Would you like to inspect the original subtitles? These are the user uploaded subtitles that are being translated:
1
00:00:01,168 --> 00:00:04,035
NARRATOR: You can find
it in a rain forest,
2
00:00:04,038 --> 00:00:08,066
on the frontiers
of medical research,
3
00:00:08,075 --> 00:00:09,167
in the movies,
4
00:00:09,176 --> 00:00:14,103
and it's all over the world
of wireless communications.
5
00:00:14,115 --> 00:00:17,016
One of nature's
biggest design secrets
6
00:00:17,018 --> 00:00:18,213
has finally been revealed.
7
00:00:18,219 --> 00:00:21,086
My God! Of course, it's obvious.
8
00:00:21,088 --> 00:00:25,116
NARRATOR: It's an odd-looking shape
you may never have heard of,
9
00:00:25,126 --> 00:00:26,082
but it's everywhere around you,
10
00:00:26,093 --> 00:00:31,156
the jagged, repeating form
called a fractal.
11
00:00:31,165 --> 00:00:32,121
They're all over in biology.
12
00:00:32,133 --> 00:00:36,195
They are solutions that natural
selection has come up with
13
00:00:36,203 --> 00:00:40,003
over and over and over again.
14
00:00:40,007 --> 00:00:42,066
NARRATOR:
Fractals are in our lungs,
15
00:00:42,076 --> 00:00:45,034
kidneys, and blood vessels.
16
00:00:45,046 --> 00:00:47,140
Flowers, plants,
17
00:00:47,148 --> 00:00:48,980
weather systems,
18
00:00:48,983 --> 00:00:50,109
the rhythms of the heart,
19
00:00:50,117 --> 00:00:53,143
the very essences of life.
20
00:00:53,154 --> 00:00:56,112
NARRATOR: But it took a
maverick mathematician
21
00:00:56,123 --> 00:00:58,990
to figure out how they work.
22
00:00:58,993 --> 00:01:01,087
I don't play with formulas;
I play with pictures,
23
00:01:01,095 --> 00:01:03,154
and that is what I've been doing
all my life.
24
00:01:03,164 --> 00:01:07,226
NARRATOR: His was a bold challenge
to centuries-old assumptions
25
00:01:07,234 --> 00:01:13,037
about the various forms
that nature takes.
26
00:01:13,040 --> 00:01:16,135
The blinders came off
and people could see forms
27
00:01:16,143 --> 00:01:24,073
that were always there,
but formerly were invisible.
28
00:01:24,085 --> 00:01:27,077
NARRATOR:
Making the invisible visible.
29
00:01:27,088 --> 00:01:30,046
Finding order in disorder.
30
00:01:30,057 --> 00:01:33,220
What mysteries can it help us
unravel?
31
00:01:33,227 --> 00:01:36,094
Coming up next on NOVA:
32
00:01:36,097 --> 00:01:45,996
"Hunting the Hidden Dimension."
33
00:01:46,006 --> 00:01:48,065
Captioning sponsored
by EXXON MOBIL,
34
00:01:48,075 --> 00:01:49,133
DAVID H. KOCH,
35
00:01:49,143 --> 00:01:51,043
the HOWARD HUGHES
MEDICAL INSTITUTE,
36
00:01:51,045 --> 00:01:53,002
the CORPORATION
FOR PUBLIC BROADCASTING
37
00:01:53,013 --> 00:01:59,009
and VIEWERS LIKE YOU.
38
00:01:59,019 --> 00:02:05,015
Major funding for NOVA
is provided by the following:
39
00:02:05,025 --> 00:02:09,189
Taking on the world's
toughest energy challenges.
40
00:02:09,196 --> 00:02:14,191
And by:
41
00:02:14,201 --> 00:02:24,168
And...
42
00:02:24,178 --> 00:02:27,204
And by the Corporation
for Public Broadcasting
43
00:02:27,214 --> 00:02:32,214
and by contributions
to your PBS station from:
44
00:02:56,210 --> 00:03:00,135
NARRATOR: In 1978, at
Boeing Aircraft in Seattle,
45
00:03:00,147 --> 00:03:04,243
engineers were designing
experimental aircraft.
46
00:03:04,251 --> 00:03:06,208
Exotic things with two wings
or two tails
47
00:03:06,220 --> 00:03:09,053
or two fuselages,
just weird stuff.
48
00:03:09,056 --> 00:03:11,081
'Cause who knows,
it might work.
49
00:03:11,091 --> 00:03:15,016
NARRATOR: A young computer
scientist named Loren Carpenter
50
00:03:15,029 --> 00:03:16,155
was helping them visualize
51
00:03:16,163 --> 00:03:19,064
what the planes might look like
in flight.
52
00:03:19,066 --> 00:03:21,091
CARPENTER:
I would get the data from them
53
00:03:21,101 --> 00:03:23,058
and make pictures, uh,
from various angles.
54
00:03:23,070 --> 00:03:26,028
But I wanted to be able to put
a mountain behind them,
55
00:03:26,040 --> 00:03:28,998
because every Boeing publicity
photo in existence
56
00:03:29,009 --> 00:03:30,067
has a mountain behind it.
57
00:03:30,077 --> 00:03:31,101
But there was no way
to do mountains.
58
00:03:31,111 --> 00:03:33,045
Mountains had millions and
millions of little triangles
59
00:03:33,047 --> 00:03:34,208
or polygons, or whatever
you want to call it,
60
00:03:34,215 --> 00:03:37,048
and, uh, we had enough trouble
with a hundred.
61
00:03:37,051 --> 00:03:39,042
Especially in those days
when our machines were, uh,
62
00:03:39,053 --> 00:03:40,111
slower than the ones
you have in your watch.
63
00:03:40,120 --> 00:03:44,114
NARRATOR: Carpenter didn't want
to make just any mountains.
64
00:03:44,124 --> 00:03:49,119
He wanted to create a landscape
the planes could fly through;
65
00:03:49,129 --> 00:03:51,063
but there was no way to do that
66
00:03:51,065 --> 00:03:54,194
with existing animation
techniques.
67
00:03:54,201 --> 00:03:56,158
From the time movies began,
68
00:03:56,170 --> 00:03:59,162
animators had to draw
each frame by hand--
69
00:03:59,173 --> 00:04:02,268
thousands of them
to make even a short cartoon.
70
00:04:02,276 --> 00:04:06,144
(echoing):
That's why they call me Thumper.
71
00:04:06,146 --> 00:04:10,208
NARRATOR: But that was before
Loren Carpenter stumbled across
72
00:04:10,217 --> 00:04:13,118
the work of a little-known
mathematician
73
00:04:13,120 --> 00:04:15,111
named Benoit Mandelbrot.
74
00:04:15,122 --> 00:04:19,150
CARPENTER: In 1978, I ran into
this book at a bookstore,
75
00:04:19,159 --> 00:04:21,116
Fractals:
Form, Chance, and Dimension
76
00:04:21,128 --> 00:04:22,118
by Benoit Mandelbrot,
77
00:04:22,129 --> 00:04:24,120
and it has to do with
the fractal geometry of nature.
78
00:04:24,131 --> 00:04:27,055
So I bought the book,
took it home and read it--
79
00:04:27,067 --> 00:04:29,092
cover to cover,
every last little word,
80
00:04:29,103 --> 00:04:31,265
including the footnotes
and references-- twice.
81
00:04:31,272 --> 00:04:36,039
NARRATOR: In his book Mandelbrot
said that many forms in nature
82
00:04:36,043 --> 00:04:39,240
can be described mathematically
as "fractals,"
83
00:04:39,246 --> 00:04:42,170
a word he invented
to define shapes
84
00:04:42,182 --> 00:04:44,207
that look jagged and broken.
85
00:04:44,218 --> 00:04:46,175
He said that you can create
a fractal
86
00:04:46,186 --> 00:04:51,147
by taking a smooth-looking shape
and breaking it into pieces,
87
00:04:51,158 --> 00:04:53,252
over and over again.
88
00:04:53,260 --> 00:04:58,164
Carpenter decided he'd try doing
that on his computer.
89
00:04:58,165 --> 00:05:00,099
CARPENTER:
Within three days,
90
00:05:00,100 --> 00:05:02,057
I was producing pictures
of mountains
91
00:05:02,069 --> 00:05:04,163
on my computer at work.
92
00:05:04,171 --> 00:05:06,162
The method is dead-simple.
93
00:05:06,173 --> 00:05:09,131
You start with a landscape made
out of very rough triangles,
94
00:05:09,143 --> 00:05:11,168
big ones,
and then for each triangle,
95
00:05:11,178 --> 00:05:15,137
break it into four triangles,
and then do that again,
96
00:05:15,149 --> 00:05:17,083
and again and again...
97
00:05:17,084 --> 00:05:18,176
NARRATOR:
Endless repetition--
98
00:05:18,185 --> 00:05:21,177
what mathematicians call
"iteration."
99
00:05:21,188 --> 00:05:27,082
It's one of the keys
to fractal geometry.
100
00:05:27,094 --> 00:05:28,220
CARPENTER:
The pictures were stunning.
101
00:05:28,228 --> 00:05:30,128
They were just totally stunning.
102
00:05:30,130 --> 00:05:32,189
No one has ever seen anything
like this,
103
00:05:32,199 --> 00:05:34,224
and I just opened
a whole new door
104
00:05:34,234 --> 00:05:38,159
to a new world
of making pictures.
105
00:05:38,172 --> 00:05:42,075
And it got the computer graphics
community excited about fractals
106
00:05:42,076 --> 00:05:43,134
because suddenly,
they were easy to do.
107
00:05:43,143 --> 00:05:47,205
And so people started doing them
all over the place.
