426 --> 00:41:41,060
PROFESSOR: Most of the species
we've been talking about

697
00:41:41,060 --> 00:41:43,830
are pretty specialized.

698
00:41:43,830 --> 00:41:50,300
Every species will show some--
I think all of the mammals

699
00:41:50,300 --> 00:41:52,640
show some curiosity.

700
00:41:52,640 --> 00:41:55,930
But the amount that they
will show will vary.

701
00:42:01,460 --> 00:42:04,460
What is characteristic about
ravens, rats, and humans?

702
00:42:08,320 --> 00:42:10,190
The first thing you
think of is that they've

703
00:42:10,190 --> 00:42:13,320
adapted to so many
different environments,

704
00:42:13,320 --> 00:42:15,710
especially if you group
the corvids together,

705
00:42:15,710 --> 00:42:16,925
the ravens and the crows.

706
00:42:20,950 --> 00:42:25,530
We know that rats can
be found everywhere.

707
00:42:25,530 --> 00:42:28,420
The same is true for
different corvids.

708
00:42:28,420 --> 00:42:31,690
They have specific
adaptations, of course,

709
00:42:31,690 --> 00:42:36,670
that makes the jackdaw
common in parts of Europe

710
00:42:36,670 --> 00:42:40,120
and the crow much
more common over here.

711
00:42:40,120 --> 00:42:43,630
Humans, of course, are
the most versatile,

712
00:42:43,630 --> 00:42:45,820
the least specialized.

713
00:42:45,820 --> 00:42:51,480
That doesn't mean-- we do
depend a lot on learning.

714
00:42:51,480 --> 00:42:54,030
But what is different
about-- for example,

715
00:42:54,030 --> 00:42:57,210
back here, when we're
talking about shaping

716
00:42:57,210 --> 00:43:02,750
of motor learning here.

717
00:43:02,750 --> 00:43:08,360
It depends a lot on how many
fixed action patterns you have.

718
00:43:08,360 --> 00:43:12,740
The more fixed action patterns
that make up your movements,

719
00:43:12,740 --> 00:43:16,140
the more versatile you are,
which is one reason humans

720
00:43:16,140 --> 00:43:17,980
are so versatile.

721
00:43:17,980 --> 00:43:21,165
We actually have more fixed
action patterns than animals.

722
00:43:23,700 --> 00:43:26,875
At least on the motor side,
we have many more choices.

723
00:43:38,840 --> 00:43:40,700
Some of the most
specialized animals

724
00:43:40,700 --> 00:43:42,740
are the ones that are
specialized for feeding

725
00:43:42,740 --> 00:43:46,820
on very particular
things, like the koala.

726
00:43:46,820 --> 00:43:49,060
They only can eat
certain things,

727
00:43:49,060 --> 00:43:52,260
and they don't survive
well in other situations.

728
00:43:52,260 --> 00:43:55,530
Why aren't hamsters--
why haven't they

729
00:43:55,530 --> 00:43:58,320
spread all over the US?

730
00:43:58,320 --> 00:44:05,530
They're just not adaptable like
the rat, or even the mouse.

731
00:44:05,530 --> 00:44:09,620
So they're only found--
they're native to Syria

732
00:44:09,620 --> 00:44:14,080
and the northern part
of what is now Israel.

733
00:44:14,080 --> 00:44:18,560
And if you even move towards
Turkey and that direction,

734
00:44:18,560 --> 00:44:19,730
the species changes.

735
00:44:19,730 --> 00:44:21,320
It's no longer Syrian hamster.

736
00:44:21,320 --> 00:44:24,730
It becomes a different hamster
that has adapted specifically

737
00:44:24,730 --> 00:44:26,480
to the higher elevations.

738
00:44:26,480 --> 00:44:35,330
If you go north, northeast,
you'll see the-- no,

739
00:44:35,330 --> 00:44:39,320
sorry, if you go further, you
go, say, into Romania, again,

740
00:44:39,320 --> 00:44:41,160
another species appears.

741
00:44:41,160 --> 00:44:45,430
You go the other way, you start
to get the Chinese hamster,

742
00:44:45,430 --> 00:44:47,890
or Siberian hamster
if you go further.

