ABOUT THE SPEAKER
Beau Lotto - Neuroscientist, Artist
Beau Lotto is founder of Lottolab, a hybrid art studio and science lab. With glowing, interactive sculpture -- and old-fashioned peer-reviewed research--he's illuminating the mysteries of the brain's visual system.

Why you should listen

"Let there be perception," was evolution's proclamation, and so it was that all creatures, from honeybees to humans, came to see the world not as it is, but as was most useful. This uncomfortable place--where what an organism's brain sees diverges from what is actually out there--is what Beau Lotto and his team at Lottolab are exploring through their dazzling art-sci experiments and public illusions. Their Bee Matrix installation, for example, places a live bee in a transparent enclosure where gallerygoers may watch it seek nectar in a virtual meadow of luminous Plexiglas flowers. (Bees, Lotto will tell you, see colors much like we humans do.) The data captured isn't just discarded, either: it's put to good use in probing scientific papers, and sometimes in more exhibits.

At their home in London’s Science Museum, the lab holds "synesthetic workshops" where kids and adults make abstract paintings that computers interpret into music, and they host regular Lates--evenings of science, music and "mass experiments." Lotto is passionate about involving people from all walks of life in research on perception--both as subjects and as fellow researchers. One such program, called "i,scientist," in fact led to the publication of the first ever peer-reviewed scientific paper written by schoolchildren ("Blackawton Bees," December 2010). It starts, "Once upon a time ..."

These and Lotto's other conjurings are slowly, charmingly bending the science of perception--and our perceptions of what science can be.

More profile about the speaker
Beau Lotto | Speaker | TED.com
TEDGlobal 2009

Beau Lotto: Optical illusions show how we see

Filmed:
5,444,247 views

Beau Lotto's color games puzzle your vision, but they also spotlight what you can't normally see: how your brain works. This fun, first-hand look at your own versatile sense of sight reveals how evolution tints your perception of what's really out there.
- Neuroscientist, Artist
Beau Lotto is founder of Lottolab, a hybrid art studio and science lab. With glowing, interactive sculpture -- and old-fashioned peer-reviewed research--he's illuminating the mysteries of the brain's visual system. Full bio

Double-click the English transcript below to play the video.