108
00:05:47,214 --> 00:05:52,141
NARRATOR: Carpenter soon left
Boeing to join Lucasfilm,
109
00:05:52,152 --> 00:05:54,143
where, instead of making
mountains,
110
00:05:54,154 --> 00:05:56,111
he created a whole new planet
111
00:05:56,123 --> 00:05:59,218
for Star Trek II:
The Wrath of Khan.
112
00:05:59,226 --> 00:06:04,153
It was the first-ever completely
computer-generated sequence
113
00:06:04,164 --> 00:06:06,064
in a feature film...
114
00:06:06,066 --> 00:06:09,195
Fascinating.
115
00:06:09,203 --> 00:06:17,202
NARRATOR: ...made possible by the new
mathematics of fractal geometry.
116
00:06:17,211 --> 00:06:22,047
Benoit Mandelbrot, whose work
had inspired that innovation,
117
00:06:22,049 --> 00:06:23,244
was someone who prided himself
118
00:06:23,250 --> 00:06:27,153
on standing
outside the mainstream.
119
00:06:27,154 --> 00:06:30,112
I can see things
that nobody else suspects
120
00:06:30,124 --> 00:06:32,081
until I point out to them.
121
00:06:32,092 --> 00:06:33,150
"Oh, of course, of course."
122
00:06:33,160 --> 00:06:34,184
But they haven't seen it before.
123
00:06:34,194 --> 00:06:38,097
NARRATOR:
You can see it in the clouds,
124
00:06:38,098 --> 00:06:40,055
in the mountains,
125
00:06:40,067 --> 00:06:44,061
even inside the human body.
126
00:06:44,071 --> 00:06:46,267
The key to fractal geometry
and the thing that evaded anyone
127
00:06:46,273 --> 00:06:48,230
until, really, Mandelbrot
sort of said
128
00:06:48,242 --> 00:06:51,143
this is the way to look
at things, is that...
129
00:06:51,145 --> 00:06:52,203
if you look on the surface,
130
00:06:52,212 --> 00:06:56,080
you see complexity, and it looks
very non-mathematical.
131
00:06:56,083 --> 00:07:00,145
What Mandelbrot said was that
think not of what you see
132
00:07:00,154 --> 00:07:04,148
but what it took to produce
what you see.
133
00:07:04,158 --> 00:07:06,217
NARRATOR:
It takes endless repetition
134
00:07:06,226 --> 00:07:10,185
and that gives rise to one
of the defining characteristics
135
00:07:10,197 --> 00:07:11,187
of a fractal,
136
00:07:11,198 --> 00:07:15,101
what mathematicians call
"self-similarity."
137
00:07:15,102 --> 00:07:19,198
The main idea is always,
as you zoom in and zoom out,
138
00:07:19,206 --> 00:07:21,231
the objects look the same.
139
00:07:21,241 --> 00:07:23,198
If you look at something
at this scale...
140
00:07:23,210 --> 00:07:27,204
and then you pick a small piece
of it and you zoom in,
141
00:07:27,214 --> 00:07:29,114
it looks very much the same.
142
00:07:29,116 --> 00:07:32,279
NARRATOR: The whole of the
fractal looks just like a part,
143
00:07:32,286 --> 00:07:36,280
which looks just like
the next smaller part.
144
00:07:36,290 --> 00:07:43,185
The similarity of the pattern
just keeps on going.
145
00:07:43,197 --> 00:07:45,291
One of the most familiar
examples
146
00:07:45,299 --> 00:07:48,132
of self-similarity is a tree.
147
00:07:48,135 --> 00:07:50,229
If we look
at each of the nodes,
148
00:07:50,237 --> 00:07:51,261
the branching nodes
of this tree,
149
00:07:51,271 --> 00:07:54,297
what you'll actually see is
that the pattern of branching
150
00:07:54,308 --> 00:07:57,107
is very similar
throughout the tree.
151
00:07:57,110 --> 00:08:00,273
As we go from the base
of the tree to higher up,
152
00:08:00,280 --> 00:08:03,113
you'll see we'll have
mother branches
153
00:08:03,116 --> 00:08:06,142
and branching then
into daughter branches.
154
00:08:06,153 --> 00:08:08,178
If we take this one branch
and node
155
00:08:08,188 --> 00:08:11,180
and then go up
to a higher branch or node,
156
00:08:11,191 --> 00:08:13,250
what we'll actually find is,
again,
157
00:08:13,260 --> 00:08:15,251
that the pattern of branching
is similar.
158
00:08:15,262 --> 00:08:20,098
Again, this pattern of branching
is repeated throughout the tree,
159
00:08:20,100 --> 00:08:23,229
all the way, ultimately,
out to the tips,
160
00:08:23,237 --> 00:08:24,227
where the leaves are.
161
00:08:24,238 --> 00:08:28,163
NARRATOR: You see
self-similarity in everything:
162
00:08:28,175 --> 00:08:30,166
from a stalk of broccoli
163
00:08:30,177 --> 00:08:33,169
to the surface of the moon,
164
00:08:33,180 --> 00:08:37,208
to the arteries that transport
blood through our bodies.
165
00:08:37,217 --> 00:08:39,151
But Mandelbrot's fascination
166
00:08:39,152 --> 00:08:41,211
with these
irregular-looking shapes
167
00:08:41,221 --> 00:08:42,279
put him squarely at odds
168
00:08:42,289 --> 00:08:45,247
with centuries
of mathematical tradition.
169
00:08:45,259 --> 00:08:51,130
In the whole of science,
the whole of mathematics,
170
00:08:51,131 --> 00:08:53,088
a smoothness with everything.
171
00:08:53,100 --> 00:09:01,269
What I did was to open up
roughness for investigation.
172
00:09:01,275 --> 00:09:04,108
DEVLIN: We used mathematics
to build the pyramids,
173
00:09:04,111 --> 00:09:05,272
to construct the Parthenon.
174
00:09:05,279 --> 00:09:06,303
We use mathematics to study
175
00:09:06,313 --> 00:09:08,304
the regular motion
of the planets
176
00:09:08,315 --> 00:09:11,080
and so forth.
177
00:09:11,084 --> 00:09:13,314
We became used to the fact
that certain patterns
178
00:09:13,320 --> 00:09:16,187
were amenable to mathematics--
the architectural ones,
179
00:09:16,189 --> 00:09:19,181
largely the patterns
of human-made structures,
180
00:09:19,192 --> 00:09:22,116
where we had straight lines
and circles
181
00:09:22,129 --> 00:09:24,154
and the perfect
geometric shapes.
182
00:09:24,164 --> 00:09:28,089
The basic assumption that
underlies classical mathematics
183
00:09:28,101 --> 00:09:30,263
is that everything
is extremely regular.
184
00:09:30,270 --> 00:09:34,264
I mean, you reduce everything
to straight lines.
185
00:09:34,274 --> 00:09:37,141
Circles, triangles.
186
00:09:37,144 --> 00:09:38,202
Flat surfaces.
187
00:09:38,211 --> 00:09:38,336
Pyramids,
188
00:09:38,345 --> 00:09:41,303
tetrahedrons, icosahedrons,
dodecahedrons.
189
00:09:41,315 --> 00:09:42,214
Smooth edges.
190
00:09:42,215 --> 00:09:47,278
DEVLIN: Classical mathematics is
really only well-suited to study
191
00:09:47,287 --> 00:09:49,187
the world that we've created,
192
00:09:49,189 --> 00:09:53,285
the things we've built using
that classical mathematics.
193
00:09:53,293 --> 00:09:55,159
The patterns in nature,
194
00:09:55,162 --> 00:09:57,153
the things that were
already there
195
00:09:57,164 --> 00:09:58,256
before we came onto the planet,
196
00:09:58,265 --> 00:10:02,133
the trees, the plants,
the clouds, the weather systems,
197
00:10:02,135 --> 00:10:06,231
those were outside
of mathematics.
198
00:10:06,239 --> 00:10:08,196
NARRATOR:
Until the 1970s,
199
00:10:08,208 --> 00:10:12,202
when Benoit Mandelbrot
introduced his new geometry.
200
00:10:12,212 --> 00:10:17,173
DEVLIN: Mandelbrot came along and
said "Hey, guys, all you need to do
201
00:10:17,184 --> 00:10:21,178
"is look at these patterns
of nature in the right way,
202
00:10:21,188 --> 00:10:23,179
"and you can apply mathematics.
203
00:10:23,190 --> 00:10:26,182
"There is an order
beneath the seeming chaos.
204
00:10:26,193 --> 00:10:28,127
"You can write down formulas
205
00:10:28,128 --> 00:10:31,086
"that describe clouds,
and flowers and plants.
206
00:10:31,098 --> 00:10:33,226
"It's just that they're
different kinds of formulas,
207
00:10:33,233 --> 00:10:38,233
and they give you
a different kind of geometry."
208
00:10:44,244 --> 00:10:49,239
The big question is
why did it take till the 1970s
209
00:10:49,249 --> 00:10:51,274
Before somebody wrote a book
210
00:10:51,284 --> 00:10:54,185
called The Fractal Geometry
of Nature.
211
00:10:54,187 --> 00:10:56,178
If they're all around us,
212
00:10:56,189 --> 00:10:57,350
why didn't we see them before?
213
00:10:57,357 --> 00:10:59,189
The answer seems to be
214
00:10:59,192 --> 00:11:02,184
well, people were seeing
them before.
215
00:11:02,195 --> 00:11:08,225
People clearly recognized this
repeating quality in nature.
216
00:11:08,235 --> 00:11:11,330
NARRATOR: People like the great
19th century Japanese artist
217
00:11:11,338 --> 00:11:14,330
Katsushika Hokusai.
