1
00:00:00,070 --> 00:00:01,670
The following
content is provided

2
00:00:01,670 --> 00:00:03,820
under a Creative
Commons license.

3
00:00:03,820 --> 00:00:06,550
Your support will help MIT
OpenCourseWare continue

4
00:00:06,550 --> 00:00:10,160
to offer high-quality
educational resources for free.

5
00:00:10,160 --> 00:00:12,700
To make a donation, or to
view additional materials

6
00:00:12,700 --> 00:00:16,620
from hundreds of MIT courses,
visit MIT OpenCourseWare

7
00:00:16,620 --> 00:00:17,275
at ocw.mit.edu.

8
00:00:25,620 --> 00:00:29,600
PROFESSOR: Last time we
talked about processing

9
00:00:29,600 --> 00:00:33,780
of cues that are used for
binaural localization of sound.

10
00:00:33,780 --> 00:00:37,936
Those being the interaural
time and interaural level

11
00:00:37,936 --> 00:00:38,435
differences.

12
00:00:41,400 --> 00:00:45,550
And we have those cues because
the ears are physically

13
00:00:45,550 --> 00:00:48,160
separated on the
sides of our head.

14
00:00:49,380 --> 00:00:52,780
And because of the physical
characteristics of sound,

15
00:00:52,780 --> 00:00:55,340
for example, the
velocity of sound in air.

16
00:00:57,420 --> 00:01:02,230
And we talked toward the
end of last time's lecture

17
00:01:02,230 --> 00:01:06,370
about the neural processing
of interaural time differences

18
00:01:06,370 --> 00:01:09,995
in the Medial Superior
Olive, or MSO.

19
00:01:11,560 --> 00:01:14,140
And we talked about
the Jeffress model

20
00:01:14,140 --> 00:01:20,880
and how it can help recreate
a neural mapping that

21
00:01:20,880 --> 00:01:22,360
has ITD sensitivity.

22
00:01:22,360 --> 00:01:27,850
And that's the subject of
the assignment for this year.

23
00:01:27,850 --> 00:01:33,040
So I just put the text
for the assignment

24
00:01:33,040 --> 00:01:36,430
up here, just to
mention that there

25
00:01:36,430 --> 00:01:39,420
was something added
on at the end.

26
00:01:39,420 --> 00:01:43,350
But the first paragraph
of the assignment

27
00:01:43,350 --> 00:01:49,005
is basically the lecture and the
sketch of the Jeffress model.

28
00:01:50,930 --> 00:01:53,990
Except that these
days, people are

29
00:01:53,990 --> 00:01:58,940
thinking that the axonal
delay lines are not

30
00:01:58,940 --> 00:02:03,190
how the delay is created in
the medial superior olive.

31
00:02:03,190 --> 00:02:09,530
And so I ask you for two other
possible neural mechanisms

32
00:02:09,530 --> 00:02:11,880
that weren't originally
postulated by Jeffress

33
00:02:11,880 --> 00:02:13,365
that could create delays.

34
00:02:15,650 --> 00:02:18,470
So that's sort of the
heart of the assignment.

35
00:02:19,800 --> 00:02:24,940
The second paragraph talks
about some further updates

36
00:02:24,940 --> 00:02:26,040
to the Jeffress model.

37
00:02:27,320 --> 00:02:31,790
And there is a paper
that is attached

38
00:02:31,790 --> 00:02:33,660
to last time's lecture.

39
00:02:33,660 --> 00:02:39,470
The paper by [? Brandt ?] et al
where they discuss extensively

40
00:02:39,470 --> 00:02:42,880
some challenges, they call
them, to the Jeffress model.

41
00:02:42,880 --> 00:02:45,420
And I might call them amendments
to the Jeffress model,

42
00:02:45,420 --> 00:02:47,470
but things that
weren't originally

43
00:02:47,470 --> 00:02:50,890
postulated by Jeffress that
have come to light because

44
00:02:50,890 --> 00:02:56,800
of more recent experimental
studies that don't fit

45
00:02:56,800 --> 00:02:58,890
with the original version
and their updates.

46
00:03:00,800 --> 00:03:02,830
And finally, we've been
talking quite a lot

47
00:03:02,830 --> 00:03:04,710
about cochlear implants.

48
00:03:04,710 --> 00:03:07,190
And there's a very nice
passage on cochlear implants

49
00:03:07,190 --> 00:03:10,120
in the textbook that I
encourage you to read.

50
00:03:10,120 --> 00:03:12,860
And the final paragraph
of the assignment

51
00:03:12,860 --> 00:03:16,860
is, what problems would
cochlear implant users, even

52
00:03:16,860 --> 00:03:22,120
those with a left and a
right cochlear implant, have

53
00:03:22,120 --> 00:03:26,865
if they use the Jeffress
model to localize sounds?

54
00:03:28,330 --> 00:03:32,760
So that's a little
bit of thinking

55
00:03:32,760 --> 00:03:34,520
to solve that last problem, too.

56
00:03:34,520 --> 00:03:37,080
But based on what
we have talked about

57
00:03:37,080 --> 00:03:39,950
and what the textbook talks
about for cochlear implants,

58
00:03:39,950 --> 00:03:43,440
you should be able to come
up with an answer for that.

59
00:03:43,440 --> 00:03:45,950
So any questions
on the assignment?

60
00:03:45,950 --> 00:03:49,160
It relates very heavily
to last time's lecture

61
00:03:49,160 --> 00:03:51,140
on the Jeffress
model and the MSO.

62
00:03:53,560 --> 00:03:56,470
And I guess-- it
doesn't say here,

63
00:03:56,470 --> 00:03:59,730
but three to five pages would
be appropriate, I think.

64
00:04:02,380 --> 00:04:05,720
And it's due on
December 4, which

65
00:04:05,720 --> 00:04:08,600
is the day of the lab tour.

66
00:04:08,600 --> 00:04:12,440
So on that class day, which
is a week from Wednesday,

67
00:04:12,440 --> 00:04:14,750
we'll meet at the
Massachusetts Eye and Ear

68
00:04:14,750 --> 00:04:17,399
Infirmary instead
of meeting here.

69
00:04:17,399 --> 00:04:19,010
So we'll send you a reminder.

70
00:04:19,010 --> 00:04:20,760
And I think now on
the website there

71
00:04:20,760 --> 00:04:23,825
are directions to get to
Mass Eye and Ear Infirmary.

72
00:04:25,080 --> 00:04:28,780
It's just a simple one-stop
on the Red Line subway

73
00:04:28,780 --> 00:04:29,440
to get there.

74
00:04:30,810 --> 00:04:32,570
So the assignment
will be due then.

75
00:04:32,570 --> 00:04:34,080
And then I'll be
able to grade it

76
00:04:34,080 --> 00:04:36,200
before we have a review session.

77
00:04:36,200 --> 00:04:41,100
And we'll talk about the correct
answers for this assignment

78
00:04:41,100 --> 00:04:43,460
at the time of the
review session, which

79
00:04:43,460 --> 00:04:46,030
is the next class
after December 4.

80
00:04:47,560 --> 00:04:51,750
So today, I want to
march into new things.

81
00:04:53,100 --> 00:04:55,630
Now, we're going to talk
about neural processing

82
00:04:55,630 --> 00:04:57,630
of interaural level differences.

83
00:04:58,660 --> 00:05:02,770
Remember if a sound is off
to one side of my head,

84
00:05:02,770 --> 00:05:06,550
it's going to appear at a higher
sound level in the ear facing

85
00:05:06,550 --> 00:05:12,190
that sound source than
it will appear in the ear

86
00:05:12,190 --> 00:05:13,860
away from the sound source.

87
00:05:13,860 --> 00:05:17,450
So we talked about how big
these cues were last time.

88
00:05:17,450 --> 00:05:21,800
And they can, for high
frequencies, be maximally 20 dB

89
00:05:21,800 --> 00:05:24,450
in Interaural Level
Difference, or ILD.

90
00:05:27,020 --> 00:05:30,820
So those are processed,
at least starting

91
00:05:30,820 --> 00:05:33,410
in the lateral
superior olive, which

92
00:05:33,410 --> 00:05:37,080
is another nucleus in the
superior olivary complex, which

93
00:05:37,080 --> 00:05:41,220
is close to the medial superior
olive we've been talking about.

94
00:05:41,220 --> 00:05:43,985
So we'll talk about
the process in the LSO.

95
00:05:45,650 --> 00:05:47,820
Then, we'll talk
about projections

96
00:05:47,820 --> 00:05:51,080
of that nucleus and the MSO.

97
00:05:51,080 --> 00:05:52,990
And other parts of
the superior olive

98
00:05:52,990 --> 00:05:56,270
to the next higher
center in the brainstem,

99
00:05:56,270 --> 00:06:00,830
which in the auditory pathway is
the Inferior Colliculus, or IC.

100
00:06:00,830 --> 00:06:04,880
So we'll be talking extensively
about the inferior colliculus,

101
00:06:04,880 --> 00:06:07,730
which is a large
nucleus in the brainstem

102
00:06:07,730 --> 00:06:10,100
just caudal to the
superior colliculus

103
00:06:10,100 --> 00:06:12,800
that you folks talked
about extensively

104
00:06:12,800 --> 00:06:14,405
in the visual part
of the course.

105
00:06:16,690 --> 00:06:20,200
In the IC, you have some
interesting neural responses

106
00:06:20,200 --> 00:06:23,630
that relate to something
called time/intensity trading.

107
00:06:25,145 --> 00:06:26,530
I misspelled it.

108
00:06:26,530 --> 00:06:28,540
Not a very good speller, sorry.

109
00:06:36,770 --> 00:06:38,590
Which we'll talk
about and define.

110
00:06:38,590 --> 00:06:41,450
We'll talk about some
interesting characteristics

111
00:06:41,450 --> 00:06:46,280
of room acoustics, like
reflections off the wall,

112
00:06:46,280 --> 00:06:50,680
and how those don't completely
upset the apple cart of knowing

113
00:06:50,680 --> 00:06:53,190
where a sound is coming
from because of things

114
00:06:53,190 --> 00:06:54,670
like the precedence effect.

115
00:06:57,690 --> 00:07:01,450
And finally, we'll end up with
auditory pathways in the barn

116
00:07:01,450 --> 00:07:01,950
owl.

117
00:07:01,950 --> 00:07:05,900
So this species of birds
has very well-developed

118
00:07:05,900 --> 00:07:07,880
auditory systems.

119
00:07:07,880 --> 00:07:11,460
And some work at
Caltech has shown

120
00:07:11,460 --> 00:07:14,440
that these animals
have so-called "space"

121
00:07:14,440 --> 00:07:17,870
maps in a certain
part of their brain.

122
00:07:17,870 --> 00:07:19,590
That is, there's
a mapping of where

123
00:07:19,590 --> 00:07:24,730
the sound is in external
space into space

124
00:07:24,730 --> 00:07:29,620
in part of their brain
called the optic tectum.

125
00:07:29,620 --> 00:07:30,700
And we'll go over that.

126
00:07:30,700 --> 00:07:35,980
And the reading for today is on
how that space map is plastic

127
00:07:35,980 --> 00:07:37,845
and can be changed
by experience.

128
00:07:39,200 --> 00:07:42,430
So we'll talk about neuronal
plasticity of the space map.

129
00:07:44,270 --> 00:07:45,660
OK, so let's get started.

130
00:07:54,380 --> 00:07:58,040
So the neural processing of the
interaural level differences

131
00:07:58,040 --> 00:08:01,210
in the lateral superior olive.

132
00:08:01,210 --> 00:08:05,500
By contrast to what Jeffress
cooked up for the MSO,

133
00:08:05,500 --> 00:08:07,890
this neural circuit
is very simple.

134
00:08:09,520 --> 00:08:11,220
And here's how it runs.

135
00:08:11,220 --> 00:08:15,470
The Lateral Superior Olive on
the left side is here, the LSO.

136
00:08:15,470 --> 00:08:18,200
There's an LSO on the
right side as well,

137
00:08:18,200 --> 00:08:20,620
but the circuit is shown for
the one on the left side.

138
00:08:21,920 --> 00:08:24,450
This LSO on the
left side gets input

139
00:08:24,450 --> 00:08:27,065
from the cochlear
nucleus on the left side.

140
00:08:29,240 --> 00:08:30,780
That's excitatory input.

141
00:08:32,530 --> 00:08:35,765
And it gets input from the
right side that's inhibitory.

142
00:08:37,429 --> 00:08:40,659
And the way the
inhibitory input works

143
00:08:40,659 --> 00:08:45,560
is the cochlear nucleus
neurons on the right side

144
00:08:45,560 --> 00:08:50,140
project across the midline
and into another sub-nucleus

145
00:08:50,140 --> 00:08:53,830
of the superior olivary
complex designated here

146
00:08:53,830 --> 00:08:58,410
as the MNTB, which I think
is spelled out here--

147
00:08:58,410 --> 00:09:03,240
the Medial Nucleus of
the Trapezoid Body.

148
00:09:03,240 --> 00:09:06,610
So if you look at these
sections under the microscope,

149
00:09:06,610 --> 00:09:09,260
there's a lot of
crossing fibers here.

150
00:09:09,260 --> 00:09:11,957
And they sort of look, if you
have a lot of imagination,

151
00:09:11,957 --> 00:09:12,665
like a trapezoid.

152
00:09:14,000 --> 00:09:17,520
And this nucleus is within
those crossing fibers.

153
00:09:17,520 --> 00:09:20,290
So it's within the
trapezoid body.

154
00:09:20,290 --> 00:09:22,470
That's how it gets its name.

155
00:09:22,470 --> 00:09:27,270
And the MNTB neurons have
inhibitory neurotransmitter.

156
00:09:29,360 --> 00:09:33,400
So the cochlear nucleus
excites these MNTB neurons,

157
00:09:33,400 --> 00:09:36,520
but then they're inhibitory
and they send their axons

158
00:09:36,520 --> 00:09:37,320
to the LSO.

159
00:09:38,620 --> 00:09:42,820
And they spill out or
release their inhibitory

160
00:09:42,820 --> 00:09:45,170
neurotransmitter on
to the LSO neurons.

161
00:09:48,590 --> 00:09:56,280
So how does this
circuit work then

162
00:09:56,280 --> 00:10:00,500
if the sound is off to
this side of the slide,

163
00:10:00,500 --> 00:10:01,800
off to the left side?

164
00:10:03,460 --> 00:10:06,430
The sound will be of higher
level in the left ear

165
00:10:06,430 --> 00:10:09,160
and it will exert a
high excitatory effect

166
00:10:09,160 --> 00:10:10,850
on the cochlear nucleus here.

167
00:10:10,850 --> 00:10:14,260
And the LSO will be
excited in a big way.

168
00:10:14,260 --> 00:10:16,180
Of course, a little
bit of that sound

169
00:10:16,180 --> 00:10:19,910
is going to come over
here to the right side,

170
00:10:19,910 --> 00:10:23,490
but it won't excite the right
auditory nerve quite as much.