00:13
I want to start with a game.
0
1000
3000
00:16
And to win this game,
1
4000
2000
00:18
all you have to do is see the reality that's in front of you
2
6000
3000
00:21
as it really is. All right?
3
9000
2000
00:23
So, we have two panels here,
4
11000
2000
00:25
of colored dots.
5
13000
2000
00:27
And one of those dots is the same
6
15000
3000
00:30
in the two panels. Okay?
7
18000
3000
00:33
And you have to tell me which one.
8
21000
2000
00:35
Now, narrow it down to
9
23000
3000
00:38
the gray one, the green one and, say, the orange one.
10
26000
3000
00:41
So, by a show of hands -- we'll start with the easiest one --
11
29000
3000
00:44
Show of hands: how many people think it's the gray one?
12
32000
4000
00:48
Really? Okay.
13
36000
2000
00:50
How many people think it's the green one?
14
38000
5000
00:55
And how many people think it's the orange one?
15
43000
4000
00:59
Pretty even split.
16
47000
3000
01:02
Let's find out what the reality is.
17
50000
3000
01:05
Here is the orange one.
18
53000
3000
01:08
(Laughter)
19
56000
2000
01:10
Here is the green one.
20
58000
3000
01:13
And here is the gray one.
21
61000
4000
01:17
(Laughter)
22
65000
3000
01:20
So, for all of you who saw that, you're a complete realist. All right?
23
68000
4000
01:24
(Laughter)
24
72000
2000
01:26
So, this is pretty amazing, actually, isn't it?
25
74000
2000
01:28
Because nearly every living system
26
76000
2000
01:30
has evolved the ability to detect light in one way or another.
27
78000
3000
01:33
So, for us, seeing color is one of the simplest things the brain does.
28
81000
5000
01:38
And yet, even at this most fundamental level,
29
86000
2000
01:40
context is everything.
30
88000
3000
01:43
What I want to talk about is not that context is everything,
31
91000
3000
01:46
but why is context everything.
32
94000
2000
01:48
Because it's answering that question that tells us not only
33
96000
4000
01:52
why we see what we do,
34
100000
2000
01:54
but who we are as individuals,
35
102000
2000
01:56
and who we are as a society.
36
104000
3000
01:59
But first, we have to ask another question,
37
107000
2000
02:01
which is, "What is color for?"
38
109000
2000
02:03
And instead of telling you, I'll just show you.
39
111000
2000
02:05
What you see here is a jungle scene,
40
113000
3000
02:08
and you see the surfaces according to the amount
41
116000
2000
02:10
of light that those surfaces reflect.
42
118000
2000
02:12
Now, can any of you see the predator that's about to jump out at you?
43
120000
5000
02:17
And if you haven't seen it yet, you're dead. Right?
44
125000
2000
02:19
(Laughter)
45
127000
2000
02:21
Can anyone see it? Anyone? No?
46
129000
2000
02:23
Now, let's see the surfaces according to the quality of light that they reflect.
47
131000
4000
02:27
And now you see it.
48
135000
3000
02:30
So, color enables us to see
49
138000
3000
02:33
the similarities and differences between surfaces,
50
141000
2000
02:35
according to the full spectrum of light that they reflect.
51
143000
3000
02:38
But what you've just done is, in many respects, mathematically impossible.
52
146000
4000
02:42
Why? Because, as Berkeley tells us,
53
150000
3000
02:45
we have no direct access to our physical world,
54
153000
3000
02:48
other than through our senses.
55
156000
2000
02:50
And the light that falls onto our eyes
56
158000
2000
02:52
is determined by multiple things in the world --
57
160000
2000
02:54
not only the color of objects,
58
162000
2000
02:56
but also the color of their illumination,
59
164000
2000
02:58
and the color of the space between us and those objects.
60
166000
3000
03:01
You vary any one of those parameters,
61
169000
2000
03:03
and you'll change the color of the light that falls onto your eye.
62
171000
5000
03:08
This is a huge problem because it means that
63
176000
2000
03:10
the same image could have an infinite number
64
178000
3000
03:13
of possible real-world sources.
65
181000
3000
03:16
So let me show you what I mean. Imagine that this is the back of your eye.
66
184000
3000
03:19
And these are two projections from the world.
67
187000
3000
03:22
They are identical in every single way.
68
190000
3000
03:25
Identical in shape, size, spectral content.
69
193000
4000
03:29
They are the same, as far as your eye is concerned.
70
197000
4000
03:33
And yet they come from completely different sources.
71
201000
5000
03:38
The one on the right
72
206000
2000
03:40
comes from a yellow surface,
73
208000
3000
03:43
in shadow, oriented facing the left,
74
211000
2000
03:45
viewed through a pinkish medium.
75
213000
3000
03:48
The one on the left comes from an orange surface,
76
216000
3000
03:51
under direct light, facing to the right,
77
219000
2000
03:53
viewed through a sort of a bluish medium.
78
221000
2000
03:55
Completely different meanings,
79
223000
3000
03:58
giving rise to the exact same retinal information.
80
226000
3000
04:01
And yet it's only the retinal information
81
229000
2000
04:03
that we get.
82
231000
2000
04:05
So how on Earth do we even see?
83
233000
3000
04:08
So, if you remember anything in this next 18 minutes,
84
236000
4000
04:12
remember this: that the light that falls on to your eye,
85
240000
3000
04:15
sensory information, is meaningless,
86
243000
2000
04:17
because it could mean literally anything.
87
245000
3000
04:20
And what's true for sensory information is true for information generally.