218
00:11:14,341 --> 00:11:19,177
If you look well enough,
you see a shadow of a cloud
219
00:11:19,179 --> 00:11:20,203
over Mount Fuji.
220
00:11:20,213 --> 00:11:26,175
The cloud is billows
upon billows upon billows.
221
00:11:26,186 --> 00:11:27,244
TAYLOR:
Hokusai, the great wave.
222
00:11:27,254 --> 00:11:29,313
You know, on top
of the great wave,
223
00:11:29,322 --> 00:11:32,314
there's smaller waves.
224
00:11:32,325 --> 00:11:34,259
MANDELBROT:
After my book
225
00:11:34,261 --> 00:11:36,252
mentioned
that Hokusai was fractal,
226
00:11:36,263 --> 00:11:40,131
I got inundated
with people saying,
227
00:11:40,133 --> 00:11:42,295
"Now we understand Hokusai."
228
00:11:42,302 --> 00:11:47,206
Hokusai was drawing fractals.
229
00:11:47,207 --> 00:11:48,163
TAYLOR:
Everybody thinks
230
00:11:48,175 --> 00:11:50,234
that mathematicians
are very different from artists.
231
00:11:50,243 --> 00:11:52,200
I've come to realize that art
232
00:11:52,212 --> 00:11:54,203
is actually really close
to mathematics,
233
00:11:54,214 --> 00:11:57,172
and that they're just using
different language.
234
00:11:57,184 --> 00:12:01,109
And so, for Mandelbrot,
it's not about equations.
235
00:12:01,121 --> 00:12:08,289
It's about how do we explain
this visual phenomenon.
236
00:12:08,295 --> 00:12:10,195
NARRATOR:
Mandelbrot's fascination
237
00:12:10,197 --> 00:12:14,259
with the visual side of math
began when he was a student.
238
00:12:14,267 --> 00:12:20,161
MANDELBROT: It is only in
January '44 that suddenly,
239
00:12:20,173 --> 00:12:21,334
I fell in love with
mathematics--
240
00:12:21,341 --> 00:12:23,207
and not mathematics in general--
241
00:12:23,210 --> 00:12:29,138
with geometry in its most
concrete, sensual form.
242
00:12:29,149 --> 00:12:31,140
That part
of geometry which...
243
00:12:31,151 --> 00:12:35,213
in which mathematics
and the eye meet.
244
00:12:35,222 --> 00:12:38,317
The professor was talking
about algebra,
245
00:12:38,325 --> 00:12:44,162
but I began to see in my mind
geometric pictures which fitted
246
00:12:44,164 --> 00:12:47,190
this algebra,
and once you see these pictures,
247
00:12:47,200 --> 00:12:48,292
the answer become obvious.
248
00:12:48,301 --> 00:12:53,171
So, I discovered something
which I had no clue before,
249
00:12:53,173 --> 00:12:57,235
that I knew how to transform
in my mind instantly
250
00:12:57,244 --> 00:13:00,202
the formulas into pictures.
251
00:13:00,213 --> 00:13:02,307
NARRATOR:
As a young man,
252
00:13:02,315 --> 00:13:05,376
Mandelbrot developed
a strong sense of self-reliance,
253
00:13:05,385 --> 00:13:07,342
shaped in large part
254
00:13:07,354 --> 00:13:09,186
by his experience as a Jew
255
00:13:09,189 --> 00:13:13,148
living under Nazi occupation
in France.
256
00:13:13,160 --> 00:13:15,117
For four years,
257
00:13:15,128 --> 00:13:17,187
he managed
to evade the constant threat
258
00:13:17,197 --> 00:13:18,187
of arrest and deportation.
259
00:13:18,198 --> 00:13:21,293
MANDELBROT: There is nothing
more, um, hardening,
260
00:13:21,301 --> 00:13:24,259
in a certain sense,
than surviving a war.
261
00:13:24,271 --> 00:13:28,196
Even not a soldier,
but as a hunted civilian.
262
00:13:28,208 --> 00:13:29,334
I knew... I knew how to act,
263
00:13:29,342 --> 00:13:33,301
and I didn't trust
people's wisdom very much.
264
00:13:33,313 --> 00:13:38,308
NARRATOR: After the war,
Mandelbrot got his Ph.D.
265
00:13:38,318 --> 00:13:41,219
He tried teaching
at a French university,
266
00:13:41,221 --> 00:13:43,212
but he didn't seem to fit in.
267
00:13:43,223 --> 00:13:44,349
MANDELBROT:
They say, well,
268
00:13:44,357 --> 00:13:46,348
I'm very gifted,
but very misled,
269
00:13:46,359 --> 00:13:48,293
and I do things
the wrong way.
270
00:13:48,295 --> 00:13:51,253
I was very much, um,
a fish out of water.
271
00:13:51,264 --> 00:13:55,167
So I abandoned this job
in France and took the gamble
272
00:13:55,168 --> 00:13:57,262
to go to IBM.
273
00:13:57,270 --> 00:13:59,204
NARRATOR:
It was 1958.
274
00:13:59,206 --> 00:14:02,369
The giant American corporation
was pioneering a technology
275
00:14:02,375 --> 00:14:05,174
that would soon revolutionize
276
00:14:05,178 --> 00:14:07,203
the way we all live:
277
00:14:07,214 --> 00:14:10,309
the computer.
278
00:14:10,317 --> 00:14:14,185
IBM was looking
for creative thinkers--
279
00:14:14,187 --> 00:14:17,248
non-conformists, even rebels.
280
00:14:17,257 --> 00:14:20,318
People like Benoit Mandelbrot.
281
00:14:20,327 --> 00:14:23,353
MANDELBROT: In fact, they
had cornered the market
282
00:14:23,363 --> 00:14:26,389
for a certain type of oddball.
283
00:14:26,399 --> 00:14:29,323
We never had
the slightest feeling
284
00:14:29,336 --> 00:14:33,330
of being the establishment.
285
00:14:33,340 --> 00:14:36,298
NARRATOR: Mandelbrot's colleagues
told the young mathematician
286
00:14:36,309 --> 00:14:39,267
about a problem of great concern
to the company.
287
00:14:39,279 --> 00:14:44,206
IBM engineers were transmitting
computer data over phone lines,
288
00:14:44,217 --> 00:14:48,245
but sometimes, the information
was not getting through.
289
00:14:48,255 --> 00:14:50,280
MANDELBROT:
They realized
290
00:14:50,290 --> 00:14:51,348
that every so often,
291
00:14:51,358 --> 00:14:56,228
the lines became,
uh, extremely noisy.
292
00:14:56,229 --> 00:14:58,186
Errors occurred
in large numbers.
293
00:14:58,198 --> 00:15:02,192
It was indeed
an extremely messy situation.
294
00:15:02,202 --> 00:15:05,399
NARRATOR: Mandelbrot
graphed the noise data,
295
00:15:05,405 --> 00:15:07,362
and what he saw surprised him.
296
00:15:07,374 --> 00:15:12,301
Regardless of the timescale,
the graph looked similar.
297
00:15:12,312 --> 00:15:14,303
One day:
298
00:15:14,314 --> 00:15:17,181
one hour, one second--
299
00:15:17,183 --> 00:15:18,275
it didn't matter.
300
00:15:18,285 --> 00:15:21,277
It looked about the same.
301
00:15:21,288 --> 00:15:24,189
It turned out to be self-similar
with a vengeance.
302
00:15:24,190 --> 00:15:26,284
NARRATOR:
Mandelbrot was amazed.
303
00:15:26,293 --> 00:15:29,217
The strange pattern reminded him
of something
304
00:15:29,229 --> 00:15:32,187
that had intrigued him
as a young man--
305
00:15:32,198 --> 00:15:33,324
a mathematical mystery
306
00:15:33,333 --> 00:15:36,394
that dated back
nearly 100 years:
307
00:15:36,403 --> 00:15:41,204
the mystery of the monsters.
308
00:15:41,207 --> 00:15:44,370
The story really begins
in the late 19th century.
309
00:15:44,377 --> 00:15:47,278
Mathematicians had written down
a formal description
310
00:15:47,280 --> 00:15:48,270
of what a curve must be.
311
00:15:48,281 --> 00:15:51,410
But within that description,
there were these other things,
312
00:15:51,418 --> 00:15:55,412
things that satisfied the formal
definition of what a curve is,
313
00:15:55,422 --> 00:15:58,346
but were so weird
that you could never draw them,
314
00:15:58,358 --> 00:16:00,258
or you couldn't even imagine
drawing them.
315
00:16:00,260 --> 00:16:03,161
They were just regarded
as monsters
316
00:16:03,163 --> 00:16:05,154
or things beyond the realm.
317
00:16:05,165 --> 00:16:06,360
ABRAHAM:
They're not lines.
318
00:16:06,366 --> 00:16:08,266
They're nothing like lines.
319
00:16:08,268 --> 00:16:09,292
They're not circles.
320
00:16:09,302 --> 00:16:12,328
They were, like, really,
really weird.
321
00:16:12,339 --> 00:16:17,175
NARRATOR: The German mathematician
Georg Cantor created the first
322
00:16:17,177 --> 00:16:19,168
of the monsters in 1883.
323
00:16:19,179 --> 00:16:22,240
RON EGLASH: He just took a
straight line, and he said,
324
00:16:22,248 --> 00:16:24,239
"I'm gonna break this line
into thirds,
325
00:16:24,250 --> 00:16:26,207
and the middle third
I'm gonna erase."
326
00:16:26,219 --> 00:16:28,415
So you're left
with two lines at each end.
327
00:16:28,421 --> 00:16:30,378
And now I'm gonna take
those two lines,
328
00:16:30,390 --> 00:16:33,416
take out the middle third,
and we'll do it again.