171
00:10:23,490 --> 00:10:27,130
And this pathway that
eventually becomes inhibitory

172
00:10:27,130 --> 00:10:30,470
won't be as strongly activated.

173
00:10:30,470 --> 00:10:33,240
And so the inhibition
will be less.

174
00:10:33,240 --> 00:10:38,210
So there's an interplay between
excitation here and inhibition.

175
00:10:38,210 --> 00:10:41,210
And in this case with the
sound off to the left side,

176
00:10:41,210 --> 00:10:43,420
the excitation will rule.

177
00:10:43,420 --> 00:10:49,180
And so here's a plot of the
firing of an individual LSO

178
00:10:49,180 --> 00:10:50,775
neuron that gets
all these inputs.

179
00:10:52,440 --> 00:10:54,730
And if the sound is off
to the left side, which

180
00:10:54,730 --> 00:10:58,520
is supposed to be this axis--
this is an axis of Interaural

181
00:10:58,520 --> 00:11:02,420
Level Difference,
or ILD, where here

182
00:11:02,420 --> 00:11:05,410
on the left side of the
graph the ipsilateral level

183
00:11:05,410 --> 00:11:08,280
is greater than the
contralateral level.

184
00:11:08,280 --> 00:11:11,390
That case, the LSO
neuron will be excited

185
00:11:11,390 --> 00:11:13,520
and it'll have a
lot of response,

186
00:11:13,520 --> 00:11:14,535
a high amount of firing.

187
00:11:17,600 --> 00:11:22,530
On the other hand, if the sound
is over here on the right side,

188
00:11:22,530 --> 00:11:25,970
it will activate in a big
way the right pathway,

189
00:11:25,970 --> 00:11:29,345
which will result in a big
inhibitory input to the LSO

190
00:11:29,345 --> 00:11:32,750
neuron on that left
side of the brain.

191
00:11:32,750 --> 00:11:35,570
Sure, some of the sound
is going to come over here

192
00:11:35,570 --> 00:11:38,460
and activate the left pathway.

193
00:11:38,460 --> 00:11:41,360
But that excitation
won't be as strong.

194
00:11:41,360 --> 00:11:43,925
So in that case, the
inhibition will rule.

195
00:11:45,350 --> 00:11:49,000
The LSO neuron gets a
large amount of inhibition

196
00:11:49,000 --> 00:11:51,130
and its firing rate will be low.

197
00:11:51,130 --> 00:11:53,490
Its response will be low
because it's inhibited.

198
00:11:56,000 --> 00:11:59,885
Now, what happens if the sound
is right in the middle, 0 ILD?

199
00:11:59,885 --> 00:12:02,280
Sound is coming straight ahead.

200
00:12:02,280 --> 00:12:04,510
The sound is the
same at the two ears.

201
00:12:04,510 --> 00:12:07,720
And thus, the ILD is 0.

202
00:12:07,720 --> 00:12:11,950
Well, it sort of depends
on how you wire this up.

203
00:12:13,490 --> 00:12:16,750
Whether the balance
is perfectly balanced.

204
00:12:16,750 --> 00:12:19,409
Then, maybe it would
be a 50% response.

205
00:12:19,409 --> 00:12:21,700
In the case of this graph,
it looks like the inhibition

206
00:12:21,700 --> 00:12:26,490
is a little bit stronger for
equal sound on the two sides.

207
00:12:26,490 --> 00:12:30,740
And so the inhibition
dominates at 0 ILD.

208
00:12:30,740 --> 00:12:34,470
But in actuality, if you
record from the LSO neurons,

209
00:12:34,470 --> 00:12:37,630
you find all sorts
of combinations.

210
00:12:37,630 --> 00:12:40,540
Those that have 50%
response rate at 0,

211
00:12:40,540 --> 00:12:44,100
those that have 90% response,
and those that have 10%

212
00:12:44,100 --> 00:12:44,810
like this guy.

213
00:12:46,800 --> 00:12:48,735
So this is a very
simple circuit.

214
00:12:50,740 --> 00:12:53,770
No coincidence detection.

215
00:12:53,770 --> 00:12:55,010
No delay lines.

216
00:12:56,390 --> 00:12:58,280
Well, you should
kind of do a heads

217
00:12:58,280 --> 00:13:01,940
up here when I talk
about delay and timing

218
00:13:01,940 --> 00:13:06,280
because this pathway coming
from this contralateral side

219
00:13:06,280 --> 00:13:07,065
is a lot longer.

220
00:13:08,950 --> 00:13:11,670
The axons have to
cross the midline.

221
00:13:11,670 --> 00:13:14,620
And then there's a
delay here because

222
00:13:14,620 --> 00:13:18,860
at the synapse between
one axon's terminal

223
00:13:18,860 --> 00:13:21,740
and the cell bodies
of the MNTB neurons,

224
00:13:21,740 --> 00:13:23,320
there's a little bit of delay.

225
00:13:23,320 --> 00:13:25,790
This neurotransmitter
has to be released.

226
00:13:27,310 --> 00:13:30,245
The MNTB neurons have to get
excited and, finally, fire.

227
00:13:31,480 --> 00:13:35,350
So all that can take a
half a millisecond or so.

228
00:13:35,350 --> 00:13:40,770
It turns out this axon
is a very thick one.

229
00:13:40,770 --> 00:13:42,790
This cochlear
nucleus neuron here

230
00:13:42,790 --> 00:13:46,590
that provides this crossing axon
is called the globular bushy

231
00:13:46,590 --> 00:13:47,740
cell.

232
00:13:47,740 --> 00:13:49,825
And we talked about
that a little bit

233
00:13:49,825 --> 00:13:51,575
when we talked about
the cochlear nucleus.

234
00:13:53,380 --> 00:13:55,590
It's not important
exactly what type it is,

235
00:13:55,590 --> 00:13:58,480
but this has the
thickest axon, really,

236
00:13:58,480 --> 00:14:01,290
of all the axons in
the auditory pathway.

237
00:14:01,290 --> 00:14:04,515
So it gets across the
midline very quickly.

238
00:14:06,030 --> 00:14:07,840
And there is a
synaptic delay here.

239
00:14:07,840 --> 00:14:10,920
So the contralateral input is
going to come in a little bit

240
00:14:10,920 --> 00:14:12,190
later.

241
00:14:12,190 --> 00:14:15,480
So sometimes in recording
some LSO neurons,

242
00:14:15,480 --> 00:14:20,140
you find a little bit of
excitation from this side.

243
00:14:20,140 --> 00:14:22,175
And then right after,
an instant later,

244
00:14:22,175 --> 00:14:25,060
a half a millisecond
or a millisecond later,

245
00:14:25,060 --> 00:14:26,330
you find a big inhibition.

246
00:14:27,670 --> 00:14:28,700
And so that can happen.

247
00:14:28,700 --> 00:14:32,070
The sum total though,
in the case where

248
00:14:32,070 --> 00:14:35,090
the sound is off to
the contralateral side,

249
00:14:35,090 --> 00:14:36,450
is a dominant inhibition.

250
00:14:41,460 --> 00:14:44,000
Now, we haven't talked
about neural inhibition

251
00:14:44,000 --> 00:14:48,520
much in our class, so maybe
we should just mention it.

252
00:14:48,520 --> 00:14:50,400
And I should definitely
mention the type

253
00:14:50,400 --> 00:14:53,000
of inhibitory transmitter
that's used here.

254
00:14:54,095 --> 00:15:15,910
So we have the MNTB neurons
are coming like this

255
00:15:15,910 --> 00:15:19,130
and sending their axons
to the LSO neurons.

256
00:15:21,750 --> 00:15:23,570
And they're inhibiting them.

257
00:15:23,570 --> 00:15:25,780
That's what I mean by
this minus sign here.

258
00:15:26,880 --> 00:15:31,230
And so this is an
inhibitory synapse

259
00:15:31,230 --> 00:15:33,275
which inhibits the LSO neurons.

260
00:15:34,800 --> 00:15:36,560
One can ask the
question is, what

261
00:15:36,560 --> 00:15:38,730
is the inhibitory
neurotransmitter?

262
00:15:38,730 --> 00:15:41,410
And in this case,
it's called glycine.

263
00:15:45,890 --> 00:15:52,210
So glycine is released from
the MNTB neuron terminals

264
00:15:52,210 --> 00:15:54,810
onto the LSO neurons.

265
00:15:54,810 --> 00:15:57,630
And so how is that known?

266
00:15:58,890 --> 00:16:03,080
Well, it's sort of a lot of
little pieces in a puzzle here.

267
00:16:03,080 --> 00:16:09,320
The MNTB neurons themselves have
the metabolic machinery that

268
00:16:09,320 --> 00:16:11,270
can make the glycine.

269
00:16:11,270 --> 00:16:13,270
They transport it down
their nerve axons.

270
00:16:14,570 --> 00:16:17,290
Glycine is actually a
fairly common chemical

271
00:16:17,290 --> 00:16:20,610
in all cells of the body,
but these MNTB neurons

272
00:16:20,610 --> 00:16:22,320
are packed with it.

273
00:16:22,320 --> 00:16:26,080
If you use antibodies
to glycine,

274
00:16:26,080 --> 00:16:29,760
they stain these
neurons much darker

275
00:16:29,760 --> 00:16:31,807
than most neurons in the brain.

276
00:16:31,807 --> 00:16:33,765
There are other glycinergic
neurons, of course.

277
00:16:35,430 --> 00:16:38,055
Their axons and their
terminals are darkly stained.

278
00:16:39,870 --> 00:16:42,720
The lateral superior
olive neurons

279
00:16:42,720 --> 00:16:45,530
have glycine receptors on them.

280
00:16:46,950 --> 00:16:50,400
When you put little
puffs of glycine,

281
00:16:50,400 --> 00:16:52,540
which you can from a pipette.

282
00:16:52,540 --> 00:16:55,590
You can release
glycine from a pipette

283
00:16:55,590 --> 00:16:58,810
in an artificial
recording situation.

284
00:16:58,810 --> 00:17:01,580
When you puff glycine
on to LSO neurons,

285
00:17:01,580 --> 00:17:03,140
they are inhibited greatly.

286
00:17:03,140 --> 00:17:05,610
So they certainly have
the receptors for glycine.

287
00:17:07,270 --> 00:17:11,910
There are uptake systems to
take up the glycine after it's

288
00:17:11,910 --> 00:17:14,455
been released so that the
inhibition doesn't stay on

289
00:17:14,455 --> 00:17:14,955
forever.

290
00:17:16,880 --> 00:17:21,569
When you stimulate, if you go
in and stimulate electrically

291
00:17:21,569 --> 00:17:26,560
these MNTB neurons, you find
the LSO neurons are inhibited.

292
00:17:26,560 --> 00:17:31,170
So a lot of little pieces of
the puzzle go into the idea

293
00:17:31,170 --> 00:17:32,700
that these are
glycinergic neurons.

294
00:17:34,270 --> 00:17:37,160
And this glycinergic
input is very

295
00:17:37,160 --> 00:17:38,566
important in ILD sensitivity.

296
00:17:40,120 --> 00:17:46,740
Now, a lot of textbooks will say
the ITD sensitivity is created

297
00:17:46,740 --> 00:17:49,780
in the MSO and the ILD
sensitivity is created here

298
00:17:49,780 --> 00:17:51,040
in the LSO.

299
00:17:51,040 --> 00:17:52,810
And we're done with it.

300
00:17:52,810 --> 00:17:55,330
But that's probably not true.

301
00:17:55,330 --> 00:17:58,210
This is such a simple
circuit, you probably

302
00:17:58,210 --> 00:18:01,450
have other places in the
auditory pathway where

303
00:18:01,450 --> 00:18:03,595
ILD sensitivity is also created.

304
00:18:04,950 --> 00:18:10,785
For example, you have nuclei
in the lateral lemniscus--

305
00:18:10,785 --> 00:18:14,070
the pathway going up to
the inferior colliculus.

306
00:18:14,070 --> 00:18:15,810
You have such circuits
probably right

307
00:18:15,810 --> 00:18:18,149
in the inferior
colliculus and maybe

308
00:18:18,149 --> 00:18:19,440
at other levels of the pathway.

309
00:18:20,450 --> 00:18:23,240
So this is not the
only place where

310
00:18:23,240 --> 00:18:25,280
we find circuits
for ILD sensitivity.

311
00:18:32,060 --> 00:18:38,440
Now, we talked about last
time how these ILD cues

312
00:18:38,440 --> 00:18:41,970
were prominent at
high frequencies

313
00:18:41,970 --> 00:18:47,710
and very-- almost nonexistent
at low frequencies.

314
00:18:47,710 --> 00:18:50,740
Because low-frequency sound
can bend around the head

315
00:18:50,740 --> 00:18:52,120
very easily.

316
00:18:52,120 --> 00:18:54,170
So we had, I think,
for 200 Hertz.

317
00:18:55,230 --> 00:18:59,260
Even for a sound source located
directly off to the side

318
00:18:59,260 --> 00:19:00,465
absolutely 0 ILD.

319
00:19:02,260 --> 00:19:05,510
For 6,000 Hertz,
we have a huge ILD.

320
00:19:06,900 --> 00:19:10,220
So ILDs are not very
important at low frequencies.

321
00:19:10,220 --> 00:19:16,530
If you go in to the LSO
and record the frequency

322
00:19:16,530 --> 00:19:21,440
responsivity of the neurons
there from measurements

323
00:19:21,440 --> 00:19:22,760
of their tuning curve.

324
00:19:22,760 --> 00:19:25,510
Now, remember what
a tuning curve was?

325
00:19:25,510 --> 00:19:28,590
Tuning curve, we had
sort of over and over.

326
00:19:28,590 --> 00:19:31,875
It was a plot of
sound frequency.

327
00:19:33,194 --> 00:19:36,960
On the y-axis is sound
pressure level for a response.

328
00:19:39,780 --> 00:19:42,520
And we have the
v-shaped functions

329
00:19:42,520 --> 00:19:45,290
and we picked off the
most sensitive frequency.

330
00:19:45,290 --> 00:19:48,710
And that's the CF of the neuron.

331
00:19:48,710 --> 00:19:51,420
Neurons in the superior olivary
complex have beautiful CFs.

332
00:19:52,440 --> 00:19:54,270
And you could do a
CF mapping, which

333
00:19:54,270 --> 00:19:55,820
is what's done in this study.

334
00:19:56,970 --> 00:20:02,550
So in the LSO-- right here, you
find CFs from 0 to 1 kilohertz.

335
00:20:05,430 --> 00:20:08,380
Right in this part
from 4 to 10 kilohertz.

336
00:20:10,730 --> 00:20:15,030
In this part here,
20 kilohertz and up.

337
00:20:15,030 --> 00:20:20,600
And this is the typical kind
of funny s-shape of the LSO

338
00:20:20,600 --> 00:20:25,360
that you see in coronal
sections in this case of the cat

339
00:20:25,360 --> 00:20:26,530
superior olivary complex.

340
00:20:29,799 --> 00:20:32,090
If you were to do this mapping
in the cochlear nucleus,

341
00:20:32,090 --> 00:20:36,040
you'd find a lot of
cochlear nucleus devoted

342
00:20:36,040 --> 00:20:40,670
to the low CFs, some to the
middle, and some to high.