743
00:44:47,890 --> 00:44:49,440
They're all different.

744
00:44:49,440 --> 00:44:50,910
They've adapted differently.

745
00:44:50,910 --> 00:44:53,069
But the rat, you
keep finding the rat

746
00:44:53,069 --> 00:44:54,485
in all these
different situations.

747
00:44:57,460 --> 00:44:59,790
So any species that's
very limited to one

748
00:44:59,790 --> 00:45:05,080
habitat you wouldn't call
a generalized species.

749
00:45:05,080 --> 00:45:09,400
And that's most of the ones
that I'm talking about.

750
00:45:09,400 --> 00:45:09,900
OK.

751
00:45:09,900 --> 00:45:11,450
What about play behavior?

752
00:45:11,450 --> 00:45:15,740
Similar to explorative behavior.

753
00:45:15,740 --> 00:45:18,710
The very primitive
play are simply--

754
00:45:18,710 --> 00:45:21,530
you could call them in
vacua reactions motivated

755
00:45:21,530 --> 00:45:26,880
by action-specific potentials
or drives that are high.

756
00:45:26,880 --> 00:45:28,170
But there is a difference.

757
00:45:28,170 --> 00:45:35,540
You don't call it play if they
execute the entire sequence

758
00:45:35,540 --> 00:45:41,440
from a high action-specific
potential, a high drive level,

759
00:45:41,440 --> 00:45:45,040
to the motor pattern,
like running and chasing

760
00:45:45,040 --> 00:45:47,610
playful fighting
or prey catching.

761
00:45:47,610 --> 00:45:50,470
The animals will enjoy
the motor patterns,

762
00:45:50,470 --> 00:45:53,080
and they will do this even
before they're using them

763
00:45:53,080 --> 00:45:56,840
to catch prey, to
feed and so forth.

764
00:45:56,840 --> 00:46:00,340
So it's very common
in young animals.

765
00:46:00,340 --> 00:46:03,470
And he points out
that play can shift

766
00:46:03,470 --> 00:46:05,990
to the real thing
in some situations.

767
00:46:05,990 --> 00:46:07,610
There's a lot of
danger in playing

768
00:46:07,610 --> 00:46:11,550
with an adult tomcat or badger.

769
00:46:11,550 --> 00:46:14,030
You might have tamed it
and think it's your pet,

770
00:46:14,030 --> 00:46:22,900
but it can switch
suddenly to the real thing

771
00:46:22,900 --> 00:46:24,840
when their motivation
becomes high.

772
00:46:27,420 --> 00:46:29,630
So if you play with
an adult tomcat,

773
00:46:29,630 --> 00:46:32,497
you have to be pretty
careful because of that.

774
00:46:32,497 --> 00:46:33,580
You don't want to trigger.

775
00:46:33,580 --> 00:46:35,100
And that's certainly true.

776
00:46:35,100 --> 00:46:37,290
People get caught off
guard all the time

777
00:46:37,290 --> 00:46:42,110
when they raise a tiger or
other big cat in their home.

778
00:46:42,110 --> 00:46:47,310
Or even a champ or a monkey,
because even though they

779
00:46:47,310 --> 00:46:49,650
seem to be playing,
it can switch suddenly

780
00:46:49,650 --> 00:46:51,840
to the real thing.

781
00:46:51,840 --> 00:46:56,590
And they can kill you or
your child or whatever.

782
00:46:56,590 --> 00:46:58,230
So it's dangerous to do that.

783
00:47:04,310 --> 00:47:08,335
He points out that, on
the motivational side,

784
00:47:08,335 --> 00:47:13,310
a cat that has no opportunity
to catch and kill prey

785
00:47:13,310 --> 00:47:16,970
can compensate by playing.

786
00:47:16,970 --> 00:47:19,190
It doesn't mean that
it's always totally safe,

787
00:47:19,190 --> 00:47:20,774
but they appear to do that.

788
00:47:20,774 --> 00:47:22,190
So how do you tell
the difference?

789
00:47:22,190 --> 00:47:25,530
Well, one way is
physiologically.