329
00:16:33,426 --> 00:16:36,350
So he does that over
and over again.
330
00:16:36,363 --> 00:16:38,229
Most people would think,
331
00:16:38,231 --> 00:16:39,392
well, if I've thrown
everything away,
332
00:16:39,399 --> 00:16:42,198
eventually,
there's nothing left.
333
00:16:42,202 --> 00:16:43,226
Not the case.
334
00:16:43,236 --> 00:16:44,397
There's not just one point left.
335
00:16:44,404 --> 00:16:46,270
There's not just
two points left.
336
00:16:46,272 --> 00:16:49,264
There's infinitely
many points left.
337
00:16:49,275 --> 00:16:52,233
NARRATOR: As you zoom
in on the Cantor set,
338
00:16:52,245 --> 00:16:53,235
the pattern stays the same,
339
00:16:53,246 --> 00:16:58,377
much like the noise patterns
that Mandelbrot had seen at IBM.
340
00:16:58,385 --> 00:17:01,411
Another strange shape
was put forward
341
00:17:01,421 --> 00:17:07,190
by the Swedish mathematician
Helge Von Koch.
342
00:17:07,193 --> 00:17:09,389
Koch said, well, you start
with an equilateral triangle,
343
00:17:09,396 --> 00:17:12,354
one of the classical
Euclidean geometric figures,
344
00:17:12,365 --> 00:17:13,355
and on each side...
345
00:17:13,366 --> 00:17:15,391
...I take a piece,
and I substitute two pieces
346
00:17:15,402 --> 00:17:17,302
that are now longer
than the original piece.
347
00:17:17,303 --> 00:17:19,397
And for each of those pieces,
I substitute two pieces
348
00:17:19,406 --> 00:17:21,431
that are each longer
than the original piece.
349
00:17:21,441 --> 00:17:23,273
Over and over again.
350
00:17:23,276 --> 00:17:24,334
You get the same shape, but now,
351
00:17:24,344 --> 00:17:27,268
each line has that little
triangular bump on it.
352
00:17:27,280 --> 00:17:28,213
And I break it again,
353
00:17:28,214 --> 00:17:29,306
and I break it again,
and I break it again,
354
00:17:29,315 --> 00:17:31,249
and each time I break it,
the line gets longer.
355
00:17:31,251 --> 00:17:32,309
Every iteration, every cycle,
356
00:17:32,318 --> 00:17:36,312
he's adding on
another little triangle.
357
00:17:36,322 --> 00:17:40,281
Imagine iterating that process
of adding little bits,
358
00:17:40,293 --> 00:17:41,419
infinitely many times.
359
00:17:41,428 --> 00:17:44,352
What you end up with
is something
360
00:17:44,364 --> 00:17:47,390
that's infinitely long.
361
00:17:47,400 --> 00:17:50,233
NARRATOR:
The Koch Curve was a paradox.
362
00:17:50,236 --> 00:17:53,433
To the eye, the curve appears
to be perfectly finite.
363
00:17:53,440 --> 00:17:57,343
But mathematically,
it is infinite,
364
00:17:57,343 --> 00:18:00,210
which means
it cannot be measured.
365
00:18:00,213 --> 00:18:03,410
EGLASH: At the time they called
it a pathological curve,
366
00:18:03,416 --> 00:18:06,249
because it made no sense,
according to the way
367
00:18:06,252 --> 00:18:07,344
people were thinking
about measurement,
368
00:18:07,353 --> 00:18:08,377
and Euclidean geometry
and so on.
369
00:18:08,388 --> 00:18:12,291
NARRATOR: But the Koch Curve
turned out to be crucial
370
00:18:12,292 --> 00:18:14,351
to a nagging measurement
problem:
371
00:18:14,360 --> 00:18:16,419
the length of a coastline.
372
00:18:16,429 --> 00:18:22,232
In the 1940s, British scientist
Lewis Richardson had observed
373
00:18:22,235 --> 00:18:23,327
that there can be
great variation
374
00:18:23,336 --> 00:18:26,362
between different measurements
of a coastline.
375
00:18:26,372 --> 00:18:28,431
It depends on how long
your yardstick is
376
00:18:28,441 --> 00:18:30,273
and how much patience you have.
377
00:18:30,276 --> 00:18:32,301
If you measure the coastline
of Britain
378
00:18:32,312 --> 00:18:35,270
with a one-mile yardstick,
you'd get so many yardsticks,
379
00:18:35,281 --> 00:18:36,373
which gives you so many miles.
380
00:18:36,382 --> 00:18:39,249
If you measure it
with a one-foot yardstick,
381
00:18:39,252 --> 00:18:40,413
it turns out that it's longer.
382
00:18:40,420 --> 00:18:43,253
And every time
you use a shorter yardstick,
383
00:18:43,256 --> 00:18:43,449
you get a longer number.
384
00:18:43,456 --> 00:18:46,323
DEVLIN: Because you can always
find finer indentations.
385
00:18:46,326 --> 00:18:50,285
NARRATOR: Mandelbrot saw that the
finer and finer indentations
386
00:18:50,296 --> 00:18:53,391
in the Koch Curve were precisely
what was needed
387
00:18:53,399 --> 00:18:56,198
to model coastlines.
388
00:18:56,202 --> 00:18:59,194
He wrote a very famous article
in Science Magazine called
389
00:18:59,205 --> 00:19:00,366
"How Long Is the Coastline
of Britain?"
390
00:19:00,373 --> 00:19:05,334
NARRATOR: A coastline, in geometric
terms, said Mandelbrot, is a fractal.
391
00:19:05,345 --> 00:19:08,303
And though he knew
he couldn't measure its length,
392
00:19:08,314 --> 00:19:13,411
he suspected he could measure
something else: its roughness.
393
00:19:13,419 --> 00:19:18,323
To do that required rethinking
one of the basic concepts
394
00:19:18,324 --> 00:19:20,418
in math: dimension.
395
00:19:20,426 --> 00:19:23,350
What we would think of
as normal geometry--
396
00:19:23,363 --> 00:19:24,455
one dimension
is the straight line,
397
00:19:24,464 --> 00:19:28,332
two dimensions is, say,
the box that has surface area.
398
00:19:28,334 --> 00:19:31,258
NARRATOR:
And three dimensions is a cube.
399
00:19:31,271 --> 00:19:33,296
But could something
have a dimension
400
00:19:33,306 --> 00:19:36,435
somewhere
in between, say, two and three?
401
00:19:36,442 --> 00:19:41,369
Mandelbrot said, yes,
fractals do.
402
00:19:41,381 --> 00:19:44,282
And the rougher they are,
403
00:19:44,284 --> 00:19:46,378
the higher
their fractal dimension.
404
00:19:46,386 --> 00:19:48,343
DEVLIN:
There are all of these
405
00:19:48,354 --> 00:19:50,413
technical terms,
like fractal dimension,
406
00:19:50,423 --> 00:19:52,289
and self-similarity,
407
00:19:52,292 --> 00:19:55,387
but those are the nuts and bolts
of the mathematics itself.
408
00:19:55,395 --> 00:20:00,299
What that fractal geometry does
is give us a way of looking at--
409
00:20:00,300 --> 00:20:03,326
in a way
that's extremely precise--
410
00:20:03,336 --> 00:20:10,402
the world in which we live,
in particular, the living world.
411
00:20:10,410 --> 00:20:13,311
NARRATOR: Mandelbrot's
fresh ways of thinking
412
00:20:13,313 --> 00:20:16,339
were made possible
by his enthusiastic embrace
413
00:20:16,349 --> 00:20:17,373
of new technology.
414
00:20:17,383 --> 00:20:22,287
Computers made it easy
for Mandelbrot to do iteration--
415
00:20:22,288 --> 00:20:24,279
the endlessly repeating cycles
of calculation
416
00:20:24,290 --> 00:20:27,419
that were demanded
by the mathematical monsters.
417
00:20:27,427 --> 00:20:31,455
MANDELBROT: The computer
was totally essential.
418
00:20:31,464 --> 00:20:34,229
Otherwise, it would have taken
a very big, long effort.
419
00:20:34,234 --> 00:20:39,365
NARRATOR: Mandelbrot decided to zero
in on yet another of the monsters--
420
00:20:39,372 --> 00:20:42,273
a problem introduced
during World War I
421
00:20:42,275 --> 00:20:47,270
by a young French mathematician
named Gaston Julia.
422
00:20:47,280 --> 00:20:49,442
DEVLIN:
Gaston Julia--
423
00:20:49,449 --> 00:20:53,317
he was actually looking at
what happens when you take
424
00:20:53,319 --> 00:20:54,252
a simple equation
425
00:20:54,254 --> 00:20:56,245
and you iterate it
through a feedback loop.
426
00:20:56,256 --> 00:20:57,280
That means you take a number,
427
00:20:57,290 --> 00:21:00,248
you plug it into the formula,
you get a number out.
428
00:21:00,260 --> 00:21:02,388
You take that number,
back to the beginning,
429
00:21:02,395 --> 00:21:03,351
and you feed it into
430
00:21:03,363 --> 00:21:05,354
the same formula,
get another number out.
431
00:21:05,365 --> 00:21:08,323
And you keep iterating that
over and over again.
432
00:21:08,334 --> 00:21:10,291
And the question is,
what happens
433
00:21:10,303 --> 00:21:12,397
when you iterate it
lots of times.
434
00:21:12,405 --> 00:21:18,276
NARRATOR: The series of numbers you
get is called a set-- the Julia set.
435
00:21:18,278 --> 00:21:20,303
But working by hand,
436
00:21:20,313 --> 00:21:21,337
you could never really know
437
00:21:21,347 --> 00:21:23,338
what the complete set
looked like.