343
00:20:40,670 --> 00:20:45,540
In the LSO, you have a
lot of the LSO devoted

344
00:20:45,540 --> 00:20:50,765
to the high CFs, which is where
ILD cues are very prominent.

345
00:20:52,000 --> 00:20:55,610
So it makes sense that where
you're processing ILDs,

346
00:20:55,610 --> 00:21:00,560
you devote a lot of neurons to
responding to the frequencies

347
00:21:00,560 --> 00:21:02,310
where the cue is very salient.

348
00:21:04,810 --> 00:21:07,540
The MNTB, which
projects into the LSO,

349
00:21:07,540 --> 00:21:11,090
has a similar disproportionately
large representation

350
00:21:11,090 --> 00:21:11,830
of high CFs.

351
00:21:15,390 --> 00:21:17,250
The MSO is just the opposite.

352
00:21:18,430 --> 00:21:22,620
There's hardly any MSO
devoted to the very highest

353
00:21:22,620 --> 00:21:23,370
frequencies.

354
00:21:23,370 --> 00:21:25,740
And remember, last
time we talked

355
00:21:25,740 --> 00:21:30,180
about ITDs being ambiguous
at high frequencies

356
00:21:30,180 --> 00:21:34,970
because the interaural time
difference is still the same,

357
00:21:34,970 --> 00:21:39,050
but the sound can go through
one or even more complete

358
00:21:39,050 --> 00:21:42,110
cycles by the time it
gets to the other ear.

359
00:21:42,110 --> 00:21:46,430
And so you can't
tell what ITD you're

360
00:21:46,430 --> 00:21:48,055
working with at these
high frequencies.

361
00:21:49,280 --> 00:21:52,240
On the other hand,
where these ILD cues

362
00:21:52,240 --> 00:21:54,460
were weak at low
frequencies, the ITDs

363
00:21:54,460 --> 00:21:56,550
are strong and salient.

364
00:21:56,550 --> 00:22:02,130
And there's a lot of MSO devoted
to the low characteristic

365
00:22:02,130 --> 00:22:04,265
frequencies where the
ITDs are prominent.

366
00:22:07,400 --> 00:22:10,350
So that's what this
text means, that there's

367
00:22:10,350 --> 00:22:14,390
sort of a disproportionate
amount of territory

368
00:22:14,390 --> 00:22:18,450
in these nuclei based
on where the cue is

369
00:22:18,450 --> 00:22:19,900
important in the
frequency domain.

370
00:22:22,870 --> 00:22:26,390
Now, we have a little bit of
an issue here with the LSO.

371
00:22:28,440 --> 00:22:35,900
This LSO on this side is going
to respond to sound sources

372
00:22:35,900 --> 00:22:37,865
on this side of the body.

373
00:22:39,060 --> 00:22:41,450
And remember in most
sensory systems,

374
00:22:41,450 --> 00:22:45,530
there's a crossing
such that stimuli

375
00:22:45,530 --> 00:22:50,540
on the right side of the
body evoke neural responses

376
00:22:50,540 --> 00:22:52,465
on the left side of the brain.

377
00:22:52,465 --> 00:22:54,590
This LSO has sort
of got it backwards.

378
00:22:54,590 --> 00:22:59,070
It's responding with excitation
to sound sources on the right.

379
00:22:59,070 --> 00:23:01,890
Well, that's taken
care of by virtue

380
00:23:01,890 --> 00:23:04,700
of the ascending
projections from the LSO

381
00:23:04,700 --> 00:23:07,350
to the next higher
center, which is

382
00:23:07,350 --> 00:23:09,220
the Inferior Colliculus, or IC.

383
00:23:11,780 --> 00:23:14,395
And that's diagrammed
here in this next slide.

384
00:23:20,430 --> 00:23:23,650
And this shows the
projections of the LSO.

385
00:23:25,190 --> 00:23:31,200
Again, the LSO on the left side
projecting across the midline

386
00:23:31,200 --> 00:23:34,210
to the inferior colliculus
on the right side.

387
00:23:35,900 --> 00:23:38,010
And that projection
would then predict

388
00:23:38,010 --> 00:23:42,630
that if a sound source was
over on this right side

389
00:23:42,630 --> 00:23:48,640
and exciting the LSO neuron,
that message would then

390
00:23:48,640 --> 00:23:53,010
get in a big way to the inferior
colliculus on the left side.

391
00:23:53,010 --> 00:23:54,840
So that inferior
colliculus would then

392
00:23:54,840 --> 00:23:57,120
respond to sounds on the
right side of the body.

393
00:23:58,930 --> 00:24:01,660
For a while, this field
was a little mystified

394
00:24:01,660 --> 00:24:03,100
because there's
also a projection

395
00:24:03,100 --> 00:24:05,980
from the left LSO
to the left IC.

396
00:24:07,180 --> 00:24:09,435
It ended up being an
inhibitory projection.

397
00:24:11,660 --> 00:24:14,120
So this projection here
that stays on the same side

398
00:24:14,120 --> 00:24:15,155
is mostly inhibitory.

399
00:24:16,420 --> 00:24:21,420
It's not exactly clear what
that does, but it's there.

400
00:24:21,420 --> 00:24:25,955
You can sort of discount it in
terms of the mapping of stimuli

401
00:24:25,955 --> 00:24:28,850
on one side of the
body to responses

402
00:24:28,850 --> 00:24:30,125
on the other side of the body.

403
00:24:32,920 --> 00:24:35,234
The MSO doesn't
have such a problem,

404
00:24:35,234 --> 00:24:36,275
I'll just say in passing.

405
00:24:37,442 --> 00:24:40,870
The MSO, just because
of its ITD map--

406
00:24:40,870 --> 00:24:44,910
if you go back and review
last week's lecture,

407
00:24:44,910 --> 00:24:46,890
you'll see that
the MSO is already

408
00:24:46,890 --> 00:24:50,440
mapping ITDs for sound
sources on the opposite side

409
00:24:50,440 --> 00:24:50,960
of the body.

410
00:24:53,400 --> 00:24:57,670
Now, there's been a lot of work,
especially in the early days

411
00:24:57,670 --> 00:25:00,890
of auditory neural
science, on looking

412
00:25:00,890 --> 00:25:01,953
at the effect of lesions.

413
00:25:08,220 --> 00:25:11,440
And lesions are a
little bit hard to do

414
00:25:11,440 --> 00:25:15,130
in a complex as hard as the
superior olivary complex.

415
00:25:15,130 --> 00:25:18,620
Because if you go in and
try to destroy the LSO,

416
00:25:18,620 --> 00:25:23,390
invariably right next door
is the MSO on one side.

417
00:25:23,390 --> 00:25:26,320
And right next door is the MNTB.

418
00:25:26,320 --> 00:25:30,750
It's very hard to make selective
lesions in the superior olive.

419
00:25:30,750 --> 00:25:32,670
When you get to the
inferior colliculus-- oh,

420
00:25:32,670 --> 00:25:34,400
there's a big nucleus.

421
00:25:34,400 --> 00:25:37,210
And you can go in and
you can destroy it

422
00:25:37,210 --> 00:25:40,190
in an experimental
animal on just one side.

423
00:25:41,950 --> 00:25:47,490
When that is done, and the
animal is trained in a task

424
00:25:47,490 --> 00:25:50,910
to localize where the
sound is coming from,

425
00:25:50,910 --> 00:25:54,880
it's very clear that a lesion
of the inferior colliculus

426
00:25:54,880 --> 00:26:00,290
on the right side makes the
animal unable to localize

427
00:26:00,290 --> 00:26:02,540
sounds on the opposite
side of the body.

428
00:26:04,000 --> 00:26:08,300
So lesion in the right
inferior colliculus,

429
00:26:08,300 --> 00:26:11,460
the animal can't localize
sounds on the opposite side.

430
00:26:13,160 --> 00:26:17,340
The animal still can localize
sounds on the ipsilateral hemi

431
00:26:17,340 --> 00:26:23,450
field, in the side that
you didn't lesion on

432
00:26:23,450 --> 00:26:27,070
because the other colliculus
that's still there that's

433
00:26:27,070 --> 00:26:29,150
not lesioned can
perform the task.

434
00:26:29,150 --> 00:26:33,190
If you lesion both
inferior colliculi,

435
00:26:33,190 --> 00:26:35,540
the animal can't
localize sounds anywhere.

436
00:26:36,810 --> 00:26:38,470
But it's very
clear that a lesion

437
00:26:38,470 --> 00:26:42,940
on one side of the
auditory pathway

438
00:26:42,940 --> 00:26:46,100
here makes the animal
unable to localize sounds

439
00:26:46,100 --> 00:26:47,550
in the opposite hemi field.

440
00:26:47,550 --> 00:26:50,660
So that's a theme for
sensory processing

441
00:26:50,660 --> 00:26:53,450
that stimuli on one
side of the body

442
00:26:53,450 --> 00:26:58,674
are mapped to neural nuclei on
the opposite side of the brain.

443
00:26:58,674 --> 00:27:00,590
And that's very clear
in the auditory pathway.

444
00:27:05,450 --> 00:27:12,140
Now, I think last time when we
had the demonstrations of ITDs

445
00:27:12,140 --> 00:27:16,480
and ILDs in headphones where
we could present one and not

446
00:27:16,480 --> 00:27:20,680
the other, someone said,
well, it sounded a little bit

447
00:27:20,680 --> 00:27:24,670
like when we had an ITD when
the sound was on the left side.

448
00:27:24,670 --> 00:27:28,210
And then later, we had an ILD
with sound on the left side.

449
00:27:28,210 --> 00:27:31,140
They sounded a little bit
the same, as if the sound

450
00:27:31,140 --> 00:27:34,090
could be put on one side
by either ITD and ILD.

451
00:27:34,090 --> 00:27:38,620
And that was an
interesting comment

452
00:27:38,620 --> 00:27:43,520
because of the phenomenon called
time intensity trading, which

453
00:27:43,520 --> 00:27:44,940
you can do with headphones.

454
00:27:44,940 --> 00:27:49,110
I won't demonstrate it because
it's so clear to explain it.

455
00:27:49,110 --> 00:27:52,180
You can have someone
listen in headphones

456
00:27:52,180 --> 00:27:54,800
and you can make-- let's see.

457
00:27:54,800 --> 00:27:59,560
An ITD such that when
that's presented alone,

458
00:27:59,560 --> 00:28:03,180
the sound is perceived as if
it's coming from the left side.

459
00:28:05,280 --> 00:28:09,460
You can also present with those
headphones the same sound,

460
00:28:09,460 --> 00:28:12,020
but make now in
this case an ILD.

461
00:28:13,290 --> 00:28:16,880
So the sound is higher
on the right side

462
00:28:16,880 --> 00:28:19,350
so that it sounds like it's
coming from the right side.

463
00:28:20,510 --> 00:28:22,730
So now we have the
time cues making

464
00:28:22,730 --> 00:28:24,590
you think it's
coming from the left.

465
00:28:24,590 --> 00:28:27,020
The intensity cues
making you think

466
00:28:27,020 --> 00:28:28,970
it's coming from the right.

467
00:28:28,970 --> 00:28:31,510
When you put those
two things together,

468
00:28:31,510 --> 00:28:35,730
which you can do
artificially with headphones,

469
00:28:35,730 --> 00:28:38,410
you find in some
cases that the sound

470
00:28:38,410 --> 00:28:40,480
sounds like it's coming
from straight ahead.

471
00:28:42,530 --> 00:28:46,800
And this is called trading
for time and intensity,

472
00:28:46,800 --> 00:28:49,090
or time/intensity trading.

473
00:28:49,090 --> 00:28:51,840
And you can balance
one with the other.

474
00:28:51,840 --> 00:28:54,512
If you do a lot of
sound level on one side

475
00:28:54,512 --> 00:28:56,220
and just a little time
on the other side,

476
00:28:56,220 --> 00:28:58,420
it sounds like it's
off to the right.

477
00:28:59,570 --> 00:29:02,360
If you do them equal, it sounds
like it's straight ahead.

478
00:29:02,360 --> 00:29:06,360
If you do a lot of time, a
big ITD and a little ILD,

479
00:29:06,360 --> 00:29:08,880
it sounds like it's a
little off to the left.

480
00:29:08,880 --> 00:29:12,220
It's a very clear
psychophysical phenomena.

481
00:29:12,220 --> 00:29:16,700
Where do you find neural
responses that correlate with

482
00:29:16,700 --> 00:29:18,560
time/ intensity trading?

483
00:29:18,560 --> 00:29:23,640
Well, you find it in places like
the inferior colliculus, where

484
00:29:23,640 --> 00:29:27,815
inputs from the LSO and
the MSO first converge.

485
00:29:29,540 --> 00:29:34,420
So both LSO and MSO project up
into the inferior colliculus.

486
00:29:35,480 --> 00:29:39,010
And here, you have
the kind of responses

487
00:29:39,010 --> 00:29:40,250
that are shown on this graph.

488
00:29:42,740 --> 00:29:48,300
This x-axis is now a time
axis, except it's ITD.

489
00:29:50,390 --> 00:29:54,270
One side is the inhibitory lead.

490
00:29:56,780 --> 00:30:03,610
The sound is delayed so that the
ITD is from the opposite ear.

491
00:30:03,610 --> 00:30:06,650
This is the side
so that the time

492
00:30:06,650 --> 00:30:08,950
is delayed from the
ear on the same side

493
00:30:08,950 --> 00:30:12,070
as the inferior colliculus
you're recording from.

494
00:30:12,070 --> 00:30:16,820
And clearly, this neuron
is sensitive to that ITD

495
00:30:16,820 --> 00:30:18,720
if you vary it.

496
00:30:18,720 --> 00:30:21,310
This is the percent
neural response.

497
00:30:21,310 --> 00:30:25,710
100% would be up here, 0
response would be up here.

498
00:30:25,710 --> 00:30:29,590
Now, what's varied
as the parameter here

499
00:30:29,590 --> 00:30:31,483
is the interaural
level difference.

500
00:30:32,760 --> 00:30:35,110
So in this case, the
contralateral ear

501
00:30:35,110 --> 00:30:38,390
is-- it looks like 35.

502
00:30:38,390 --> 00:30:40,273
And the ipsilateral ear is 45.

503
00:30:41,510 --> 00:30:45,490
They've kept the contralateral
ear the same for the most part.

504
00:30:45,490 --> 00:30:48,170
And the ipsilateral
ear level has changed.

505
00:30:48,170 --> 00:30:51,840
And clearly, changing
the level also

506
00:30:51,840 --> 00:30:54,780
has a big effect on
the neural response.

507
00:30:54,780 --> 00:30:58,130
So here, for the first time
in the auditory pathway, where

508
00:30:58,130 --> 00:31:04,490
we're finding ITD and
ILD responses together

509
00:31:04,490 --> 00:31:07,590
in single neurons--
in a big way.

510
00:31:07,590 --> 00:31:11,100
You find them in a small
way in the MSO and the LSO.