790
00:47:25,530 --> 00:47:28,890
Because in play, the
autonomic nervous system

791
00:47:28,890 --> 00:47:30,840
isn't involved in the same way.

792
00:47:30,840 --> 00:47:33,030
So if you major
autonomic responses,

793
00:47:33,030 --> 00:47:38,730
you'll find big differences
between the way the heart rate,

794
00:47:38,730 --> 00:47:40,430
the breathing, blood
pressure changes,

795
00:47:40,430 --> 00:47:42,530
and so forth are responding
in the two situations.

796
00:47:45,062 --> 00:47:46,520
Now what are the
functions of play?

797
00:47:46,520 --> 00:47:51,010
And I just felt-- in the last
half of his career studied

798
00:47:51,010 --> 00:47:54,270
humans and wrote the
book Human Ecology.

799
00:47:54,270 --> 00:47:59,600
He studied development of play
in cats and other carnivores.

800
00:47:59,600 --> 00:48:03,970
He found evidence of their
learning of coordination,

801
00:48:03,970 --> 00:48:08,380
learning of stimulus selection,
so they could change,

802
00:48:08,380 --> 00:48:11,605
so they would respond to
more relevant stimuli.

803
00:48:14,260 --> 00:48:17,150
And also they would
invent, in their play

804
00:48:17,150 --> 00:48:19,850
they would invent
movement patterns simply

805
00:48:19,850 --> 00:48:25,940
by linking different elements,
the different inherited

806
00:48:25,940 --> 00:48:29,070
elements of their behavior.

807
00:48:29,070 --> 00:48:31,120
And they will do
that even though it

808
00:48:31,120 --> 00:48:32,530
doesn't lead to any reward.

809
00:48:32,530 --> 00:48:38,050
The reward is simply
doing it in the play.

810
00:48:38,050 --> 00:48:41,675
And he has evidence for all
these advantages of play.

811
00:48:45,490 --> 00:48:47,890
And then-- this is
interesting-- he

812
00:48:47,890 --> 00:48:50,630
says human research is
exploratory behavior

813
00:48:50,630 --> 00:48:53,800
plus play, mostly
exploratory behavior.

814
00:48:53,800 --> 00:48:57,010
And human art is mostly playing.

815
00:48:57,010 --> 00:49:00,610
He's not putting it down at all.

816
00:49:00,610 --> 00:49:05,800
He's just pointed out the value
of these things in animals.

817
00:49:05,800 --> 00:49:09,370
And now he says, well,
if we talk about humans,

818
00:49:09,370 --> 00:49:11,665
this is how he sees
human research and art.

819
00:49:17,130 --> 00:49:18,960
Just about all these
types of learning,

820
00:49:18,960 --> 00:49:21,190
remember there are
species differences,

821
00:49:21,190 --> 00:49:24,770
but we don't know fully
all aspects of it.

822
00:49:24,770 --> 00:49:26,510
And what I will
talk about next time

823
00:49:26,510 --> 00:49:29,040
is I'll talk a little bit
about brain localization.

824
00:49:29,040 --> 00:49:32,180
We can at least make educated
guesses in some cases.

825
00:49:32,180 --> 00:49:35,100
Real experiments have been done.

826
00:49:35,100 --> 00:49:39,750
And whether there are
any general rules.

827
00:49:39,750 --> 00:49:41,380
The rules seem to be
a little different

828
00:49:41,380 --> 00:49:43,720
for all of these different
types of learning

829
00:49:43,720 --> 00:49:46,190
that Lorenz has described.

830
00:49:46,190 --> 00:49:53,370
So next time we'll be talking
here more about brain pathways.

831
00:49:53,370 --> 00:49:57,440
Neuroscience has made
studies, for example,

832
00:49:57,440 --> 00:49:59,670
of hippocampal function
and the function

833
00:49:59,670 --> 00:50:04,340
of other parts of the brain--
have given us a different way

834
00:50:04,340 --> 00:50:06,730
to categorize types of learning.

835
00:50:06,730 --> 00:50:09,315
So we'll talk about
that next time.