438
00:21:23,349 --> 00:21:25,408
ABRAHAM:
There were attempts to draw it.
439
00:21:25,418 --> 00:21:27,409
Doing a bunch of arithmetic
by hand
440
00:21:27,420 --> 00:21:29,377
and putting a point
on graph paper.
441
00:21:29,389 --> 00:21:32,347
You would have to feed it back
hundreds, thousands,
442
00:21:32,358 --> 00:21:34,224
millions of times.
443
00:21:34,227 --> 00:21:37,322
The development of that new kind
of mathematics had to wait
444
00:21:37,330 --> 00:21:41,233
until fast computers
were invented.
445
00:21:41,234 --> 00:21:43,396
NARRATOR:
At IBM, Mandelbrot did something
446
00:21:43,403 --> 00:21:45,428
Julia could never do:
447
00:21:45,438 --> 00:21:50,308
use a computer to run
the equations millions of times.
448
00:21:50,310 --> 00:21:51,471
He then turned the numbers
449
00:21:51,477 --> 00:21:55,436
from his Julia sets
into points on a graph.
450
00:21:55,448 --> 00:22:01,342
MANDELBROT: My first step
was to just draw mindlessly
451
00:22:01,354 --> 00:22:03,311
a large number of Julia sets.
452
00:22:03,323 --> 00:22:06,281
Not one picture,
hundreds of pictures.
453
00:22:06,292 --> 00:22:10,320
NARRATOR: Those images led
Mandelbrot to a breakthrough.
454
00:22:10,330 --> 00:22:14,233
In 1980, he created
an equation of his own,
455
00:22:14,233 --> 00:22:17,294
one that combined
all of the Julia sets
456
00:22:17,303 --> 00:22:19,260
into a single image.
457
00:22:19,272 --> 00:22:21,434
When Mandelbrot
iterated his equation,
458
00:22:21,441 --> 00:22:24,240
he got his own set of numbers.
459
00:22:24,243 --> 00:22:27,338
Graphed on a computer,
it was a kind of road map
460
00:22:27,347 --> 00:22:30,442
of all the Julia sets
and quickly became famous
461
00:22:30,450 --> 00:22:34,353
as the emblem
of fractal geometry...
462
00:22:34,354 --> 00:22:38,279
the Mandelbrot set.
463
00:22:38,291 --> 00:22:41,386
They intersect at certain areas,
and it's got like a, you know...
464
00:22:41,394 --> 00:22:44,352
And they have little curlicues
built into them.
465
00:22:44,364 --> 00:22:46,389
Black beetle-like thing.
466
00:22:46,399 --> 00:22:48,333
Crawling across the floor.
467
00:22:48,334 --> 00:22:49,324
Seahorses.
Dragons.
468
00:22:49,335 --> 00:22:51,394
Something similar
to my hair, actually.
469
00:22:51,404 --> 00:22:54,396
(laughing)
470
00:22:54,407 --> 00:22:56,341
NARRATOR:
With this mysterious image,
471
00:22:56,342 --> 00:22:58,436
Mandelbrot was issuing
a bold challenge
472
00:22:58,444 --> 00:23:04,315
to long-standing ideas
about the limits of mathematics.
473
00:23:04,317 --> 00:23:07,309
The blinders came off,
and people could see forms
474
00:23:07,320 --> 00:23:12,486
that were always there,
but formerly were invisible.
475
00:23:12,492 --> 00:23:16,360
DEVLIN: The Mandelbrot
set was a great example
476
00:23:16,362 --> 00:23:19,388
of what you could do
in fractal geometry,
477
00:23:19,399 --> 00:23:22,266
just as the archetypical example
478
00:23:22,268 --> 00:23:30,301
of classical geometry
is the circle.
479
00:23:30,309 --> 00:23:33,335
ABRAHAM: When you zoom in, you
see them coming up again,
480
00:23:33,346 --> 00:23:34,472
so you see self-similarity.
481
00:23:34,480 --> 00:23:37,381
You see, by zooming in,
you zoom, zoom, zoom,
482
00:23:37,383 --> 00:23:38,407
you're zooming in,
you're zooming in,
483
00:23:38,418 --> 00:23:41,319
and pop, suddenly it seems
like you're exactly
484
00:23:41,320 --> 00:23:42,412
where you were before,
but you're not.
485
00:23:42,422 --> 00:23:45,289
It's just that way down there,
it has the same kind
486
00:23:45,291 --> 00:23:50,291
of structure as way up here,
and the sameness can be grokked.
487
00:23:59,472 --> 00:24:01,429
NARRATOR:
Mandelbrot's mesmerizing images
488
00:24:01,441 --> 00:24:05,435
launched a fad
in the world of popular culture.
489
00:24:05,445 --> 00:24:07,345
MANDELBROT:
Suddenly, this thing caught
490
00:24:07,346 --> 00:24:09,508
like... like a bush fire.
491
00:24:09,515 --> 00:24:14,515
Everybody wanted to have it.
492
00:24:21,394 --> 00:24:24,420
DEVLIN: I thought, this is
something big going on.
493
00:24:24,430 --> 00:24:30,358
This was a cultural event
of great proportions.
494
00:24:30,369 --> 00:24:33,327
NARRATOR:
In the late 1970s, Jhane Barnes
495
00:24:33,339 --> 00:24:37,435
had just launched a business
designing men's clothing.
496
00:24:37,443 --> 00:24:39,434
JHANE BARNES: When I started
my business in '76,
497
00:24:39,445 --> 00:24:43,313
I was doing fabrics
the old-fashioned way,
498
00:24:43,316 --> 00:24:44,442
just on graph paper;
499
00:24:44,450 --> 00:24:47,283
weaving them
on a little handloom.
500
00:24:47,286 --> 00:24:49,380
NARRATOR: But then, she
discovered fractals
501
00:24:49,388 --> 00:24:52,414
and realized that the
simple rules that made them
502
00:24:52,425 --> 00:24:55,417
could be used to create
intricate clothing designs.
503
00:24:55,428 --> 00:24:59,387
BARNES: I thought, this is amazing,
so that very simple concept,
504
00:24:59,398 --> 00:25:02,322
I said, "Oh, I can make designs
with that."
505
00:25:02,335 --> 00:25:05,293
But in the '80s,
I really didn't know
506
00:25:05,304 --> 00:25:07,466
how to design a fractal,
because there wasn't software.
507
00:25:07,473 --> 00:25:09,373
NARRATOR:
So Barnes got help
508
00:25:09,375 --> 00:25:10,467
from two people who knew a lot
509
00:25:10,476 --> 00:25:14,310
about math and computers:
Bill Jones
510
00:25:14,313 --> 00:25:15,508
and Dana Cartwright.
511
00:25:15,515 --> 00:25:19,315
BARNES: I had Dana and Bill
writing my software for me.
512
00:25:19,318 --> 00:25:23,380
They said, "Oh, your work
is very mathematical."
513
00:25:23,389 --> 00:25:24,379
And I was like, "It is?
514
00:25:24,390 --> 00:25:26,290
That's my weakest subject
in school."
515
00:25:26,292 --> 00:25:29,387
We had a physicist
and a mathematician
516
00:25:29,395 --> 00:25:30,487
and a textile designer.
517
00:25:30,496 --> 00:25:32,453
BARNES: We had so much to
learn from each other.
518
00:25:32,465 --> 00:25:36,424
DANA CARTWRIGHT: I did not know
what a warp and a weft is.
519
00:25:36,435 --> 00:25:37,459
You know, Jhane...
520
00:25:37,470 --> 00:25:40,337
her ability with numbers
is fairly restricted,
521
00:25:40,339 --> 00:25:42,296
if I can put that politely.
522
00:25:42,308 --> 00:25:43,469
All, um,
the parameters here...
523
00:25:43,476 --> 00:25:46,502
BARNES: There was a way we
were going to communicate.
524
00:25:46,512 --> 00:25:48,378
We were going
to get together somehow,
525
00:25:48,381 --> 00:25:49,507
and it really did happen
pretty quickly.
526
00:25:49,515 --> 00:25:53,509
The general fashion press
thought "Jhane's a little nuts."
527
00:25:53,519 --> 00:25:57,285
They started calling me
the Fashion Nerd,
528
00:25:57,290 --> 00:25:58,314
you know, but that was okay.
529
00:25:58,324 --> 00:26:00,418
That was okay with me
because I was learning a lot.
530
00:26:00,426 --> 00:26:05,387
This was fun
and very, very inspirational.
531
00:26:05,398 --> 00:26:11,394
I'm getting things that wouldn't
be possible by hand.
532
00:26:11,404 --> 00:26:15,466
You know, sometimes when I think
about things in my head
533
00:26:15,474 --> 00:26:19,377
and I say, "You know,
I just saw light coming
534
00:26:19,378 --> 00:26:20,470
"through that screen door,
535
00:26:20,479 --> 00:26:22,504
"and look at
the moir,ing effects
536
00:26:22,515 --> 00:26:25,382
that are happening
on the ground."
537
00:26:25,384 --> 00:26:26,408
Can I go draw that?
538
00:26:26,419 --> 00:26:31,289
No way, but I can describe that
to my mathematician.
539
00:26:31,290 --> 00:26:32,280
This kind of
reminds me of...
540
00:26:32,291 --> 00:26:37,286
He sends me back the generator,
all ready for me to try,
541
00:26:37,296 --> 00:26:39,355
and I sit down
at the computer and say,
542
00:26:39,365 --> 00:26:40,491
"well let's see
what it's doing."
543
00:26:40,499 --> 00:26:44,299
And I have parameters
that I can control.
544
00:26:44,303 --> 00:26:46,397
And I keep pushing, and I go,
545
00:26:46,405 --> 00:26:50,330
"well, this is not
what I expected at all...