511
00:31:11,100 --> 00:31:14,100
But this is a huge effect here.

512
00:31:14,100 --> 00:31:17,600
And probably, there
is the first place

513
00:31:17,600 --> 00:31:21,900
where you might have neural
correlates of your perception

514
00:31:21,900 --> 00:31:24,100
for time/intensity trading.

515
00:31:28,070 --> 00:31:32,550
So that's one phenomenon I
want to cover that probably

516
00:31:32,550 --> 00:31:35,420
has a neural
correlate at the level

517
00:31:35,420 --> 00:31:37,030
of the inferior colliculus.

518
00:31:37,030 --> 00:31:38,200
And here's another one.

519
00:31:39,900 --> 00:31:44,180
We've been dealing with
very simple stimuli that

520
00:31:44,180 --> 00:31:47,770
have just, say,
one ITD or one ILD.

521
00:31:49,660 --> 00:31:53,570
When you get into a room, it
becomes much more complicated

522
00:31:53,570 --> 00:31:58,250
because of echoes off
the walls and ceiling

523
00:31:58,250 --> 00:32:00,110
and floor of the room.

524
00:32:00,110 --> 00:32:02,750
And there's some very
interesting experiments

525
00:32:02,750 --> 00:32:06,990
that are done with more natural
stimuli like you find in rooms.

526
00:32:06,990 --> 00:32:08,940
And we'll go over
just a few of them.

527
00:32:08,940 --> 00:32:13,930
So here-- and this
is an observer

528
00:32:13,930 --> 00:32:17,220
listening to a sound source
off to the observer's left.

529
00:32:18,240 --> 00:32:20,900
And the direct sound
is indicated by A here.

530
00:32:22,850 --> 00:32:28,820
And most of these data deal with
the interaural time difference

531
00:32:28,820 --> 00:32:30,870
because this is a
fairly low frequency.

532
00:32:32,060 --> 00:32:35,390
This interaural time difference
of the direct sound indicated

533
00:32:35,390 --> 00:32:40,090
by this big arrow in the air
here favors the left ear.

534
00:32:40,090 --> 00:32:44,380
It's going to be arriving
at the left ear first

535
00:32:44,380 --> 00:32:46,420
and the right ear
a little bit later.

536
00:32:46,420 --> 00:32:50,310
So if you just had that
sound, the subject,

537
00:32:50,310 --> 00:32:52,340
obviously, would
report that the sound

538
00:32:52,340 --> 00:32:53,705
is coming from the left side.

539
00:32:55,060 --> 00:33:04,260
And this is a plot of the sound
for the left and right ear

540
00:33:04,260 --> 00:33:08,590
if you just had this
arrow in A here.

541
00:33:08,590 --> 00:33:12,630
And in green, it's the
interaural time difference

542
00:33:12,630 --> 00:33:14,590
for just that direct sound.

543
00:33:14,590 --> 00:33:19,080
So you can see it's
a fairly stable ITD

544
00:33:19,080 --> 00:33:22,330
and it's a negative
ITD just by convention.

545
00:33:23,430 --> 00:33:26,400
Left ear is going to
be negative here, ITD.

546
00:33:26,400 --> 00:33:28,680
And this is an ITD
in milliseconds

547
00:33:28,680 --> 00:33:30,215
of 0.4 milliseconds.

548
00:33:31,500 --> 00:33:34,220
And remember, we said if it was
directly off to the left side,

549
00:33:34,220 --> 00:33:37,247
it would be about 0.6.

550
00:33:37,247 --> 00:33:38,955
If it were straight
ahead, it would be 0.

551
00:33:40,120 --> 00:33:42,480
So what's the sound
stimulus here?

552
00:33:42,480 --> 00:33:45,720
Well, this is, again,
changing things a little bit.

553
00:33:45,720 --> 00:33:48,650
We've been talking about very
simple pure tone stimuli,

554
00:33:48,650 --> 00:33:49,185
or clicks.

555
00:33:51,080 --> 00:33:54,805
This paper is from
a German group.

556
00:33:56,090 --> 00:34:02,380
And French or German speakers
have this lovely speech sound,

557
00:34:02,380 --> 00:34:08,199
which is called like a
trill or rolled R. And I,

558
00:34:08,199 --> 00:34:11,120
for the life of me,
cannot do this stimulus.

559
00:34:11,120 --> 00:34:12,020
But I'll try.

560
00:34:12,020 --> 00:34:14,260
It's something like [ROLLING R].

561
00:34:14,260 --> 00:34:19,790
As if you were to pronounce
the word in German "reich."

562
00:34:19,790 --> 00:34:21,540
Or in French, [INAUDIBLE].

563
00:34:23,780 --> 00:34:26,469
It's impossible for me
to do that because I

564
00:34:26,469 --> 00:34:28,040
don't speak those languages.

565
00:34:28,040 --> 00:34:30,860
But anyway, here's
the sound stimulus

566
00:34:30,860 --> 00:34:32,929
with all the echoes added in.

567
00:34:34,850 --> 00:34:37,679
And one of these traits just
shows the left ear input

568
00:34:37,679 --> 00:34:38,694
and the right ear input.

569
00:34:39,780 --> 00:34:43,449
And these peaks here
are the trills of the R.

570
00:34:43,449 --> 00:34:47,169
You can see them of just the
left ear input as considered.

571
00:34:47,169 --> 00:34:50,620
So it's a trill
of the R. And you

572
00:34:50,620 --> 00:34:53,639
can see how many
milliseconds happened

573
00:34:53,639 --> 00:34:56,270
between each of those
parts of the trill.

574
00:34:56,270 --> 00:34:59,030
Maybe like 40
milliseconds or so.

575
00:35:01,740 --> 00:35:05,750
Now, when the subject
is in a normal room,

576
00:35:05,750 --> 00:35:10,060
a lot more happens than
just the direct sound.

577
00:35:10,060 --> 00:35:13,380
Some of the sound comes
from the source, the R here,

578
00:35:13,380 --> 00:35:16,880
and bounces off the wall
to the subject's right.

579
00:35:18,070 --> 00:35:20,700
Some of the sound goes
beyond the subject's head

580
00:35:20,700 --> 00:35:23,320
and bounces off the
wall to the right,

581
00:35:23,320 --> 00:35:26,880
and then the wall behind the
subject and comes back there.

582
00:35:28,730 --> 00:35:34,760
And this drawing
here is when you

583
00:35:34,760 --> 00:35:40,080
take into account several of
these reflections in addition

584
00:35:40,080 --> 00:35:41,115
to the direct sound.

585
00:35:43,020 --> 00:35:47,825
And the overall ITD is
still plotted in green.

586
00:35:49,530 --> 00:35:51,330
And look what a mess it is.

587
00:35:51,330 --> 00:35:53,120
The ITD is all over the place.

588
00:35:54,743 --> 00:35:59,440
It starts out like it should
from just the direct sound.

589
00:35:59,440 --> 00:36:00,930
It's negative here.

590
00:36:00,930 --> 00:36:03,140
But then real quickly,
it goes past 0.

591
00:36:03,140 --> 00:36:06,120
And it goes way up here
and it bounces around,

592
00:36:06,120 --> 00:36:07,485
then it goes back down again.

593
00:36:09,350 --> 00:36:10,670
It's all over the map.

594
00:36:12,130 --> 00:36:15,080
So where do you
think that subject

595
00:36:15,080 --> 00:36:17,140
says the sound is coming from?

596
00:36:18,430 --> 00:36:21,390
Do you think the subject
says, I can't tell,

597
00:36:21,390 --> 00:36:22,390
it's all over the place?

598
00:36:25,810 --> 00:36:28,120
Well, you've been
in rooms and you've

599
00:36:28,120 --> 00:36:30,760
listened to speakers in rooms.

600
00:36:30,760 --> 00:36:32,680
You're hearing me
right now and there's

601
00:36:32,680 --> 00:36:38,330
a lot of sound coming off the
walls to get into your ears

602
00:36:38,330 --> 00:36:39,890
on both sides.

603
00:36:39,890 --> 00:36:44,330
If you add up the
reflections from the walls

604
00:36:44,330 --> 00:36:47,390
to the side and beyond,
it's a lot greater

605
00:36:47,390 --> 00:36:50,790
than the direct sound in
terms of the total energy.

606
00:36:50,790 --> 00:36:53,100
But you can close
your eyes and you

607
00:36:53,100 --> 00:36:56,170
can know that I'm standing
up here to this side of you.

608
00:36:56,170 --> 00:36:59,670
You don't have any
trouble with that.

609
00:36:59,670 --> 00:37:03,430
And if you just do a
very careful experiment

610
00:37:03,430 --> 00:37:05,190
with this subject,
the direct sound,

611
00:37:05,190 --> 00:37:07,200
and a couple of
reflections like here.

612
00:37:07,200 --> 00:37:10,080
And you have the
subject with a pointer.

613
00:37:10,080 --> 00:37:12,650
That subject will say the
speaker is over there.

614
00:37:12,650 --> 00:37:13,980
It's on the left side.

615
00:37:13,980 --> 00:37:17,520
It's not behind or
to the wrong side.

616
00:37:18,750 --> 00:37:20,000
So how do we do that?

617
00:37:21,460 --> 00:37:25,510
Well, there is something
called the precedence effect,

618
00:37:25,510 --> 00:37:27,108
which is very important here.

619
00:37:29,740 --> 00:37:31,780
Which helps you in
a situation where

620
00:37:31,780 --> 00:37:33,085
you have lots of reflections.

621
00:37:36,420 --> 00:37:37,950
And what does precedence mean?

622
00:37:37,950 --> 00:37:40,555
Well, precedence means
something is dominating

623
00:37:40,555 --> 00:37:42,335
or something is
the most important.

624
00:37:44,840 --> 00:37:50,710
And in these cases, if
you look very carefully

625
00:37:50,710 --> 00:37:52,960
at the complex sound,
when you add up

626
00:37:52,960 --> 00:37:57,000
all the reflections
with the direct sound,

627
00:37:57,000 --> 00:37:59,900
you can see, if you have
really good eyesight,

628
00:37:59,900 --> 00:38:05,260
that right at the beginning of
the sort of burst of energy,

629
00:38:05,260 --> 00:38:08,910
the first thing to get
to the subject ears

630
00:38:08,910 --> 00:38:10,615
is coming from the left side.

631
00:38:12,310 --> 00:38:15,270
So the left ear input here
right at the beginning

632
00:38:15,270 --> 00:38:18,480
of these trills starts out.

633
00:38:20,010 --> 00:38:21,850
There's a bigger left ear input.

634
00:38:21,850 --> 00:38:25,820
And that's why you
have a negative ITD

635
00:38:25,820 --> 00:38:28,030
right at the
beginning of the trill

636
00:38:28,030 --> 00:38:29,375
before it starts going crazy.

637
00:38:31,250 --> 00:38:34,160
It turns out that the
most important part

638
00:38:34,160 --> 00:38:38,110
of this signal for localizing
where the sound comes from

639
00:38:38,110 --> 00:38:40,910
is in the initial
few milliseconds.

640
00:38:40,910 --> 00:38:44,750
So subjects bias
their impression

641
00:38:44,750 --> 00:38:46,990
of where the sound
is coming from right

642
00:38:46,990 --> 00:38:48,690
at the beginning of the sound.

643
00:38:48,690 --> 00:38:53,250
And they tend to suppress or
ignore all the remaining parts

644
00:38:53,250 --> 00:38:56,125
of the sound, at least
for localization purposes.

645
00:38:57,320 --> 00:39:02,350
So what takes precedence
here is the very initial part

646
00:39:02,350 --> 00:39:04,410
of the sound signal.

647
00:39:04,410 --> 00:39:06,420
And that's what the
precedence effect is.

648
00:39:07,610 --> 00:39:12,460
It's sometimes called by a
different name when you're

649
00:39:12,460 --> 00:39:15,720
dealing with speech--
the Haas effect.

650
00:39:18,950 --> 00:39:20,470
But the precedence
effect is more

651
00:39:20,470 --> 00:39:22,650
general for any type of sound.

652
00:39:22,650 --> 00:39:24,590
And it really means
that when you're

653
00:39:24,590 --> 00:39:28,290
in a complex environment with
all sorts of reflections,

654
00:39:28,290 --> 00:39:32,490
you pay attention to the
very first, or initial, sound

655
00:39:32,490 --> 00:39:33,590
and ignore the others.

656
00:39:34,770 --> 00:39:42,030
So the precedence effect
can be studied carefully

657
00:39:42,030 --> 00:39:47,930
by just narrowing it down to a
direct sound and one reflection

658
00:39:47,930 --> 00:39:50,120
by this kind of setup here.

659
00:39:50,120 --> 00:39:55,690
So here's a subject
listening to two speakers

660
00:39:55,690 --> 00:39:57,590
in an anechoic room.

661
00:39:57,590 --> 00:40:00,710
And I think we've talked
about anechoic rooms before.

662
00:40:00,710 --> 00:40:02,870
Anechoic rooms
have these baffles

663
00:40:02,870 --> 00:40:05,780
on the walls and the floor.

664
00:40:05,780 --> 00:40:10,090
This person is seated
in a chair in a mesh,

665
00:40:10,090 --> 00:40:12,240
so he doesn't fall
into the baffles

666
00:40:12,240 --> 00:40:15,690
below him, which are
also absorbing sound

667
00:40:15,690 --> 00:40:17,570
on the floor of the room.

668
00:40:17,570 --> 00:40:19,400
There are baffles in
the ceiling as well.

669
00:40:19,400 --> 00:40:21,410
So whatever sound
this is presented

670
00:40:21,410 --> 00:40:22,655
goes right to the subject.

671
00:40:23,800 --> 00:40:28,200
If it goes beyond him, it
goes to this wall over here

672
00:40:28,200 --> 00:40:30,160
and is completely absorbed.

673
00:40:30,160 --> 00:40:31,480
So there are no reflections.

674
00:40:31,480 --> 00:40:33,040
There's just the direct sound.

675
00:40:34,800 --> 00:40:37,970
So you can say, well, one of
these is the direct sound.

676
00:40:37,970 --> 00:40:39,940
And a little bit
later, I'm going

677
00:40:39,940 --> 00:40:41,740
to introduce a
second sound coming

678
00:40:41,740 --> 00:40:45,700
from somewhere else,
which is the echo.

679
00:40:46,740 --> 00:40:51,250
So this might simulate
a wall on this side

680
00:40:51,250 --> 00:40:53,320
of the subject that reflects.

681
00:40:53,320 --> 00:40:54,940
In this case, there's
no reflection.

682
00:40:54,940 --> 00:40:58,840
But you can say this is an
artificial echo presented

683
00:40:58,840 --> 00:41:00,735
by this second loudspeaker.

684
00:41:02,190 --> 00:41:06,216
What happens if we change
the timing between those two

685
00:41:06,216 --> 00:41:06,715
sounds?

686
00:41:07,900 --> 00:41:09,375
And that's what's plotted here.

687
00:41:10,940 --> 00:41:14,100
This is the delay
between the first sound

688
00:41:14,100 --> 00:41:15,730
and the second sound
in milliseconds.