88
248000
3000
04:23
There is no inherent meaning in information.
89
251000
2000
04:25
It's what we do with that information that matters.
90
253000
4000
04:29
So, how do we see? Well, we see by learning to see.
91
257000
3000
04:32
So, the brain evolved the mechanisms for finding patterns,
92
260000
4000
04:36
finding relationships in information
93
264000
2000
04:38
and associating those relationships
94
266000
2000
04:40
with a behavioral meaning,
95
268000
2000
04:42
a significance, by interacting with the world.
96
270000
3000
04:45
We're very aware of this
97
273000
2000
04:47
in the form of more cognitive attributes, like language.
98
275000
3000
04:50
So, I'm going to give you some letter strings. And I want you to read them out for me,
99
278000
2000
04:52
if you can.
100
280000
2000
04:54
Audience: "Can you read this?"
101
282000
3000
04:57
"You are not reading this."
102
285000
2000
04:59
"What are you reading?"
103
287000
2000
05:01
Beau Lotto: "What are you reading?" Half the letters are missing. Right?
104
289000
3000
05:04
There is no a priori reason why an "H" has to go
105
292000
2000
05:06
between that "W" and "A."
106
294000
2000
05:08
But you put one there. Why?
107
296000
2000
05:10
Because in the statistics of your past experience
108
298000
2000
05:12
it would have been useful to do so. So you do so again.
109
300000
3000
05:15
And yet you don't put a letter after that first "T."
110
303000
3000
05:18
Why? Because it wouldn't have been useful in the past.
111
306000
3000
05:21
So you don't do it again.
112
309000
2000
05:23
So let me show you how quickly our brains can redefine normality,
113
311000
4000
05:27
even at the simplest thing the brain does, which is color.
114
315000
2000
05:29
So, if I could have the lights down up here.
115
317000
3000
05:32
I want you to first notice that those two desert scenes are physically the same.
116
320000
3000
05:35
One is simply the flipping of the other. Okay?
117
323000
5000
05:40
Now I want you to look at that dot
118
328000
2000
05:42
between the green and the red. Okay?
119
330000
3000
05:45
And I want you to stare at that dot. Don't look anywhere else.
120
333000
3000
05:48
And we're going to look at that for about 30 seconds,
121
336000
1000
05:49
which is a bit of a killer in an 18-minute talk.
122
337000
3000
05:52
(Laughter)
123
340000
1000
05:53
But I really want you to learn.
124
341000
2000
05:55
And I'll tell you -- don't look anywhere else --
125
343000
3000
05:58
and I'll tell you what's happening inside your head.
126
346000
2000
06:00
Your brain is learning. And it's learning that the right side of its visual field
127
348000
3000
06:03
is under red illumination;
128
351000
2000
06:05
the left side of its visual field is under green illumination.
129
353000
3000
06:08
That's what it's learning. Okay?
130
356000
3000
06:11
Now, when I tell you, I want you to look at the dot between the two desert scenes.
131
359000
5000
06:16
So why don't you do that now?
132
364000
2000
06:18
(Laughter)
133
366000
3000
06:21
Can I have the lights up again?
134
369000
2000
06:23
I take it from your response they don't look the same anymore. Right?
135
371000
4000
06:27
(Applause)
136
375000
1000
06:28
Why? Because your brain is seeing that same information
137
376000
3000
06:31
as if the right one is still under red light,
138
379000
2000
06:33
and the left one is still under green light.
139
381000
2000
06:35
That's your new normal.
140
383000
2000
06:37
So, what does this mean for context?
141
385000
2000
06:39
It means that I can take these two identical squares,
142
387000
2000
06:41
and I can put them in light and dark surrounds.
143
389000
2000
06:43
And now the one on the dark surround looks lighter than the one on the light surround.
144
391000
3000
06:46
What's significant is not simply the light and dark surrounds that matter.
145
394000
4000
06:50
It's what those light and dark surrounds meant for your behavior in the past.
146
398000
4000
06:54
So I'll show you what I mean. Here we have
147
402000
2000
06:56
that exact same illusion.
148
404000
2000
06:58
We have two identical tiles, on the left,
149
406000
2000
07:00
one in a dark surround, one in a light surround.
150
408000
2000
07:02
And the same thing over on the right.
151
410000
2000
07:04
Now, what I'm going to do is I'm going to review those two scenes.
152
412000
3000
07:07
But I'm not going to change anything within those boxes,
153
415000
2000
07:09
except their meaning.
154
417000
2000
07:11
And see what happens to your perception.
155
419000
2000
07:13
Notice that on the left
156
421000
2000
07:15
the two tiles look nearly completely opposite:
157
423000
3000
07:18
one very white and one very dark.
158
426000
2000
07:20
All right? Whereas on the right,
159
428000
2000
07:22
the two tiles look nearly the same.
160
430000
2000
07:24
And yet there is still one on a dark surround and one on a light surround.
161
432000
4000
07:28
Why? Because if the tile in that shadow
162
436000
3000
07:31
were in fact in shadow,
163
439000
2000
07:33
and reflecting the same amount of light to your eye
164
441000
2000
07:35
as the one outside the shadow,
165
443000
2000
07:37
it would have to be more reflective -- just the laws of physics.
166
445000
3000
07:40
So you see it that way.
167
448000
2000
07:42
Whereas on the right, the information is consistent
168
450000
3000