546
00:26:50,343 --> 00:26:56,373
um, at all, but it's cool."
547
00:26:56,382 --> 00:26:57,372
(weapons blasting)
548
00:26:57,383 --> 00:26:59,442
OBI-WAN KENOBI:
Use the Force, Luke.
549
00:26:59,452 --> 00:27:01,352
(weapons blasting)
550
00:27:01,354 --> 00:27:04,449
NARRATOR: The same kinds of
fractal design principles
551
00:27:04,457 --> 00:27:08,485
have completely transformed
the magic of special effects.
552
00:27:08,494 --> 00:27:14,388
DAN PIPONI: This is a key moment
from Star Wars: Episode III,
553
00:27:14,400 --> 00:27:16,459
where our two heroes
have run out
554
00:27:16,469 --> 00:27:19,461
onto the end
of this giant mechanical arm
555
00:27:19,472 --> 00:27:24,376
and the lava splashes down
onto the arm.
556
00:27:24,377 --> 00:27:25,503
My starting point here
is to actually take
557
00:27:25,511 --> 00:27:29,470
the three-dimensional model
and take essentially a jet
558
00:27:29,482 --> 00:27:33,350
and just shoot lava
up into the air.
559
00:27:33,352 --> 00:27:34,444
This looks kind of boring.
560
00:27:34,453 --> 00:27:36,319
It's doing roughly
the right thing,
561
00:27:36,322 --> 00:27:39,519
but the motion has no kind
of visual interest to it.
562
00:27:39,525 --> 00:27:41,391
Let's look at what happens here
563
00:27:41,394 --> 00:27:44,352
when I add the fractal swirl
to it.
564
00:27:44,363 --> 00:27:46,559
Where this becomes fractal is,
565
00:27:46,565 --> 00:27:48,499
we take that same swirl pattern,
566
00:27:48,501 --> 00:27:51,425
we shrink it down
and reapply it.
567
00:27:51,437 --> 00:27:55,305
We take that, we shrink it
down again, we reapply it.
568
00:27:55,307 --> 00:27:57,401
We shrink it down again,
we reapply it.
569
00:27:57,410 --> 00:27:59,504
And from here on,
it's just a case
570
00:27:59,512 --> 00:28:02,470
of layering up more
and more and more.
571
00:28:02,481 --> 00:28:03,505
I've used the same technique
572
00:28:03,516 --> 00:28:06,474
to create these additional
lava streams.
573
00:28:06,485 --> 00:28:09,318
I then do it again here
574
00:28:09,321 --> 00:28:12,347
to get some just red hot embers.
575
00:28:12,358 --> 00:28:14,554
Then, we take all of those
layers, and we add them up,
576
00:28:14,560 --> 00:28:17,518
and we get the final
composite image.
577
00:28:17,530 --> 00:28:19,487
My hero lava in the foreground,
578
00:28:19,498 --> 00:28:22,365
the extra lava
in the background.
579
00:28:22,368 --> 00:28:29,365
The embers, sparks,
steam, smoke.
580
00:28:29,375 --> 00:28:34,404
(grunting)
581
00:28:34,413 --> 00:28:36,438
NARRATOR: Designers and
artists the world over
582
00:28:36,449 --> 00:28:39,544
have embraced the visual
potential of fractals,
583
00:28:39,552 --> 00:28:43,420
but when the Mandelbrot set
was first published,
584
00:28:43,422 --> 00:28:45,481
mathematicians,
for the most part,
585
00:28:45,491 --> 00:28:47,448
reacted with scorn.
586
00:28:47,460 --> 00:28:50,452
ABRAHAM: In the
Mathematical Intelligencer,
587
00:28:50,463 --> 00:28:53,558
which is a gossip sheet for
professional mathematicians,
588
00:28:53,566 --> 00:28:55,523
there were article after article
589
00:28:55,534 --> 00:28:57,525
saying he wasn't
a mathematician;
590
00:28:57,536 --> 00:29:00,460
he was a bad mathematician;
it's not mathematics;
591
00:29:00,473 --> 00:29:02,407
fractal geometry is worthless.
592
00:29:02,408 --> 00:29:05,434
The eye had been banished
out of science.
593
00:29:05,444 --> 00:29:08,345
The eye had been excommunicated.
594
00:29:08,347 --> 00:29:13,342
ABRAHAM: His colleagues,
especially the really good ones,
595
00:29:13,352 --> 00:29:15,514
pure mathematicians
that he respected,
596
00:29:15,521 --> 00:29:17,319
they turned against him.
597
00:29:17,323 --> 00:29:19,451
Because, see now, you get used
to the world
598
00:29:19,458 --> 00:29:21,392
that you've created
and that you live in,
599
00:29:21,393 --> 00:29:22,485
and mathematicians
had become very used
600
00:29:22,495 --> 00:29:26,420
to this world of smooth curves
that they could do things with.
601
00:29:26,432 --> 00:29:31,495
ABRAHAM: They were clinging
to the old paradigm
602
00:29:31,504 --> 00:29:34,405
when Mandelbrot and a few people
603
00:29:34,406 --> 00:29:40,470
were way out there
bringing in the new paradigm.
604
00:29:40,479 --> 00:29:44,541
And he used to call me up
on the telephone late at night,
605
00:29:44,550 --> 00:29:48,384
because he was bothered,
and we'd talk about it.
606
00:29:48,387 --> 00:29:49,513
Mandelbrot was saying,
607
00:29:49,522 --> 00:29:52,480
"This is a branch of geometry
just like Euclid."
608
00:29:52,491 --> 00:29:53,481
Well, that offended them.
609
00:29:53,492 --> 00:29:56,393
They said, "No, this
is an artifact
610
00:29:56,395 --> 00:30:01,390
of your stupid
computing machine."
611
00:30:01,400 --> 00:30:04,392
MANDELBROT: I know very well
that there is this line
612
00:30:04,403 --> 00:30:05,495
that fractals
are pretty pictures,
613
00:30:05,504 --> 00:30:06,494
but are pretty useless.
614
00:30:06,505 --> 00:30:08,371
Well, it's a pretty jingle,
615
00:30:08,374 --> 00:30:10,468
but it's completely ridiculous.
616
00:30:10,476 --> 00:30:12,570
NARRATOR: Mandelbrot
replied to his critics
617
00:30:12,578 --> 00:30:17,414
with his new book:
The Fractal Geometry of Nature.
618
00:30:17,416 --> 00:30:18,577
It was filled with examples
619
00:30:18,584 --> 00:30:21,542
of how his ideas
could be useful to science.
620
00:30:21,554 --> 00:30:24,455
Mandelbrot argued
that with fractals,
621
00:30:24,456 --> 00:30:27,448
he could precisely
measure natural shapes
622
00:30:27,459 --> 00:30:30,417
and make calculations
that could be applied
623
00:30:30,429 --> 00:30:32,523
to all kinds of formations,
624
00:30:32,531 --> 00:30:35,557
from the drainage patterns
of rivers
625
00:30:35,568 --> 00:30:37,525
to the movements of clouds.
626
00:30:37,536 --> 00:30:41,370
DEVLIN: So this domain of
growing, living systems,
627
00:30:41,373 --> 00:30:42,568
which I, along with most other
mathematicians,
628
00:30:42,575 --> 00:30:45,476
had always regarded
as pretty well off-limits
629
00:30:45,477 --> 00:30:48,435
for mathematics and certainly
off-limits for geometry,
630
00:30:48,447 --> 00:30:50,347
suddenly was center stage.
631
00:30:50,349 --> 00:30:52,545
It was Mandelbrot's book
that convinced us
632
00:30:52,551 --> 00:30:54,542
that this wasn't just artwork.
633
00:30:54,553 --> 00:30:57,454
This was new science
in the making.
634
00:30:57,456 --> 00:31:00,414
This was a completely new way
of looking
635
00:31:00,426 --> 00:31:01,552
at the world in which we live
636
00:31:01,560 --> 00:31:04,518
that allowed us not just
to look at it,
637
00:31:04,530 --> 00:31:05,554
not just to measure it,
638
00:31:05,564 --> 00:31:08,556
but to do mathematics
and thereby understand it
639
00:31:08,567 --> 00:31:12,492
in a deeper way
than we had before.
640
00:31:12,504 --> 00:31:15,462
As someone who's been working
with fractals for 20 years,
641
00:31:15,474 --> 00:31:17,465
I'm not going to tell you
fractals are cool.
642
00:31:17,476 --> 00:31:20,434
I'm going to tell you fractals
are useful,
643
00:31:20,446 --> 00:31:22,574
and that's what's important
to me.
644
00:31:22,581 --> 00:31:26,484
NARRATOR: In the 1990s, a
Boston radio astronomer
645
00:31:26,485 --> 00:31:29,511
named Nathan Cohen
used fractal mathematics
646
00:31:29,521 --> 00:31:31,512
to make a technological
breakthrough
647
00:31:31,523 --> 00:31:33,514
in electronic communication.
648
00:31:33,525 --> 00:31:34,458
:( beeping)
649
00:31:34,460 --> 00:31:38,419
Cohen had a hobby:
he was a ham radio operator,
650
00:31:38,430 --> 00:31:39,522
but his landlord had a rule
651
00:31:39,531 --> 00:31:42,523
against rigging antennas
on the building.
652
00:31:42,534 --> 00:31:45,492
NATHAN COHEN: I was at an
astronomy conference in Hungary,
653
00:31:45,504 --> 00:31:47,438
and Dr. Mandelbrot
was giving a talk
654
00:31:47,439 --> 00:31:50,363
about the large-scale structure
of the universe
655
00:31:50,376 --> 00:31:55,542
and reporting how using fractals
is a very good way
656
00:31:55,547 --> 00:31:57,379
of understanding
that kind of structure,
657
00:31:57,383 --> 00:32:02,344
which really wowed
the entire group of astronomers.