689
00:41:17,110 --> 00:41:20,470
Now, for ITDS and
sound localization,

690
00:41:20,470 --> 00:41:27,030
remember we've been talking
about ITDs way down here from 0

691
00:41:27,030 --> 00:41:30,000
to about 0.6 milliseconds.

692
00:41:30,000 --> 00:41:35,510
That's where you have sound
localization way down here.

693
00:41:35,510 --> 00:41:40,570
And when you do a delay
that's that short,

694
00:41:40,570 --> 00:41:43,345
it sounds to a subject like
there's just one sound.

695
00:41:45,580 --> 00:41:48,900
If the ITD is at
exactly 0, this subject

696
00:41:48,900 --> 00:41:50,710
will perceive the
sound source being

697
00:41:50,710 --> 00:41:52,705
directly in between
these two speakers.

698
00:41:55,510 --> 00:41:59,660
As the lagging sound
gets greater and greater,

699
00:41:59,660 --> 00:42:04,540
the subject will start to
perceive that the source goes

700
00:42:04,540 --> 00:42:07,180
to the original or
first sound emitted.

701
00:42:09,930 --> 00:42:14,210
And when the delay comes
to the maximal delay

702
00:42:14,210 --> 00:42:16,040
for the size of the
human head, which

703
00:42:16,040 --> 00:42:19,080
is about 0.6
milliseconds right here,

704
00:42:19,080 --> 00:42:21,260
the subject will
perceive that it's only

705
00:42:21,260 --> 00:42:23,405
coming from that
initial loudspeaker.

706
00:42:25,780 --> 00:42:30,530
Then, as delays get
further and further on,

707
00:42:30,530 --> 00:42:33,090
up between about
0.6 milliseconds

708
00:42:33,090 --> 00:42:35,850
and a little over
10 milliseconds,

709
00:42:35,850 --> 00:42:39,490
the subject will still say,
I still hear one sound.

710
00:42:39,490 --> 00:42:41,860
And it's still coming
from that speaker.

711
00:42:41,860 --> 00:42:45,040
But now it's starting to
sound a little bit different.

712
00:42:45,040 --> 00:42:48,080
In fact, it doesn't
sound as dead anymore.

713
00:42:48,080 --> 00:42:51,540
It sounds like I'm in
a pretty live room.

714
00:42:51,540 --> 00:42:52,770
So what's a live room?

715
00:42:52,770 --> 00:42:57,015
A room that's reverberant,
like a church or a cathedral.

716
00:42:57,015 --> 00:42:58,345
It sounds roomy.

717
00:42:58,345 --> 00:42:59,595
It sounds like there's volume.

718
00:43:01,650 --> 00:43:06,130
Over here, this is the region
called the precedence effect,

719
00:43:06,130 --> 00:43:09,170
where you ignore
that lagging sound.

720
00:43:09,170 --> 00:43:11,420
And it still sounds like
there's just one sound,

721
00:43:11,420 --> 00:43:13,160
but it sounds different.

722
00:43:13,160 --> 00:43:17,002
This is the region of the
precedence effect from 0.6

723
00:43:17,002 --> 00:43:19,570
to maybe 10 milliseconds or so.

724
00:43:19,570 --> 00:43:23,570
Then, as the delay
becomes longer and longer,

725
00:43:23,570 --> 00:43:26,520
you start to
perceive two sounds.

726
00:43:26,520 --> 00:43:29,680
You hear, let's say, a
click from the first speaker

727
00:43:29,680 --> 00:43:33,760
and a second click a little bit
later from the second speaker.

728
00:43:33,760 --> 00:43:36,320
And now, the delay
is long enough

729
00:43:36,320 --> 00:43:37,870
so you actually hear an echo.

730
00:43:37,870 --> 00:43:41,210
You hear two sounds, an
initial sound and an echo.

731
00:43:43,310 --> 00:43:45,550
So that's the perception.

732
00:43:45,550 --> 00:43:48,195
And there's this big region here
called the precedence effect.

733
00:43:49,390 --> 00:43:51,640
So I have a
demonstration, if you

734
00:43:51,640 --> 00:43:54,520
don't believe me, about echoes.

735
00:43:54,520 --> 00:43:57,770
And the demonstration is
really vivid, I think.

736
00:43:59,260 --> 00:44:04,160
The echoes become more-- you
don't hear the echoes here

737
00:44:04,160 --> 00:44:06,520
because it's a
precedence effect.

738
00:44:06,520 --> 00:44:08,540
Maybe it sounds a
little bit more roomy,

739
00:44:08,540 --> 00:44:10,140
but you don't hear an echo.

740
00:44:10,140 --> 00:44:12,719
But there definitely
are echoes there.

741
00:44:12,719 --> 00:44:14,260
You may not be able
to perceive them.

742
00:44:15,440 --> 00:44:18,110
So let me play
this demonstration.

743
00:44:18,110 --> 00:44:20,480
The demonstration is--
I think, the best part

744
00:44:20,480 --> 00:44:23,590
of the demonstration is
someone taking a brick

745
00:44:23,590 --> 00:44:25,800
and hitting it with a hammer.

746
00:44:25,800 --> 00:44:28,160
And that makes a
big click, right?

747
00:44:28,160 --> 00:44:32,447
Well, after that big
sound, that impact sound,

748
00:44:32,447 --> 00:44:33,280
there's some echoes.

749
00:44:35,750 --> 00:44:39,100
You can't hear them
very well except when

750
00:44:39,100 --> 00:44:42,130
they do this neat trick on
this demonstration, which

751
00:44:42,130 --> 00:44:44,445
is to play the sound
recording backwards.

752
00:44:45,900 --> 00:44:48,540
And then, the
echoes start first,

753
00:44:48,540 --> 00:44:50,900
and then you hear the
hammer hitting the brick.

754
00:44:52,820 --> 00:44:55,120
They also have
some text on here.

755
00:44:55,120 --> 00:44:58,660
And they read the text, and then
they play the text backward.

756
00:44:58,660 --> 00:45:00,480
To me, that's not so obvious.

757
00:45:00,480 --> 00:45:03,080
So they do this demonstration
of hitting the brick.

758
00:45:03,080 --> 00:45:07,110
They do it in an anechoic
room right here first.

759
00:45:07,110 --> 00:45:07,900
There's no echoes.

760
00:45:09,140 --> 00:45:11,620
Second, they do it
in a normal room,

761
00:45:11,620 --> 00:45:15,170
like this room where
there's some reverberation.

762
00:45:15,170 --> 00:45:17,020
But a lot of the
reverberation is

763
00:45:17,020 --> 00:45:19,410
stopped by the
carpet on the floor

764
00:45:19,410 --> 00:45:21,670
and the clothes I'm wearing.

765
00:45:21,670 --> 00:45:25,470
And the seat cushions
absorb some of the echoes.

766
00:45:25,470 --> 00:45:27,890
Then finally, they do this
demonstration a third time.

767
00:45:27,890 --> 00:45:29,290
And they do that backwards, too.

768
00:45:29,290 --> 00:45:32,960
Then they do it a third time
in a very reverberant room,

769
00:45:32,960 --> 00:45:36,610
like a church or a cathedral,
where you hit the brick

770
00:45:36,610 --> 00:45:38,940
and it just sort of
rings for quite a ways.

771
00:45:40,350 --> 00:45:42,320
So let's play this
demonstration and see

772
00:45:42,320 --> 00:45:45,324
if it lives up to
my description.

773
00:45:45,324 --> 00:45:45,990
[AUDIO PLAYBACK]

774
00:45:45,990 --> 00:45:47,982
-[INAUDIBLE] echoes.

775
00:45:47,982 --> 00:45:50,970
First, in an anechoic room.

776
00:45:50,970 --> 00:45:52,962
Then, in a conference room.

777
00:45:52,962 --> 00:45:55,950
Then finally, in a
very reverberant space.

778
00:45:55,950 --> 00:45:59,436
You will hear a hammer
striking a brick followed

779
00:45:59,436 --> 00:46:02,424
by an old Scottish [INAUDIBLE].

780
00:46:02,424 --> 00:46:04,914
Playing these sounds
backwards focuses

781
00:46:04,914 --> 00:46:07,902
our attention on the
echoes that occur.

782
00:46:12,890 --> 00:46:16,005
From ghoulies and ghosties,
and long-leggedy beasties

783
00:46:16,005 --> 00:46:19,900
and things that go bump in the
night, good lord deliver us.

784
00:46:19,900 --> 00:46:21,742
PROFESSOR: OK, that's forward.

785
00:46:21,742 --> 00:46:22,450
This is backward.

786
00:46:24,414 --> 00:46:31,775
-[SPEAKING BACKWARD]

787
00:46:31,775 --> 00:46:33,400
PROFESSOR: OK, now
the conference room.

788
00:46:40,750 --> 00:46:44,670
-From ghoulies and ghosties,
and long-leggedy beasties

789
00:46:44,670 --> 00:46:48,590
and things that go bump in the
night, good lord deliver us.

790
00:46:53,000 --> 00:46:59,537
[SPEAKING BACKWARD]

791
00:46:59,537 --> 00:47:01,620
PROFESSOR: All right, now
in the reverberant room.

792
00:47:10,380 --> 00:47:14,104
-From ghoulies and ghosties,
and long-leggedy beasties

793
00:47:14,104 --> 00:47:17,527
and things that go bump in the
night, good lord deliver us.

794
00:47:23,395 --> 00:47:31,010
[SPEAKING BACKWARD]

795
00:47:31,010 --> 00:47:34,445
[END AUDIO PLAYBACK]

796
00:47:34,445 --> 00:47:39,040
PROFESSOR: All right, so I
like especially the sound

797
00:47:39,040 --> 00:47:42,370
of the hammer hitting the
brick played backwards

798
00:47:42,370 --> 00:47:46,070
in the reverberant room
because it's going to pssew.

799
00:47:46,070 --> 00:47:49,460
And all that pssew
leading up to the impact

800
00:47:49,460 --> 00:47:52,650
is the echo that you
just completely discount

801
00:47:52,650 --> 00:47:57,030
because of the precedence
effect in the normal hearing.

802
00:47:59,490 --> 00:48:04,805
OK, so that brings me
up to the special part

803
00:48:04,805 --> 00:48:06,305
of my lecture, which
is the reading.

804
00:48:07,660 --> 00:48:12,480
And this little quotation
is by a musician.

805
00:48:13,660 --> 00:48:17,540
And of course, musicians
love reverberant rooms,

806
00:48:17,540 --> 00:48:21,140
like churches or cathedrals,
or concert halls, or whatever.

807
00:48:21,140 --> 00:48:25,730
So E. Power Biggs, who was
the organist at Harvard

808
00:48:25,730 --> 00:48:28,480
for a long time, made
many famous recordings

809
00:48:28,480 --> 00:48:31,800
of organ music said,
"An organist will

810
00:48:31,800 --> 00:48:35,190
take all the reverberation
time he is given,

811
00:48:35,190 --> 00:48:40,220
and then ask for a bit more,
for ample reverberation is

812
00:48:40,220 --> 00:48:42,640
part of organ music itself.

813
00:48:42,640 --> 00:48:45,780
Many of Bach's organ works
are designed actually

814
00:48:45,780 --> 00:48:47,143
to exploit reverberation.

815
00:48:48,520 --> 00:48:52,150
Consider the pause that follows
the ornamented proclamation

816
00:48:52,150 --> 00:48:56,260
that opens the famous 'Toccata
in D Minor.' Obviously,

817
00:48:56,260 --> 00:48:59,970
this is for the enjoyment of the
notes as they remain suspended

818
00:48:59,970 --> 00:49:02,710
in the air."

819
00:49:02,710 --> 00:49:06,286
So musicians love
reverberations.

820
00:49:07,410 --> 00:49:13,620
And that's the reason that
halls where we appreciate music,

821
00:49:13,620 --> 00:49:17,980
like Boston Symphony Hall--
and they have some measurements

822
00:49:17,980 --> 00:49:22,300
of reverberation time
here for opera houses.

823
00:49:22,300 --> 00:49:25,256
1.3 seconds.

824
00:49:26,260 --> 00:49:30,410
OK, that's the time the
echoes take to decay.

825
00:49:31,700 --> 00:49:34,625
Symphony Hall in
Birmingham, 2.4 seconds.

826
00:49:35,760 --> 00:49:42,090
St. Paul's Cathedral in London,
13 seconds reverberation time.

827
00:49:42,090 --> 00:49:46,110
Now in contrast, when you
have theaters for speech,

828
00:49:46,110 --> 00:49:48,580
like Shakespeare
drama theaters, you

829
00:49:48,580 --> 00:49:50,480
don't want all those
reverberations.

830
00:49:50,480 --> 00:49:53,360
You want, for example,
a theater for speech

831
00:49:53,360 --> 00:49:57,470
here is quoted at having
a reverb time of 0.9

832
00:49:57,470 --> 00:50:00,300
seconds because you don't
want all these echoes

833
00:50:00,300 --> 00:50:04,135
to interfere with your
interpretation of the speech.

834
00:50:06,370 --> 00:50:10,800
And average living
room, 0.4 seconds.

835
00:50:10,800 --> 00:50:13,830
Reverberation time is given
for the great outdoors.

836
00:50:13,830 --> 00:50:14,668
Anybody guess?

837
00:50:17,108 --> 00:50:18,506
0.0.

838
00:50:18,506 --> 00:50:19,960
All right.

839
00:50:19,960 --> 00:50:23,185
The outdoors has no
walls, ceiling, or floor.

840
00:50:25,770 --> 00:50:27,650
Now, why are we talking
about this here?

841
00:50:27,650 --> 00:50:31,840
Because in the
inferior colliculus,

842
00:50:31,840 --> 00:50:36,163
you find some neurons that
show precedence-like responses.

843
00:50:37,700 --> 00:50:42,475
And so here is a recording from
an inferior colliculus neuron.

844
00:50:43,520 --> 00:50:47,400
And this was a study where
they used infants and adults,

845
00:50:47,400 --> 00:50:50,405
but these are just the
data for the adult.

846
00:50:51,580 --> 00:50:55,500
And the stimuli are two sounds.

847
00:50:55,500 --> 00:50:56,560
I think they are clicks.

848
00:50:57,950 --> 00:51:01,532
The first sound starts
at time 10 milliseconds.

849
00:51:01,532 --> 00:51:03,680
So this is the time axis.

850
00:51:03,680 --> 00:51:05,490
And this is the
dot raster display.

851
00:51:05,490 --> 00:51:07,320
So each little
dot here-- they're

852
00:51:07,320 --> 00:51:09,710
a little bit hard to
see-- is a neural spike.

853
00:51:10,880 --> 00:51:12,870
And there are many, many trials.

854
00:51:12,870 --> 00:51:15,120
OK, perhaps 50 or 100 trials.

855
00:51:15,120 --> 00:51:18,670
And you can see reliably on
all the trials, the neuron

856
00:51:18,670 --> 00:51:20,260
responded to the first stimulus.