07:45
with those two tiles being under the same light.
169
453000
2000
07:47
If they are under the same light, reflecting the same amount of light
170
455000
2000
07:49
to your eye,
171
457000
2000
07:51
then they must be equally reflective.
172
459000
2000
07:53
So you see it that way.
173
461000
2000
07:55
Which means we can bring all this information together
174
463000
2000
07:57
to create some incredibly strong illusions.
175
465000
2000
07:59
This is one I made a few years ago.
176
467000
2000
08:01
And you'll notice you see a dark brown tile at the top,
177
469000
3000
08:04
and a bright orange tile at the side.
178
472000
3000
08:07
That is your perceptual reality. The physical reality
179
475000
2000
08:09
is that those two tiles are the same.
180
477000
5000
08:14
Here you see four gray tiles on your left,
181
482000
3000
08:17
seven gray tiles on the right.
182
485000
2000
08:19
I'm not going to change those tiles at all,
183
487000
2000
08:21
but I'm going to reveal the rest of the scene
184
489000
2000
08:23
and see what happens to your perception.
185
491000
3000
08:26
The four blue tiles on the left are gray.
186
494000
4000
08:30
The seven yellow tiles on the right are also gray.
187
498000
3000
08:33
They are the same. Okay?
188
501000
2000
08:35
Don't believe me? Let's watch it again.
189
503000
4000
08:39
What's true for color is also true for complex perceptions of motion.
190
507000
4000
08:43
So here we have --
191
511000
3000
08:46
let's turn this around -- a diamond.
192
514000
5000
08:51
And what I'm going to do is, I'm going to hold it here,
193
519000
2000
08:53
and I'm going to spin it.
194
521000
4000
08:57
And for all of you, you'll see it probably spinning this direction.
195
525000
3000
09:00
Now I want you to keep looking at it.
196
528000
3000
09:03
Move your eyes around, blink, maybe close one eye.
197
531000
2000
09:05
And suddenly it will flip, and start spinning the opposite direction.
198
533000
4000
09:09
Yes? Raise your hand if you got that. Yes?
199
537000
3000
09:12
Keep blinking. Every time you blink it will switch. Alright?
200
540000
4000
09:16
So I can ask you, which direction is it rotating?
201
544000
4000
09:20
How do you know?
202
548000
2000
09:22
Your brain doesn't know. Because both are equally likely.
203
550000
3000
09:25
So depending on where it looks, it flips
204
553000
2000
09:27
between the two possibilities.
205
555000
3000
09:30
Are we the only ones that see illusions?
206
558000
2000
09:32
The answer to this question is no.
207
560000
2000
09:34
Even the beautiful bumblebee,
208
562000
2000
09:36
with its mere one million brain cells,
209
564000
2000
09:38
which is 250 times fewer cells than you have in one retina,
210
566000
3000
09:41
sees illusions, does the most complicated things
211
569000
3000
09:44
that even our most sophisticated computers can't do.
212
572000
3000
09:47
So in my lab, we of course work on bumblebees.
213
575000
2000
09:49
Because we can completely control their experience,
214
577000
2000
09:51
and see how that alters the architecture of their brain.
215
579000
2000
09:53
And we do this in what we call the Bee Matrix.
216
581000
3000
09:56
And here you have the hive. You can see the queen bee,
217
584000
2000
09:58
that large bee in the middle there. Those are all her daughters, the eggs.
218
586000
3000
10:01
And they go back and forth between this hive
219
589000
3000
10:04
and the arena, via this tube.
220
592000
5000
10:09
And you'll see one of the bees come out here.
221
597000
2000
10:11
You see how she has a little number on her?
222
599000
3000
10:14
Yeah there is another one coming out. She has another number on her.
223
602000
3000
10:17
Now, they are not born that way. Right?
224
605000
3000
10:20
We pull them out, put them in the fridge, and they fall asleep.
225
608000
2000
10:22
And then you can superglue little numbers on them.
226
610000
2000
10:24
(Laughter)
227
612000
2000
10:26
And now, in this experiment they get rewarded if they go to the blue flowers.
228
614000
4000
10:30
And they land on the flower. They stick their tongue in there,
229
618000
3000
10:33
called a proboscis, and they drink sugar water.
230
621000
2000
10:35
Now she is drinking a glass of water that's about that big to you and I,
231
623000
3000
10:38
will do that about three times, and then fly.
232
626000
6000
10:44
And sometimes they learn not to go to the blue,
233
632000
2000
10:46
but to go to where the other bees go.
234
634000
2000
10:48
So they copy each other. They can count to five. They can recognize faces.
235
636000
3000
10:51
And here she comes down the ladder.
236
639000
3000
10:54
And she'll come into the hive, find an empty honey pot
237
642000
2000
10:56
and throw up, and that's honey.
238
644000
2000
10:58
(Laughter)
239
646000
1000
10:59
Now remember -- (Laughter)
240
647000
3000
11:02
-- she's supposed to be going to the blue flowers.
241
650000
2000
11:04
But what are these bees doing in the upper right corner?
242
652000
3000
11:07
It looks like they're going to green flowers.
243
655000
2000
11:09
Now, are they getting it wrong?
244
657000
3000
11:12
And the answer to the question is no. Those are actually blue flowers.
245
660000
3000
11:15
But those are blue flowers under green light.
246
663000
4000
11:19
So they are using the relationships between the colors to solve the puzzle,
247
667000
4000
11:23