658
00:32:02,354 --> 00:32:03,549
He showed several
different fractals
659
00:32:03,555 --> 00:32:06,547
that I, in my own mind,
looked at and said,
660
00:32:06,558 --> 00:32:07,514
"Oh, wouldn't it be funny
661
00:32:07,526 --> 00:32:10,393
"if you made an antenna
out of that shape?
662
00:32:10,396 --> 00:32:11,522
I wonder what it would do."
663
00:32:11,530 --> 00:32:13,589
NARRATOR: One of the
first designs he tried
664
00:32:13,599 --> 00:32:17,524
was inspired by one of the
19th century "monsters":
665
00:32:17,536 --> 00:32:20,437
the snowflake
of Helge von Koch.
666
00:32:20,439 --> 00:32:22,464
I thought back
to the lecture and said,
667
00:32:22,474 --> 00:32:23,600
"well, I've got a piece
of wire.
668
00:32:23,609 --> 00:32:27,603
What happens if I bend it?"
669
00:32:27,613 --> 00:32:30,412
After I bent the wire,
I hooked it up
670
00:32:30,416 --> 00:32:31,542
to the cable and my ham radio,
671
00:32:31,550 --> 00:32:34,451
and I was quite surprised
to see that it worked
672
00:32:34,453 --> 00:32:36,410
the first time out of the box.
673
00:32:36,422 --> 00:32:37,514
It worked very well,
and I discovered
674
00:32:37,523 --> 00:32:39,514
that, much of a surprise to me,
675
00:32:39,525 --> 00:32:41,425
that I could actually
make the antenna
676
00:32:41,427 --> 00:32:45,421
much smaller
using the fractal design,
677
00:32:45,431 --> 00:32:47,388
so it was, frankly,
an interesting way
678
00:32:47,399 --> 00:32:51,427
to beat a bad rap
with the landlord.
679
00:32:51,437 --> 00:32:54,532
NARRATOR: Cohen's experiments soon
led him to another discovery.
680
00:32:54,540 --> 00:32:58,534
Using a fractal design not only
made antennas smaller,
681
00:32:58,544 --> 00:33:00,501
but enabled them to receive
682
00:33:00,512 --> 00:33:03,436
a much wider range
of frequencies.
683
00:33:03,449 --> 00:33:06,441
COHEN: Using fractals,
experimentally I came up
684
00:33:06,452 --> 00:33:07,613
with a very wideband antenna.
685
00:33:07,619 --> 00:33:08,609
And then I worked backwards
686
00:33:08,620 --> 00:33:10,577
and said,
"Why is it working this way?
687
00:33:10,589 --> 00:33:13,581
"What is it about nature
that requires you
688
00:33:13,592 --> 00:33:16,493
to use the fractal
to get there?"
689
00:33:16,495 --> 00:33:18,395
The result of that work was
690
00:33:18,397 --> 00:33:20,456
a mathematical theorem
that showed
691
00:33:20,466 --> 00:33:22,366
if you want to get something
692
00:33:22,368 --> 00:33:23,563
that works as an antenna
693
00:33:23,569 --> 00:33:26,402
over a very wide range
of frequencies,
694
00:33:26,405 --> 00:33:28,396
you need to have
self-similarity.
695
00:33:28,407 --> 00:33:32,401
It has to be fractal
in its shape to make it work.
696
00:33:32,411 --> 00:33:35,437
Now, that was an exact solution.
It wasn't like,
697
00:33:35,447 --> 00:33:36,471
"Oh, here's a way of doing it
698
00:33:36,482 --> 00:33:38,473
and there's a lot
of other ways of doing it."
699
00:33:38,484 --> 00:33:39,610
It turned out mathematically,
700
00:33:39,618 --> 00:33:42,542
we were able to demonstrate
that was the only technique
701
00:33:42,554 --> 00:33:44,386
you would use to get there.
702
00:33:44,390 --> 00:33:45,380
(cell phone ringing)
703
00:33:45,391 --> 00:33:46,449
NARRATOR:
Cohen made his discovery
704
00:33:46,458 --> 00:33:49,587
at a time when cell phone
companies were facing a problem.
705
00:33:49,595 --> 00:33:53,554
They were offering new features
to their customers,
706
00:33:53,565 --> 00:33:56,557
like Bluetooth,
walkie-talkie, and Wi-Fi,
707
00:33:56,568 --> 00:34:00,471
but each of them ran
on a separate frequency.
708
00:34:00,472 --> 00:34:02,372
COHEN:
You need to be able to use all
709
00:34:02,374 --> 00:34:03,535
those different frequencies
and have access to them
710
00:34:03,542 --> 00:34:08,446
without ten stubby antennas
sticking out at the same time.
711
00:34:08,447 --> 00:34:09,403
The alternative option is
712
00:34:09,415 --> 00:34:11,406
you can let your cell phone
look like a porcupine.
713
00:34:11,417 --> 00:34:16,412
But most people don't want
to carry around a porcupine.
714
00:34:16,422 --> 00:34:18,584
NARRATOR:
Today, fractal antennas are used
715
00:34:18,590 --> 00:34:20,547
in tens of millions
of cell phones,
716
00:34:20,559 --> 00:34:24,587
and other wireless communication
devices all over the world.
717
00:34:24,596 --> 00:34:29,466
COHEN: We're going to see over the
next ten to 15 to 20 years that
718
00:34:29,468 --> 00:34:30,594
you're going
to have to use fractals
719
00:34:30,602 --> 00:34:33,469
because it's the only way
to get, uh, cheaper costs
720
00:34:33,472 --> 00:34:35,600
and smaller size
for all the complex
721
00:34:35,607 --> 00:34:40,534
telecommunication needs
we're having.
722
00:34:40,546 --> 00:34:42,480
MANDELBROT:
Once you realize that
723
00:34:42,481 --> 00:34:46,475
a shrewd engineer would use
fractals in many, many contexts,
724
00:34:46,485 --> 00:34:50,615
you better understand
why nature, which is shrewder,
725
00:34:50,622 --> 00:34:52,613
uses them in its ways.
726
00:34:52,624 --> 00:34:54,558
They're all over in biology.
727
00:34:54,560 --> 00:34:55,550
They're solutions
728
00:34:55,561 --> 00:34:58,462
that natural selection
has come up with
729
00:34:58,464 --> 00:35:01,593
over and over and over
and over again.
730
00:35:01,600 --> 00:35:03,466
NARRATOR:
One powerful example:
731
00:35:03,469 --> 00:35:06,495
the rhythms of the heart.
(beating)
732
00:35:06,505 --> 00:35:09,531
Something that Boston
cardiologist Ary Goldberger
733
00:35:09,541 --> 00:35:13,409
has been studying
his entire professional life.
734
00:35:13,412 --> 00:35:15,437
ARY GOLDBERGER: The notion
of sort of the human body
735
00:35:15,447 --> 00:35:17,575
as a machine goes
back through the tradition
736
00:35:17,583 --> 00:35:19,540
of Newton
and the machinelike universe.
737
00:35:19,551 --> 00:35:20,609
So somehow we're,
we're machines,
738
00:35:20,619 --> 00:35:23,577
we're mechanisms; the heartbeat
is this timekeeper.
739
00:35:23,589 --> 00:35:26,490
Galileo was reported
to have used
740
00:35:26,492 --> 00:35:28,392
his pulse to time
741
00:35:28,393 --> 00:35:30,384
the swinging
of a pendular motion.
742
00:35:30,395 --> 00:35:34,525
So that all fit in with the idea
that a normal heartbeat
743
00:35:34,533 --> 00:35:35,625
is like a metronome.
744
00:35:35,634 --> 00:35:38,558
NARRATOR: But when Goldberger
and his colleagues
745
00:35:38,570 --> 00:35:41,494
analyzed data
from thousands of people,
746
00:35:41,507 --> 00:35:44,499
they found the old theory
was wrong.
747
00:35:44,510 --> 00:35:45,500
MADALENA DAMASIO COSTA:
This is, um,
748
00:35:45,511 --> 00:35:49,573
where I show the heartbeat time
series of a healthy subject.
749
00:35:49,581 --> 00:35:50,639
And as you can see,
750
00:35:50,649 --> 00:35:54,483
the heartbeat
is not constant over time.
751
00:35:54,486 --> 00:35:56,477
It fluctuates,
and it fluctuates a lot.
752
00:35:56,488 --> 00:35:57,614
For example, in this case
it fluctuates between
753
00:35:57,623 --> 00:36:02,550
60 beats per minute
and 120 beats per minute.
754
00:36:02,561 --> 00:36:05,428
NARRATOR: The patterns looked
familiar to Goldberger,
755
00:36:05,430 --> 00:36:08,491
who happened to have read
Benoit Mandelbrot's book.
756
00:36:08,500 --> 00:36:11,424
GOLDBERGER: When you actually
plotted out the intervals
757
00:36:11,436 --> 00:36:14,633
between heartbeats,
what you saw was very close
758
00:36:14,640 --> 00:36:18,543
to the rough edges
of the mountain ranges
759
00:36:18,544 --> 00:36:21,570
that were in Mandelbrot's book.
760
00:36:21,580 --> 00:36:24,481
You blow them up,
uh, expand them,
761
00:36:24,483 --> 00:36:26,542
you actually see that
there are more of these
762
00:36:26,552 --> 00:36:27,610
wrinkles upon wrinkles.
763
00:36:27,619 --> 00:36:30,452
The healthy heartbeat,
it turned out,
764
00:36:30,455 --> 00:36:33,516
had this fractal architecture.