857
00:51:22,090 --> 00:51:24,010
When the second
stimulus occurred

858
00:51:24,010 --> 00:51:28,370
at 101 millisecond
delayed, the neuron

859
00:51:28,370 --> 00:51:31,050
also faithfully responded
to the second stimulus.

860
00:51:32,500 --> 00:51:36,100
But as the delay was
shortened less and less,

861
00:51:36,100 --> 00:51:40,130
the neuron eventually stopped
responding for a delay

862
00:51:40,130 --> 00:51:42,840
that's-- in this case,
it looks like about 20

863
00:51:42,840 --> 00:51:43,903
or so milliseconds.

864
00:51:48,680 --> 00:51:50,930
Well, this is certainly
precedence-like behavior.

865
00:51:52,270 --> 00:51:56,655
That the neural response to the
second stimulus is attenuated.

866
00:51:58,280 --> 00:52:00,630
This delay, where
this particular neuron

867
00:52:00,630 --> 00:52:03,770
starts cutting out,
is not exactly where

868
00:52:03,770 --> 00:52:07,447
we stop losing the
precedence effect in humans.

869
00:52:07,447 --> 00:52:08,405
This is from an animal.

870
00:52:09,580 --> 00:52:14,380
And this animal is anesthetized,
so many of the processes

871
00:52:14,380 --> 00:52:16,920
are slowed down by anesthesia.

872
00:52:16,920 --> 00:52:19,360
And perhaps the animal
is a little cool,

873
00:52:19,360 --> 00:52:22,600
which might make these
things abnormally long.

874
00:52:22,600 --> 00:52:27,060
But certainly, this kind of
precedence-like responses

875
00:52:27,060 --> 00:52:30,270
are on the way toward
explaining the precedence effect

876
00:52:30,270 --> 00:52:32,207
at the level of the
inferior colliculus.

877
00:52:35,340 --> 00:52:37,570
And this is the work
of Ruth Litovsky

878
00:52:37,570 --> 00:52:39,675
from University of Wisconsin.

879
00:52:42,650 --> 00:52:46,300
OK, so now I want to
shift gears a little bit

880
00:52:46,300 --> 00:52:53,830
and go onto a different
species, which is the barn owl.

881
00:52:53,830 --> 00:52:56,880
And this is mostly the
work of-- originally,

882
00:52:56,880 --> 00:52:58,345
Mark Konishi at Caltech.

883
00:52:59,480 --> 00:53:03,645
And now, Eric Knudsen at
Stanford in California.

884
00:53:05,150 --> 00:53:08,430
And why did they choose
to study the barn owl?

885
00:53:08,430 --> 00:53:12,350
So many of us study
mammalian models

886
00:53:12,350 --> 00:53:16,930
because we want to know what
human hearing is all about.

887
00:53:16,930 --> 00:53:21,650
And it's hard to record
from human brains,

888
00:53:21,650 --> 00:53:24,890
but we want to choose an
animal that's like the human.

889
00:53:24,890 --> 00:53:25,780
So we use a mammal.

890
00:53:25,780 --> 00:53:28,030
Why did they choose
the barn owl?

891
00:53:29,130 --> 00:53:31,570
A lot of successes
in neuroscience

892
00:53:31,570 --> 00:53:33,410
have been choosing
an animal that's

893
00:53:33,410 --> 00:53:35,560
specialized for a certain task.

894
00:53:35,560 --> 00:53:37,905
And barn owls are very
specialized for hearing.

895
00:53:39,310 --> 00:53:44,730
So if you take a barn
owl and blindfold the owl

896
00:53:44,730 --> 00:53:49,200
and turn a mouse
loose in a room.

897
00:53:49,200 --> 00:53:51,840
As long as the mouse is
making a little bit of sound,

898
00:53:51,840 --> 00:53:54,370
the barn owl can quickly
fly over to the mouse

899
00:53:54,370 --> 00:53:56,350
and catch it and eat it.

900
00:53:56,350 --> 00:54:01,290
So the prey of barn owls
are mice and insects,

901
00:54:01,290 --> 00:54:04,465
like grasshoppers that
are down on the floor.

902
00:54:05,490 --> 00:54:09,850
And if you've ever seen or
watched a barn owl hunting,

903
00:54:09,850 --> 00:54:12,850
you can clearly
see them at night.

904
00:54:12,850 --> 00:54:14,710
They don't come
out during the day.

905
00:54:14,710 --> 00:54:16,080
But at night they come out.

906
00:54:16,080 --> 00:54:21,340
I watched one a lot of evenings
when I was in California.

907
00:54:21,340 --> 00:54:24,140
The owl would come and
sit on a basketball hoop.

908
00:54:24,140 --> 00:54:26,050
And it would just perch there.

909
00:54:26,050 --> 00:54:28,095
And it would move
its head all around.

910
00:54:28,919 --> 00:54:31,210
And I didn't really know what
it was doing at the time,

911
00:54:31,210 --> 00:54:36,530
until later, when I read some
of this work and it said that

912
00:54:36,530 --> 00:54:39,890
owls' eyes cannot
move in their head.

913
00:54:39,890 --> 00:54:41,210
The owl's eyes are fixed.

914
00:54:42,660 --> 00:54:46,950
So this beautiful control
of the eyeball position

915
00:54:46,950 --> 00:54:50,175
that we have in
mammalian eye control

916
00:54:50,175 --> 00:54:52,670
is not present in most birds.

917
00:54:52,670 --> 00:54:54,965
And certainly is not
present in the owl.

918
00:54:54,965 --> 00:54:58,000
So to move eyes, you
have to move the head.

919
00:54:58,000 --> 00:55:01,280
And of course, you're
moving the ears as well.

920
00:55:01,280 --> 00:55:03,450
In the barn owl,
and in other words,

921
00:55:03,450 --> 00:55:06,990
you don't have an external
pinna, which we said

922
00:55:06,990 --> 00:55:10,120
introduces a lot of help
to localizing sounds.

923
00:55:10,120 --> 00:55:12,200
But the barn owl
is a unique bird

924
00:55:12,200 --> 00:55:16,900
in that there is some
external stuff going on,

925
00:55:16,900 --> 00:55:19,630
which is called
the facial rough.

926
00:55:19,630 --> 00:55:25,240
And that's formed by feathers
on the face of the owl.

927
00:55:25,240 --> 00:55:29,250
You can see they're sort of
like fan-shaped down here.

928
00:55:29,250 --> 00:55:34,350
And over here, they
go over the ear canal.

929
00:55:34,350 --> 00:55:36,215
There's an ear canal, certainly.

930
00:55:36,215 --> 00:55:41,490
They go over the ear canal
in this perioral flap.

931
00:55:41,490 --> 00:55:44,420
And the opening for sound
to get in is below that.

932
00:55:44,420 --> 00:55:46,170
And also, above it
there's an opening.

933
00:55:47,229 --> 00:55:48,770
So the barn owl
doesn't have a pinna,

934
00:55:48,770 --> 00:55:51,360
but it has some
modified feathers

935
00:55:51,360 --> 00:55:52,475
on the front of its face.

936
00:55:54,320 --> 00:55:56,955
Barn owl feathers are
also interesting that they

937
00:55:56,955 --> 00:55:57,720
are modified.

938
00:55:57,720 --> 00:56:00,430
The feathers on the
wings are modified.

939
00:56:00,430 --> 00:56:03,100
When this owl took off, when I
watched it on that basketball

940
00:56:03,100 --> 00:56:05,890
hoop, every now and then it
would take off and go down

941
00:56:05,890 --> 00:56:08,120
to the basketball court below.

942
00:56:08,120 --> 00:56:09,960
And you couldn't hear anything.

943
00:56:09,960 --> 00:56:14,430
Owl feathers are
specifically designed so even

944
00:56:14,430 --> 00:56:19,270
the air going over them
when the owl flaps its wings

945
00:56:19,270 --> 00:56:23,780
is completely silent because the
owls don't want whatever it's

946
00:56:23,780 --> 00:56:26,700
hunting to hear
them approaching it.

947
00:56:26,700 --> 00:56:28,930
So barn owl wing
feathers are specifically

948
00:56:28,930 --> 00:56:31,250
designed to be
acoustically silent.

949
00:56:35,150 --> 00:56:39,590
OK, so a lot of work has been
done on the barn owl pathway.

950
00:56:39,590 --> 00:56:41,630
Now, this is a
little bit different

951
00:56:41,630 --> 00:56:47,640
because if you've looked at
the brains of avian species,

952
00:56:47,640 --> 00:56:49,050
they're a little bit different.

953
00:56:49,050 --> 00:56:50,765
They evolve differently
than mammals.

954
00:56:51,780 --> 00:56:54,040
But they have some
analogous nuclei.

955
00:56:55,830 --> 00:56:58,100
Here are the cochlear nuclei.

956
00:56:58,100 --> 00:57:00,350
That's with this
little text here.

957
00:57:00,350 --> 00:57:03,030
This is supposed to be the
owl pathway on the left side

958
00:57:03,030 --> 00:57:04,430
and the right side of the brain.

959
00:57:04,430 --> 00:57:07,260
This is the midline
in dashed lines.

960
00:57:07,260 --> 00:57:12,940
The owl cochlear nuclei are
split up into two parts.

961
00:57:12,940 --> 00:57:15,740
And the one that's featured
here is labeled NM.

962
00:57:18,700 --> 00:57:27,790
So NM stands for
Nucleus Magnocellularis.

963
00:57:36,240 --> 00:57:39,030
And so we can all figure
out what this means.

964
00:57:39,030 --> 00:57:42,850
Cellularlaris means the
cell or the nerve cells.

965
00:57:42,850 --> 00:57:44,640
Magno means big.

966
00:57:44,640 --> 00:57:46,580
These are the big cells.

967
00:57:46,580 --> 00:57:50,050
So there's some other parts
of the bird cochlear nuclei

968
00:57:50,050 --> 00:57:51,620
where the cells are smaller.

969
00:57:51,620 --> 00:57:54,955
But this is the
big-- big cell part.

970
00:57:56,180 --> 00:57:58,930
And in nucleus
magnocellularis, you

971
00:57:58,930 --> 00:58:01,405
have beautiful phase locking.

972
00:58:03,450 --> 00:58:06,350
We talked about
that being typical

973
00:58:06,350 --> 00:58:08,770
of the mammalian pathway.

974
00:58:08,770 --> 00:58:11,120
In the cochlear
nucleus, the bushy cells

975
00:58:11,120 --> 00:58:13,260
have good phase locking.

976
00:58:13,260 --> 00:58:15,310
Maybe even better than
the auditory nerve.

977
00:58:16,930 --> 00:58:20,960
In the owl, the
nucleus magnocellularis

978
00:58:20,960 --> 00:58:23,140
neurons have excellent
phase locking.

979
00:58:23,140 --> 00:58:26,430
And so they're keeping track
of the stimulus waveform.

980
00:58:26,430 --> 00:58:28,210
The timing is important to them.

981
00:58:30,070 --> 00:58:32,420
They project centrally.

982
00:58:35,360 --> 00:58:37,960
The one on the left side and
the one on the right side

983
00:58:37,960 --> 00:58:42,330
converge onto a nucleus
that's sensitive to interaural

984
00:58:42,330 --> 00:58:43,256
time differences.

985
00:58:44,480 --> 00:58:47,250
That's the avian
equivalent of the MSO.

986
00:58:49,930 --> 00:59:02,031
And it's called NL,
Nucleus Laminaris.

987
00:59:05,260 --> 00:59:06,910
OK, lamina means sheet.

988
00:59:07,816 --> 00:59:08,815
OK, and this is a sheet.

989
00:59:10,194 --> 00:59:14,160
It looks like that
anyway in the anatomy.

990
00:59:14,160 --> 00:59:16,840
And there, the neurons
are sensitive to ITDs.

991
00:59:18,160 --> 00:59:21,065
And there's a beautiful
Jeffress model there.

992
00:59:22,570 --> 00:59:25,870
Most of the papers
on the mammalian MSO

993
00:59:25,870 --> 00:59:28,500
say, we know there's
a beautiful Jeffress

994
00:59:28,500 --> 00:59:32,960
model in the avian
nucleus laminaris.

995
00:59:32,960 --> 00:59:35,540
But in the mammal, we're
starting to rethink it.

996
00:59:36,820 --> 00:59:38,650
This is a beautiful
Jeffress model

997
00:59:38,650 --> 00:59:46,416
where you find
neural responses that

998
00:59:46,416 --> 00:59:50,010
are very strongly peaked to ITD.

999
00:59:50,010 --> 00:59:51,840
So they fire for a certain ITD.

1000
00:59:53,100 --> 00:59:54,360
This is the firing rate.

1001
00:59:58,260 --> 01:00:01,150
But don't fire much
at all to other ITDs.

1002
01:00:01,150 --> 01:00:03,260
They're strongly tuned to ITD.

1003
01:00:04,830 --> 01:00:09,620
The nucleus laminaris in turn
projects across the midline

1004
01:00:09,620 --> 01:00:12,376
here to the inferior colliculus.

1005
01:00:14,230 --> 01:00:17,120
And we haven't talked
about it, but there

1006
01:00:17,120 --> 01:00:20,250
are several subregions of
the inferior colliculus.

1007
01:00:21,610 --> 01:00:23,160
The big one is called the ICC.

1008
01:00:24,530 --> 01:00:28,600
And that's called the
Inferior Colliculus Central

1009
01:00:28,600 --> 01:00:30,840
part, or central nucleus.

1010
01:00:30,840 --> 01:00:32,250
That's true in mammals as well.

1011
01:00:32,250 --> 01:00:34,180
That's the big part.

1012
01:00:34,180 --> 01:00:35,445
Some would call it the core.

1013
01:00:36,700 --> 01:00:39,440
That's what's indicated
here, the core

1014
01:00:39,440 --> 01:00:40,565
of the inferior colliculus.

1015
01:00:42,720 --> 01:00:46,790
And it, in turn,
projects to other places,

1016
01:00:46,790 --> 01:00:48,155
like the lateral part.

1017
01:00:50,310 --> 01:00:56,230
And it finally projects from
the lateral part to the ICX.

1018
01:01:00,910 --> 01:01:04,800
And the ICX is the
Inferior Colliculus.

1019
01:01:06,050 --> 01:01:07,823
And X stands for External.

1020
01:01:11,170 --> 01:01:13,875
So the external part of
the inferior colliculus.

1021
01:01:15,100 --> 01:01:18,450
And there is where some
very interesting responses

1022
01:01:18,450 --> 01:01:20,147
take place in the barn owl.

1023
01:01:21,520 --> 01:01:23,500
And we'll look at those
responses right now.

1024
01:01:30,760 --> 01:01:33,270
So these experiments
were first done

1025
01:01:33,270 --> 01:01:43,640
by Mark Konishi in, I
believe, the 1970s at Caltech.

1026
01:01:43,640 --> 01:01:45,590
And the experimental
subject is seen.

1027
01:01:47,060 --> 01:01:48,740
Here is the barn owl right here.

1028
01:01:50,280 --> 01:01:51,430
And where are his wings?

1029
01:01:51,430 --> 01:01:54,040
Well, his wings are folded down.