which is exactly what we do.
248
671000
2000
11:25
So, illusions are often used,
249
673000
2000
11:27
especially in art, in the words of a more contemporary artist,
250
675000
4000
11:31
"to demonstrate the fragility of our senses."
251
679000
2000
11:33
Okay, this is complete rubbish.
252
681000
3000
11:36
The senses aren't fragile. And if they were, we wouldn't be here.
253
684000
3000
11:39
Instead, color tells us something completely different,
254
687000
4000
11:43
that the brain didn't actually evolve to see the world the way it is.
255
691000
3000
11:46
We can't. Instead, the brain evolved to see the world
256
694000
4000
11:50
the way it was useful to see in the past.
257
698000
3000
11:53
And how we see is by continually redefining normality.
258
701000
6000
11:59
So how can we take this
259
707000
4000
12:03
incredible capacity of plasticity of the brain
260
711000
3000
12:06
and get people to experience their world differently?
261
714000
3000
12:09
Well, one of the ways we do in my lab and studio
262
717000
3000
12:12
is we translate the light into sound
263
720000
3000
12:15
and we enable people to hear their visual world.
264
723000
4000
12:19
And they can navigate the world using their ears.
265
727000
3000
12:22
Here is David, in the right. And he is holding a camera.
266
730000
3000
12:25
On the left is what his camera sees.
267
733000
2000
12:27
And you'll see there is a line, a faint line going across that image.
268
735000
3000
12:30
That line is broken up into 32 squares.
269
738000
3000
12:33
In each square we calculate the average color.
270
741000
2000
12:35
And then we just simply translate that into sound.
271
743000
2000
12:37
And now he's going to
272
745000
3000
12:40
turn around, close his eyes,
273
748000
4000
12:44
and find a plate on the ground with his eyes closed.
274
752000
3000
13:06
He finds it. Amazing. Right?
275
774000
2000
13:08
So not only can we create a prosthetic for the visually impaired,
276
776000
2000
13:10
but we can also investigate how people
277
778000
3000
13:13
literally make sense of the world.
278
781000
3000
13:16
But we can also do something else. We can also make music with color.
279
784000
4000
13:20
So, working with kids,
280
788000
2000
13:22
they created images,
281
790000
2000
13:24
thinking about what might the images you see
282
792000
2000
13:26
sound like if we could listen to them.
283
794000
2000
13:28
And then we translated these images.
284
796000
2000
13:30
And this is one of those images.
285
798000
2000
13:32
And this is a six-year-old child composing a piece of music
286
800000
3000
13:35
for a 32-piece orchestra.
287
803000
3000
13:38
And this is what it sounds like.
288
806000
2000
14:06
So, a six-year-old child. Okay?
289
834000
3000
14:09
Now, what does all this mean?
290
837000
3000
14:12
What this suggests is that no one is an outside observer
291
840000
3000
14:15
of nature. Okay?
292
843000
2000
14:17
We are not defined by our central properties,
293
845000
2000
14:19
by the bits that make us up.
294
847000
2000
14:21
We're defined by our environment and our interaction with that environment --
295
849000
3000
14:24
by our ecology.
296
852000
2000
14:26
And that ecology is necessarily relative,
297
854000
4000
14:30
historical and empirical.
298
858000
2000
14:32
So what I'd like to finish with is this over here.
299
860000
6000
14:38
Because what I've been trying to do is really celebrate uncertainty.
300
866000
3000
14:41
Because I think only through uncertainty is there potential for understanding.
301
869000
4000
14:45
So, if some of you are still feeling a bit too certain,
302
873000
3000
14:48
I'd like to do this one.
303
876000
2000
14:50
So, if we have the lights down.
304
878000
2000
14:52
And what we have here --
305
880000
6000
14:58
Can everyone see 25 purple surfaces
306
886000
3000
15:01
on your left,
307
889000
2000
15:03
and 25, call it yellowish, surfaces on your right?
308
891000
4000
15:07
So, now, what I want to do:
309
895000
2000
15:09
I'm going to put the middle nine surfaces here
310
897000
2000
15:11
under yellow illumination
311
899000
2000
15:13
by simply putting a filter behind them.
312
901000
4000
15:17
All right. Now you can see that changes the light
313
905000
3000
15:20
that's coming through there. Right?
314
908000
2000
15:22
Because now the light is going through a yellowish filter
315
910000
2000
15:24
and then a purplish filter.
316
912000
2000
15:26
I'm going to do this opposite on the left here.
317
914000
5000
15:31
I'm going to put the middle nine under a purplish light.
318
919000
7000
15:38
Now, some of you will notice that the consequence is that
319
926000
4000
15:42
the light coming through those middle nine on the right,
320
930000
3000
15:45
or your left,
321
933000
2000
15:47
is exactly the same as the light coming through
322
935000
2000
15:49
the middle nine on your right.
323
937000
2000
15:51
Agreed? Yes?
324
939000
3000
15:54
Okay. So they are physically the same.
325
942000
2000
15:56
Let's pull the covers off.
326
944000
6000
16:02
Now remember,
327
950000
4000
16:06
you know the middle nine are exactly the same.
328
954000
3000
16:09
Do they look the same? No.
329
957000
4000
16:13
The question is, "Is that an illusion?"
330
961000
2000
16:15
And I'll leave you with that.
331
963000
2000
16:17
So, thank you very much.
332
965000
2000
16:19
(Applause)
333
967000
3000