1030
01:01:54,040 --> 01:01:58,830
And he's in a little, sort of
like a tube sock, if you will.

1031
01:01:58,830 --> 01:01:59,940
His wings are constrained.

1032
01:01:59,940 --> 01:02:05,350
And the tube sock is mounted
on a pole or a pedestal.

1033
01:02:05,350 --> 01:02:06,710
And he's just sitting there.

1034
01:02:06,710 --> 01:02:07,740
He can move his head.

1035
01:02:09,320 --> 01:02:11,090
Actually, in many of
these experiments,

1036
01:02:11,090 --> 01:02:12,692
the head is clamped.

1037
01:02:12,692 --> 01:02:13,650
But he's sitting there.

1038
01:02:13,650 --> 01:02:14,150
He's awake.

1039
01:02:14,917 --> 01:02:16,125
He's listening to the sounds.

1040
01:02:18,300 --> 01:02:20,375
And the sound is
presented by a speaker.

1041
01:02:21,860 --> 01:02:23,465
The speaker is on a big hoop.

1042
01:02:24,740 --> 01:02:25,810
You see that hoop?

1043
01:02:27,040 --> 01:02:29,320
And the speaker is
sitting on the hoop.

1044
01:02:29,320 --> 01:02:31,900
And there's a little
motor in the speaker.

1045
01:02:31,900 --> 01:02:35,540
And the speaker can be moved
by driving the motor over here.

1046
01:02:36,634 --> 01:02:39,070
Or you can move the motor
the other direction.

1047
01:02:39,070 --> 01:02:42,970
You can move the speaker over
here or down here or up here,

1048
01:02:42,970 --> 01:02:45,820
wherever you want to on
that hoop the speaker can

1049
01:02:45,820 --> 01:02:46,930
be driven by the motor.

1050
01:02:47,990 --> 01:02:53,120
And because the hoop is mounted
on two posts on the side,

1051
01:02:53,120 --> 01:02:57,000
the whole hoop can be swung
up or it can be swung down.

1052
01:02:58,580 --> 01:03:03,380
OK, so you can put that speaker
anywhere this way and anywhere

1053
01:03:03,380 --> 01:03:05,370
up or down that you want to.

1054
01:03:05,370 --> 01:03:08,840
And so you can put that speaker
in the entire-- any position

1055
01:03:08,840 --> 01:03:14,080
you want to in the entire
frontal hemi field of the owl.

1056
01:03:14,080 --> 01:03:16,120
And I suppose they
didn't do that.

1057
01:03:16,120 --> 01:03:20,570
I suppose you could put it in
the rear hemi field as well.

1058
01:03:20,570 --> 01:03:25,260
But these data are just from the
frontal hemi field of the owl.

1059
01:03:27,550 --> 01:03:33,230
And since the owl's head is
mounted and is not moving,

1060
01:03:33,230 --> 01:03:35,350
you can apply a little
local anesthetic

1061
01:03:35,350 --> 01:03:36,990
and open up the skull.

1062
01:03:36,990 --> 01:03:42,440
And you can advance an electrode
into the ICX, the External

1063
01:03:42,440 --> 01:03:46,380
Nucleus of the
Inferior Colliculus,

1064
01:03:46,380 --> 01:03:48,900
and make recordings from
single neurons there.

1065
01:03:51,170 --> 01:03:54,120
And in this case, the
recordings are made from the ICX

1066
01:03:54,120 --> 01:03:55,050
on the right side.

1067
01:03:58,800 --> 01:04:02,200
And what's found
in this response

1068
01:04:02,200 --> 01:04:07,980
plot here is these neurons have
restricted receptive fields

1069
01:04:07,980 --> 01:04:08,480
in space.

1070
01:04:10,090 --> 01:04:13,960
So what's plotted here is
the neuro-responsive field

1071
01:04:13,960 --> 01:04:17,940
in dashed lines and
the most vigorous part

1072
01:04:17,940 --> 01:04:19,920
in the shaded area there.

1073
01:04:21,180 --> 01:04:24,750
And this axis, the
x-axis, is azimuth.

1074
01:04:26,202 --> 01:04:28,740
And the y-axis is elevation.

1075
01:04:31,560 --> 01:04:36,850
And this is a fairly restrictive
part of the whole hemi field.

1076
01:04:36,850 --> 01:04:40,810
And one of these-- this is one
response area for one neuron.

1077
01:04:41,910 --> 01:04:45,410
And a whole bunch
of these rectangles

1078
01:04:45,410 --> 01:04:48,810
are plotted down here with
a little diagram of the owl

1079
01:04:48,810 --> 01:04:50,955
right in the center
of this globe.

1080
01:04:53,420 --> 01:04:57,130
So there's 1, 2, 3,
4, 5, 6, 7, 8, 9,

1081
01:04:57,130 --> 01:05:02,090
10-- about a dozen neural
receptive fields for ICX

1082
01:05:02,090 --> 01:05:04,150
neurons in that owl.

1083
01:05:05,240 --> 01:05:08,710
Notice that they're
all fairly discrete.

1084
01:05:08,710 --> 01:05:11,520
That is, it only
responds when the speaker

1085
01:05:11,520 --> 01:05:17,630
is in a certain part of the
frontal field of the owl.

1086
01:05:19,790 --> 01:05:22,360
Notice also that we're
recording from the right side.

1087
01:05:23,390 --> 01:05:29,630
And most of the receptive fields
are off to the left of the owl.

1088
01:05:29,630 --> 01:05:31,850
They're on the opposite side.

1089
01:05:31,850 --> 01:05:35,770
So clearly, sound
stimuli on one side

1090
01:05:35,770 --> 01:05:39,880
are mapped to the
opposite brain side.

1091
01:05:39,880 --> 01:05:42,630
That's not true for
these three, but they're

1092
01:05:42,630 --> 01:05:43,660
close to the midline.

1093
01:05:45,930 --> 01:05:49,930
Most of the receptive fields
are not straight ahead.

1094
01:05:49,930 --> 01:05:51,890
They're actually
down below the owl.

1095
01:05:52,900 --> 01:05:55,427
There's one or two that are
just a little bit above,

1096
01:05:55,427 --> 01:05:57,010
but there are none
that are way above.

1097
01:05:57,010 --> 01:05:58,710
Most of them are
down below the owl.

1098
01:05:59,720 --> 01:06:03,360
Remember, the owl is sitting
on the basketball hoop

1099
01:06:03,360 --> 01:06:05,990
and he's listening for
targets down below.

1100
01:06:07,240 --> 01:06:11,030
This makes a lot of sense that
most of the receptive fields

1101
01:06:11,030 --> 01:06:15,790
are down below the owl, not
at the same level of the owl.

1102
01:06:15,790 --> 01:06:18,750
It doesn't care about
a mouse making a slam

1103
01:06:18,750 --> 01:06:20,270
dunk at the other hoop.

1104
01:06:20,270 --> 01:06:25,320
It cares about the mouse
down on the basketball court.

1105
01:06:28,540 --> 01:06:31,590
And finally, maybe the most
important and most interesting

1106
01:06:31,590 --> 01:06:36,980
part of these responses
is the progression

1107
01:06:36,980 --> 01:06:38,870
of where the
receptive fields are

1108
01:06:38,870 --> 01:06:43,010
in space versus where they
are along the dimensions

1109
01:06:43,010 --> 01:06:43,760
of the ICX.

1110
01:06:44,990 --> 01:06:50,820
So that this receptive field was
located over here in the ICX.

1111
01:06:51,940 --> 01:06:55,000
And as you move this way and
encountered a different neuron,

1112
01:06:55,000 --> 01:06:57,330
it was located over here
and its receptive field

1113
01:06:57,330 --> 01:07:00,230
was moved over
this way a little.

1114
01:07:00,230 --> 01:07:02,850
As you move further
in that direction,

1115
01:07:02,850 --> 01:07:05,680
you encountered another
receptive field even further

1116
01:07:05,680 --> 01:07:07,200
in this dimension.

1117
01:07:07,200 --> 01:07:10,930
And then finally, way
over here laterally,

1118
01:07:10,930 --> 01:07:13,180
you encountered these
receptive fields

1119
01:07:13,180 --> 01:07:14,505
that were way off to the side.

1120
01:07:15,780 --> 01:07:20,460
So along this dimension
of auditory space,

1121
01:07:20,460 --> 01:07:23,260
receptive fields were
found along this dimension

1122
01:07:23,260 --> 01:07:24,920
of the ICX.

1123
01:07:24,920 --> 01:07:27,740
There was also a
mapping going this way.

1124
01:07:29,220 --> 01:07:33,120
This is clearly then what
some people call a space map.

1125
01:07:33,120 --> 01:07:37,410
A mapping of auditory
space into position

1126
01:07:37,410 --> 01:07:41,830
within a certain part, the
ICX, of the owl's brain.

1127
01:07:43,160 --> 01:07:45,040
OK, it's a beautiful mapping.

1128
01:07:45,040 --> 01:07:47,750
Neuroscientists love mappings.

1129
01:07:47,750 --> 01:07:51,480
You've heard me talk
about tonotopic mappings

1130
01:07:51,480 --> 01:07:52,590
out the wazoo.

1131
01:07:52,590 --> 01:07:55,790
We love tonotopic mappings
because CF is then

1132
01:07:55,790 --> 01:07:56,430
very important.

1133
01:07:58,520 --> 01:08:03,134
This clearly is important
to have organization

1134
01:08:03,134 --> 01:08:05,300
of the receptive fields on
a dimension in the brain.

1135
01:08:06,470 --> 01:08:11,530
People have spent many,
many years looking

1136
01:08:11,530 --> 01:08:16,960
for these kind of mappings in
the mammalian auditory pathway

1137
01:08:16,960 --> 01:08:18,529
and not found them.

1138
01:08:20,109 --> 01:08:22,800
People have looked in
the inferior colliculus,

1139
01:08:22,800 --> 01:08:26,600
in the analogous part,
the central part,

1140
01:08:26,600 --> 01:08:28,180
the external part.

1141
01:08:28,180 --> 01:08:31,660
It is hard to record
from the external part

1142
01:08:31,660 --> 01:08:33,930
of the mammalian
inferior colliculus

1143
01:08:33,930 --> 01:08:35,899
because it's a small
edge structure,

1144
01:08:35,899 --> 01:08:37,900
but it has not been found.

1145
01:08:38,930 --> 01:08:42,460
People have looked in
the medial geniculate,

1146
01:08:42,460 --> 01:08:47,100
in the auditory cortex,
and looked for organization

1147
01:08:47,100 --> 01:08:51,520
of spatial receptive
fields and not found them.

1148
01:08:51,520 --> 01:08:56,170
So on nuclei in the main
part of the auditory pathway,

1149
01:08:56,170 --> 01:09:00,149
you do not find space maps
in the mammalian system.

1150
01:09:00,149 --> 01:09:04,139
So the one place you
find spatial organization

1151
01:09:04,139 --> 01:09:11,100
in the mammal is in the
mammalian superior colliculus.

1152
01:09:19,688 --> 01:09:20,979
And you're probably going, huh?

1153
01:09:20,979 --> 01:09:23,990
I thought you just said the
superior colliculus is visual?

1154
01:09:25,010 --> 01:09:25,700
Well, it is.

1155
01:09:25,700 --> 01:09:29,609
But it's a layered structure
that, if I'm not mistaken,

1156
01:09:29,609 --> 01:09:32,040
the top three layers
are exclusively visual.

1157
01:09:32,040 --> 01:09:35,520
But if you go down to lower
layers, the bottom layers

1158
01:09:35,520 --> 01:09:38,670
of the superior colliculus,
you start to encounter neurons

1159
01:09:38,670 --> 01:09:42,710
that respond to visual as
well as auditory stimuli.

1160
01:09:43,910 --> 01:09:47,730
And you may have talked
about the visual mapping

1161
01:09:47,730 --> 01:09:48,855
of the superior colliculus.

1162
01:09:50,160 --> 01:09:52,840
And those neurons
in the deep layers

1163
01:09:52,840 --> 01:09:56,740
that are also responsive
to auditory stimuli--

1164
01:09:56,740 --> 01:09:59,270
they're mostly ILD sensitive.

1165
01:09:59,270 --> 01:10:03,770
They're mapped in line with
the visual receptive fields.

1166
01:10:03,770 --> 01:10:07,860
They're also space mapped
to a certain extent.

1167
01:10:07,860 --> 01:10:11,730
Now, that nucleus clearly
is not on the main drag

1168
01:10:11,730 --> 01:10:13,110
of the auditory pathway.

1169
01:10:13,110 --> 01:10:16,910
The auditory pathway is cochlear
nucleus, superior olive,

1170
01:10:16,910 --> 01:10:20,950
inferior colliculus, medial
geniculate and cortex.

1171
01:10:20,950 --> 01:10:26,740
So you do have a space map
in the mammalian deep layers

1172
01:10:26,740 --> 01:10:28,350
of the superior
colliculus, but not

1173
01:10:28,350 --> 01:10:32,600
on the main parts of the
mammalian auditory pathway.

1174
01:10:32,600 --> 01:10:34,210
So that's been the finding.

1175
01:10:34,210 --> 01:10:36,700
Now, that's negative evidence.

1176
01:10:36,700 --> 01:10:39,620
It's not clear that I
won't be back here teaching

1177
01:10:39,620 --> 01:10:41,660
the course next year.

1178
01:10:41,660 --> 01:10:43,245
And we'll read a
paper that says,

1179
01:10:43,245 --> 01:10:45,210
ah, space map in wherever.

1180
01:10:45,210 --> 01:10:48,770
But it hasn't been found so
far with this one exception.

1181
01:10:51,650 --> 01:10:58,010
Now, the paper that we
read for today's class

1182
01:10:58,010 --> 01:11:04,040
talks about-- back
to the barn owl.

1183
01:11:05,810 --> 01:11:08,580
A place in the barn
owl, which is-- they

1184
01:11:08,580 --> 01:11:11,300
call it the optic
tectum in the birds.

1185
01:11:11,300 --> 01:11:15,000
But it's analogous to the
superior colliculus in mammals.

1186
01:11:15,000 --> 01:11:20,930
A place where, as I said, you
find auditory spacing maps that

1187
01:11:20,930 --> 01:11:23,375
are in line with
visual space maps.

1188
01:11:24,450 --> 01:11:29,470
And they do a very interesting
and elegant experiment

1189
01:11:29,470 --> 01:11:31,445
where one of those
maps is distorted.

1190
01:11:33,220 --> 01:11:37,390
And you study the resulting
effect on the other map.

1191
01:11:37,390 --> 01:11:38,810
OK, so how did they do that?

1192
01:11:44,380 --> 01:11:48,400
So this is n owl, but
it's a juvenile owl.

1193
01:11:48,400 --> 01:11:49,430
An owl chick.

1194
01:11:51,320 --> 01:11:55,460
And it's wearing some
interesting things on its eyes.

1195
01:11:55,460 --> 01:11:56,470
Those aren't its eyes.

1196
01:11:56,470 --> 01:11:59,820
Those are some prisms
that the investigators

1197
01:11:59,820 --> 01:12:03,530
have put on the owl's eyes.

1198
01:12:03,530 --> 01:12:07,580
And they deflect the visual
field a certain amount

1199
01:12:07,580 --> 01:12:09,520
depending on the
size of the prism.

1200
01:12:10,620 --> 01:12:17,130
And I can't remember
what the deflection was.

1201
01:12:17,130 --> 01:12:18,770
I seem to remember 30 degrees.

1202
01:12:20,730 --> 01:12:25,960
So the visual field is
deflected 30 degrees.

1203
01:12:25,960 --> 01:12:30,800
And as I said before, the owl's
eyes are fixed in the head.

1204
01:12:30,800 --> 01:12:33,530
So putting on these
goggles, no matter

1205
01:12:33,530 --> 01:12:37,540
what-- the goggles are going to
move if the owl moves its head.

1206
01:12:37,540 --> 01:12:41,370
So no matter what the owl does,
the visual receptive fields

1207
01:12:41,370 --> 01:12:43,420
of all these neurons,
everything in vision

1208
01:12:43,420 --> 01:12:44,905
is shifted 30 degrees.

1209
01:12:46,740 --> 01:12:47,945
This is normal.

1210
01:12:49,700 --> 01:12:52,151
These might be receptive fields
from neurons in the brain

1211
01:12:52,151 --> 01:12:52,650
somewhere.

1212
01:12:53,720 --> 01:12:55,450
This is when the
prisms are added.

1213
01:12:56,560 --> 01:13:01,470
Here, you've shifted the
visual receptive field.

1214
01:13:01,470 --> 01:13:03,010
The auditory
receptive field-- you

1215
01:13:03,010 --> 01:13:05,400
haven't changed the ears at all.

1216
01:13:05,400 --> 01:13:07,225
The auditory receptive
field is the same.

1217
01:13:08,820 --> 01:13:13,920
What's found when you do
that in juvenile owls,

1218
01:13:13,920 --> 01:13:15,730
you come back eight weeks later.

1219
01:13:15,730 --> 01:13:19,040
You find, oh my gosh, the
auditory receptive field

1220
01:13:19,040 --> 01:13:20,820
has actually moved.

1221
01:13:20,820 --> 01:13:22,230
It shifted.

1222
01:13:22,230 --> 01:13:25,720
You knew where you ere
recording from in the brain.

1223
01:13:25,720 --> 01:13:28,726
In this case, the recordings
are made in the optic tectum,

1224
01:13:28,726 --> 01:13:29,850
in the superior colliculus.

1225
01:13:31,460 --> 01:13:33,760
You know the dimensions
and you come back

1226
01:13:33,760 --> 01:13:36,540
and you're expecting to see
auditory receptive fields

1227
01:13:36,540 --> 01:13:39,670
like that, but
they've been shifted.

1228
01:13:39,670 --> 01:13:43,480
So juvenile owls with
prism experience,

1229
01:13:43,480 --> 01:13:46,170
given a number of
weeks to compensate,

1230
01:13:46,170 --> 01:13:49,340
shift the auditory
receptive fields

1231
01:13:49,340 --> 01:13:50,985
so that the two are
back in alignment.

1232
01:13:52,910 --> 01:13:55,030
This group has also
shown that if you

1233
01:13:55,030 --> 01:13:58,702
do this experiment
with adult owls,

1234
01:13:58,702 --> 01:14:00,195
you don't get such shifts.

1235
01:14:01,420 --> 01:14:05,860
You come back to an adult
owl with these same prisms,

1236
01:14:05,860 --> 01:14:08,230
eight weeks later you
have still a mismatch.

1237
01:14:09,260 --> 01:14:13,630
So plasticity clearly takes
place in the juveniles

1238
01:14:13,630 --> 01:14:15,740
and not in the adults.

1239
01:14:15,740 --> 01:14:20,230
And they've then likened
it to those old folks of us

1240
01:14:20,230 --> 01:14:22,120
trying to learn a
foreign language.

1241
01:14:22,120 --> 01:14:24,570
It's really tough because
we didn't have experience

1242
01:14:24,570 --> 01:14:25,973
with it while we were juveniles.

1243
01:14:27,860 --> 01:14:31,920
So neural plasticity and
these learning new things--

1244
01:14:31,920 --> 01:14:37,130
if this is learning something,
it's a bit of a stretch.

1245
01:14:37,130 --> 01:14:40,910
But it's more difficult to
learn things as an adult

1246
01:14:40,910 --> 01:14:42,015
than it is juveniles.

1247
01:14:43,720 --> 01:14:49,710
Now, the experiment-- an even
further twist in this paper

1248
01:14:49,710 --> 01:14:55,525
that we read for today, which is
now the subject is adult owls.

1249
01:14:55,525 --> 01:14:57,475
The recordings were
made from adult owls.

1250
01:14:58,840 --> 01:15:00,300
But there are two groups.

1251
01:15:00,300 --> 01:15:02,560
One is a plain, old
control adult owl.

1252
01:15:03,840 --> 01:15:08,300
The other is an adult owl
that when it was a juvenile

1253
01:15:08,300 --> 01:15:10,080
had experience with the prisms.

1254
01:15:12,290 --> 01:15:15,415
That experience was long
ago, six months before.

1255
01:15:17,240 --> 01:15:19,820
The prisms were on
for eight weeks or so.

1256
01:15:20,840 --> 01:15:22,910
These shifts took place.

1257
01:15:22,910 --> 01:15:25,000
The prisms were removed.

1258
01:15:25,000 --> 01:15:27,630
The owl is allowed
to grow up and become

1259
01:15:27,630 --> 01:15:29,550
an adult with normal vision.

1260
01:15:31,310 --> 01:15:36,040
Then, take these two groups of
owls, put the prisms on again.

1261
01:15:36,040 --> 01:15:39,910
We've already said that
adults don't have the capacity

1262
01:15:39,910 --> 01:15:40,930
to remap.

1263
01:15:42,270 --> 01:15:46,390
So many of them just
stayed the same.

1264
01:15:46,390 --> 01:15:48,640
These are the open circles here.

1265
01:15:48,640 --> 01:15:50,810
The auditory receptive
field is the same.

1266
01:15:50,810 --> 01:15:52,740
This is before the prisms.

1267
01:15:52,740 --> 01:15:54,180
This is eight weeks after.

1268
01:15:54,180 --> 01:15:57,295
The open circles are
from the adult owls

1269
01:15:57,295 --> 01:15:59,770
that didn't have any
juvenile prism experience.

1270
01:16:01,050 --> 01:16:04,960
But the adult owls-- they're
now adults with the prisms on.

1271
01:16:06,310 --> 01:16:09,480
The adult owls with the
juvenile prism experiences,

1272
01:16:09,480 --> 01:16:11,790
one of the neurons
is recorded here.

1273
01:16:11,790 --> 01:16:13,850
It has an auditory
receptive field

1274
01:16:13,850 --> 01:16:17,110
that's shifted to
make it in line

1275
01:16:17,110 --> 01:16:21,600
with the shifted visual
receptive fields.

1276
01:16:21,600 --> 01:16:24,500
This is now showing
then that adults

1277
01:16:24,500 --> 01:16:28,390
that have the
juvenile experience

1278
01:16:28,390 --> 01:16:33,210
have some plastic ability to
re-map their auditory receptive

1279
01:16:33,210 --> 01:16:36,870
field so that you have alignment
with the visual receptive

1280
01:16:36,870 --> 01:16:37,370
fields.

1281
01:16:38,550 --> 01:16:41,130
Clearly, a very
interesting experiment.

1282
01:16:41,130 --> 01:16:43,720
Showing, if you will,
a type of maybe memory

1283
01:16:43,720 --> 01:16:48,530
trace that these owls
have retained something

1284
01:16:48,530 --> 01:16:51,580
that was altered by the
juvenile experience.

1285
01:16:51,580 --> 01:16:53,170
So what could that be?

1286
01:16:59,440 --> 01:17:02,480
The group has in
other studies gone on

1287
01:17:02,480 --> 01:17:07,300
and looked at the projections
between these two boxes

1288
01:17:07,300 --> 01:17:10,590
here from the central
nucleus of the colliculus

1289
01:17:10,590 --> 01:17:12,160
to the external nucleus.

1290
01:17:13,470 --> 01:17:17,030
And clearly, shown in juvenile
animals that have the prism

1291
01:17:17,030 --> 01:17:19,650
experiences that the
axons that were headed

1292
01:17:19,650 --> 01:17:24,880
for direct innervation
take a little change

1293
01:17:24,880 --> 01:17:28,375
and they regrow some
axonal projections.

1294
01:17:29,690 --> 01:17:31,570
They've studied these.

1295
01:17:31,570 --> 01:17:38,070
And it looks like this regrowth
of axon between those two

1296
01:17:38,070 --> 01:17:38,640
areas.

1297
01:17:38,640 --> 01:17:41,410
And that's what was meant
by this little circle that

1298
01:17:41,410 --> 01:17:46,610
says locus or site of
plasticity is manifested

1299
01:17:46,610 --> 01:17:49,820
by a change in growth
of axons right there.

1300
01:17:49,820 --> 01:17:52,560
And maybe those
adult animals that

1301
01:17:52,560 --> 01:17:54,910
have juvenile
experience retain some

1302
01:17:54,910 --> 01:18:00,490
of those axonal projections
that have been changed

1303
01:18:00,490 --> 01:18:02,330
as a result of the experience.

1304
01:18:02,330 --> 01:18:04,600
And clearly, it takes
many weeks for axons

1305
01:18:04,600 --> 01:18:07,120
to grow or change
their connections.

1306
01:18:07,120 --> 01:18:09,340
Maybe that's the thing
that's much easier

1307
01:18:09,340 --> 01:18:12,620
to do if you're
a juvenile animal

1308
01:18:12,620 --> 01:18:14,890
and you're reacting to
these changed stimuli.

1309
01:18:17,550 --> 01:18:21,120
Another way to change
responses there

1310
01:18:21,120 --> 01:18:24,330
is to have everything
connected to everything,

1311
01:18:24,330 --> 01:18:28,780
but certain things emphasize
certain synapses very strong

1312
01:18:28,780 --> 01:18:30,850
and the weak ones not emphasize.

1313
01:18:30,850 --> 01:18:33,760
Maybe when you have
the prism experience,

1314
01:18:33,760 --> 01:18:37,260
the previously
de-emphasized synapses

1315
01:18:37,260 --> 01:18:41,670
become upregulated without
any change of axons.

1316
01:18:41,670 --> 01:18:43,240
But clearly, this
group has shown

1317
01:18:43,240 --> 01:18:46,440
that the axons have changed
their growth patterns.

1318
01:18:46,440 --> 01:18:49,664
So that could be a mechanism
for the plasticity.

1319
01:18:55,940 --> 01:18:59,990
And I think that's all
I wanted to say today.

1320
01:18:59,990 --> 01:19:02,580
So I have five minutes
if you guys want

1321
01:19:02,580 --> 01:19:05,960
to ask questions about anything.

1322
01:19:05,960 --> 01:19:06,794
Yeah.

1323
01:19:06,794 --> 01:19:09,214
AUDIENCE: Can you go
over how prisms actually

1324
01:19:09,214 --> 01:19:11,150
change auditory
perception again?

1325
01:19:11,150 --> 01:19:12,605
Like, how are the prisms--

1326
01:19:12,605 --> 01:19:13,855
PROFESSOR: Back to precedence?

1327
01:19:15,125 --> 01:19:16,000
Is that what you're--

1328
01:19:16,000 --> 01:19:18,090
AUDIENCE: How are
prisms changing--

1329
01:19:18,090 --> 01:19:19,610
PROFESSOR: How are prisms?

1330
01:19:19,610 --> 01:19:20,340
AUDIENCE: Yeah.

1331
01:19:20,340 --> 01:19:22,220
PROFESSOR: That's not clear.

1332
01:19:22,220 --> 01:19:24,316
What's clear is that the
fields are mismatched.

1333
01:19:27,396 --> 01:19:30,330
AUDIENCE: So we don't
know why that happens?

1334
01:19:30,330 --> 01:19:31,580
PROFESSOR: That's correct.

1335
01:19:31,580 --> 01:19:32,170
Yeah.

1336
01:19:32,170 --> 01:19:37,940
But I presume what is
happening to the owl

1337
01:19:37,940 --> 01:19:41,670
during these eight weeks is
the owl is seeing an object

1338
01:19:41,670 --> 01:19:43,150
and hearing it as well.

1339
01:19:43,150 --> 01:19:45,980
For example, a mouse down here.

1340
01:19:45,980 --> 01:19:48,390
The owl sees it
and it goes for it.

1341
01:19:48,390 --> 01:19:51,620
But actually, because its
visual fields are off,

1342
01:19:51,620 --> 01:19:54,210
it goes over here and
the object is over here.

1343
01:19:55,350 --> 01:19:59,060
But the auditory cues, if
it paid attention to them,

1344
01:19:59,060 --> 01:20:00,870
it would go here.

1345
01:20:00,870 --> 01:20:05,760
So it's sensing a
misalignment in experience.

1346
01:20:05,760 --> 01:20:07,220
There's no training
involved here,

1347
01:20:07,220 --> 01:20:11,530
but the owls are allowed
to hunt prey and experience

1348
01:20:11,530 --> 01:20:13,450
environmental sounds.

1349
01:20:13,450 --> 01:20:16,830
So they clearly then have
a mismatch between vision

1350
01:20:16,830 --> 01:20:19,160
and audition in
these eight weeks.

1351
01:20:20,457 --> 01:20:20,957
Yeah.

1352
01:20:20,957 --> 01:20:22,332
AUDIENCE: Does
this sort of then,

1353
01:20:22,332 --> 01:20:25,448
I guess suggest that the
auditory input is somehow

1354
01:20:25,448 --> 01:20:26,945
more important?

1355
01:20:26,945 --> 01:20:31,436
Because rather than say a visual
input shifting [INAUDIBLE]?

1356
01:20:34,450 --> 01:20:35,250
PROFESSOR: Yes.

1357
01:20:35,250 --> 01:20:36,695
Yes, you could say that.

1358
01:20:38,210 --> 01:20:41,180
I mean, it would be interesting
to do the converse experiment.

1359
01:20:41,180 --> 01:20:45,670
Keep the eyeballs normal
and somehow distort

1360
01:20:45,670 --> 01:20:49,040
the auditory receptive fields.

1361
01:20:49,040 --> 01:20:53,470
So you could do that with
ITDs by putting a tube

1362
01:20:53,470 --> 01:20:56,050
and lengthening the
ITD on one side.

1363
01:20:57,170 --> 01:21:01,550
That would be an interesting
sort of counter experiment

1364
01:21:01,550 --> 01:21:04,260
to get at what
you're asking about.

1365
01:21:04,260 --> 01:21:06,063
It would distort the other cue.

1366
01:21:13,170 --> 01:21:14,090
OK, great.

1367
01:21:14,090 --> 01:21:17,420
We'll see you back
on Wednesday then.