▲Back to top

ABOUT THE SPEAKER
Beau Lotto - Neuroscientist, Artist
Beau Lotto is founder of Lottolab, a hybrid art studio and science lab. With glowing, interactive sculpture -- and old-fashioned peer-reviewed research--he's illuminating the mysteries of the brain's visual system.

Why you should listen

"Let there be perception," was evolution's proclamation, and so it was that all creatures, from honeybees to humans, came to see the world not as it is, but as was most useful. This uncomfortable place--where what an organism's brain sees diverges from what is actually out there--is what Beau Lotto and his team at Lottolab are exploring through their dazzling art-sci experiments and public illusions. Their Bee Matrix installation, for example, places a live bee in a transparent enclosure where gallerygoers may watch it seek nectar in a virtual meadow of luminous Plexiglas flowers. (Bees, Lotto will tell you, see colors much like we humans do.) The data captured isn't just discarded, either: it's put to good use in probing scientific papers, and sometimes in more exhibits.

At their home in London’s Science Museum, the lab holds "synesthetic workshops" where kids and adults make abstract paintings that computers interpret into music, and they host regular Lates--evenings of science, music and "mass experiments." Lotto is passionate about involving people from all walks of life in research on perception--both as subjects and as fellow researchers. One such program, called "i,scientist," in fact led to the publication of the first ever peer-reviewed scientific paper written by schoolchildren ("Blackawton Bees," December 2010). It starts, "Once upon a time ..."

These and Lotto's other conjurings are slowly, charmingly bending the science of perception--and our perceptions of what science can be.

More profile about the speaker
Beau Lotto | Speaker | TED.com