765
00:36:33,525 --> 00:36:35,425
People said,
"This isn't cardiology.
766
00:36:35,427 --> 00:36:37,589
Do cardiology if you want
to get funded."
767
00:36:37,596 --> 00:36:41,430
But it turns out
it is cardiology.
768
00:36:41,433 --> 00:36:44,562
NARRATOR: Goldberger found
that the healthy heartbeat
769
00:36:44,570 --> 00:36:46,561
has a distinctive
fractal pattern,
770
00:36:46,572 --> 00:36:50,497
a signature that one day
may help cardiologists
771
00:36:50,509 --> 00:36:55,538
spot heart problems sooner.
772
00:36:55,547 --> 00:36:58,414
Please look around
the screen for me.
773
00:36:58,417 --> 00:37:00,613
All right, Cooper,
we're going to do
774
00:37:00,619 --> 00:37:01,609
the calibration.
775
00:37:01,620 --> 00:37:04,487
NARRATOR:
At the University of Oregon,
776
00:37:04,489 --> 00:37:07,481
Richard Taylor is using
fractals to reveal
777
00:37:07,492 --> 00:37:11,554
the secrets of another part
of the body: the eye.
778
00:37:11,563 --> 00:37:13,497
TAYLOR:
What we want to do
779
00:37:13,498 --> 00:37:15,523
is see what is that eye doing
780
00:37:15,534 --> 00:37:19,596
that allows it to absorb
so much visual information.
781
00:37:19,605 --> 00:37:23,508
And so that's what led us
into the eye trajectories.
782
00:37:23,508 --> 00:37:26,603
Under the monitor
is a little infrared camera,
783
00:37:26,612 --> 00:37:28,671
which will actually monitor
784
00:37:28,680 --> 00:37:30,444
where the eye is looking.
785
00:37:30,449 --> 00:37:32,577
And it actually
records that data.
786
00:37:32,584 --> 00:37:35,508
And so what we get out
is a trajectory
787
00:37:35,520 --> 00:37:37,511
of where the eye
has been looking.
788
00:37:37,522 --> 00:37:39,616
Oh, it's interesting
how they go around
789
00:37:39,625 --> 00:37:40,615
in the patterns...
790
00:37:40,626 --> 00:37:43,527
TAYLOR: And so the computer
will get out this graph,
791
00:37:43,528 --> 00:37:46,486
and it will look, you know,
have all of these various,
792
00:37:46,498 --> 00:37:48,489
uh, little structure in it.
793
00:37:48,500 --> 00:37:50,628
And it's that pattern
that we zoom in--
794
00:37:50,636 --> 00:37:52,627
we tell the computer
to zoom in on--
795
00:37:52,638 --> 00:37:54,663
and, and see
the fractal dimension.
796
00:37:54,673 --> 00:38:00,442
NARRATOR: The tests show that the
eye does not always look at things
797
00:38:00,445 --> 00:38:02,504
in an orderly or smooth way.
798
00:38:02,514 --> 00:38:05,438
If we could understand more
about how the eye
799
00:38:05,450 --> 00:38:07,509
takes in information,
we could do
800
00:38:07,519 --> 00:38:09,510
a better job
of designing the things
801
00:38:09,521 --> 00:38:11,649
that we really need to see.
802
00:38:11,657 --> 00:38:13,421
TAYLOR:
A traffic light.
803
00:38:13,425 --> 00:38:14,620
You're looking
at the traffic light.
804
00:38:14,626 --> 00:38:16,458
You've got traffic.
805
00:38:16,461 --> 00:38:16,620
You've got pedestrians.
806
00:38:16,628 --> 00:38:19,461
Your eye is looking
all over the place
807
00:38:19,464 --> 00:38:22,525
trying to assess
all of this information.
808
00:38:22,534 --> 00:38:26,596
People design aircraft cockpits,
rows of dials
809
00:38:26,605 --> 00:38:27,595
and things like that.
810
00:38:27,606 --> 00:38:30,564
If your eye is darting around
811
00:38:30,575 --> 00:38:32,634
based on a fractal geometry,
though,
812
00:38:32,644 --> 00:38:34,544
maybe that's not the best way.
813
00:38:34,546 --> 00:38:39,473
Maybe you don't want
these things in a simple row.
814
00:38:39,484 --> 00:38:41,475
We're trying to work out
the natural way
815
00:38:41,486 --> 00:38:44,444
that the eye wants
to absorb the information.
816
00:38:44,456 --> 00:38:45,514
Is it going to be similar
817
00:38:45,524 --> 00:38:48,585
to a lot of these other
subconscious processes?
818
00:38:48,593 --> 00:38:51,494
Body motion,
when you're balancing,
819
00:38:51,496 --> 00:38:53,487
what are you
actually doing there?
820
00:38:53,498 --> 00:38:55,626
It's something subconscious,
and it works.
821
00:38:55,634 --> 00:38:59,537
And you're stringing
together big sways
822
00:38:59,538 --> 00:39:00,596
and small sways
and smaller sways.
823
00:39:00,605 --> 00:39:03,438
Could those all
be connected together
824
00:39:03,442 --> 00:39:07,538
to actually be doing
a fractal pattern there?
825
00:39:07,546 --> 00:39:11,449
More and more
physiological processes
826
00:39:11,450 --> 00:39:14,613
have been found to be fractal.
827
00:39:14,619 --> 00:39:17,611
NARRATOR: Not everyone
in science is convinced
828
00:39:17,622 --> 00:39:21,650
of fractal geometry's potential
for delivering new knowledge.
829
00:39:21,660 --> 00:39:24,527
Skeptics argue
that it's done little
830
00:39:24,529 --> 00:39:26,554
to advance mathematical theory.
831
00:39:26,565 --> 00:39:29,694
But in Toronto,
biophysicist Peter Burns
832
00:39:29,701 --> 00:39:31,499
and his colleagues
833
00:39:31,503 --> 00:39:34,666
see fractals as a practical
tool, a way to develop
834
00:39:34,673 --> 00:39:37,574
mathematical models
that might help
835
00:39:37,576 --> 00:39:39,635
in diagnosing cancer earlier.
836
00:39:39,644 --> 00:39:43,569
Detecting very small tumors
is one of the big challenges
837
00:39:43,582 --> 00:39:46,483
in medical imaging.
838
00:39:46,485 --> 00:39:47,611
NARRATOR:
Burns knew that one
839
00:39:47,619 --> 00:39:50,452
early sign of cancer
is particularly
840
00:39:50,455 --> 00:39:52,583
difficult to see:
a network
841
00:39:52,591 --> 00:39:55,686
of tiny blood vessels
that forms with the tumor.
842
00:39:55,694 --> 00:39:57,685
Conventional imaging techniques,
843
00:39:57,696 --> 00:40:01,530
like ultrasound, aren't powerful
enough to show them.
844
00:40:01,533 --> 00:40:03,592
BURNS: We need to be able
to see structures which are
845
00:40:03,602 --> 00:40:07,470
just a few tenths of a
millionths of a meter across.
846
00:40:07,472 --> 00:40:09,463
When it comes
to a living patient,
847
00:40:09,474 --> 00:40:11,568
we don't have
the tools to be able
848
00:40:11,576 --> 00:40:12,702
to see these tiny blood vessels.
849
00:40:12,711 --> 00:40:17,615
NARRATOR: But ultrasound does
provide a very good picture
850
00:40:17,616 --> 00:40:20,483
of the overall movement
of blood.
851
00:40:20,485 --> 00:40:22,544
"Is there any way,"
Burns wondered,
852
00:40:22,554 --> 00:40:25,615
"that images of blood flow
could reveal the hidden
853
00:40:25,624 --> 00:40:27,558
structure of the blood vessels?"
854
00:40:27,559 --> 00:40:30,517
To find out,
Burns and his colleagues
855
00:40:30,529 --> 00:40:32,486
used fractal geometry
856
00:40:32,497 --> 00:40:33,658
to make a mathematical model.
857
00:40:33,665 --> 00:40:35,565
BURNS:
If you have a mathematical way
858
00:40:35,567 --> 00:40:37,626
of analyzing a structure,
859
00:40:37,636 --> 00:40:38,694
you can make a model.
860
00:40:38,703 --> 00:40:41,661
What fractals do is they give
you some simple rules
861
00:40:41,673 --> 00:40:43,505
by which you can create models.
862
00:40:43,508 --> 00:40:46,637
And by changing some
of the parameters of the model,
863
00:40:46,645 --> 00:40:49,603
we can change
how the structure looks.
864
00:40:49,614 --> 00:40:50,638
NARRATOR:
The model showed
865
00:40:50,649 --> 00:40:53,607
the flow of blood in a kidney,
866
00:40:53,618 --> 00:40:54,676
first through
normal blood vessels,
867
00:40:54,686 --> 00:40:58,554
and then through vessels feeding
a cancerous tumor.
868
00:40:58,557 --> 00:41:01,583
Burns discovered that
the two kinds of networks
869
00:41:01,593 --> 00:41:05,587
had very different
fractal dimensions.
870
00:41:05,597 --> 00:41:07,531
Instead of being neatly
bifurcating,
871
00:41:07,532 --> 00:41:09,591
looking like a, a nice elm tree,
872
00:41:09,601 --> 00:41:13,663
the tumor vasculature
is chaotic and tangled
873
00:41:13,672 --> 00:41:17,666
and disorganized, looking more
like a mistletoe bush.
874
00:41:17,676 --> 00:41:20,475
NARRATOR:
And the flow of blood
875
00:41:20,479 --> 00:41:22,641
through these tangled vessels
looked very different
876
00:41:22,647 --> 00:41:25,514
than in a normal network--
a difference
877
00:41:25,517 --> 00:41:30,580
doctors might one day be able
to detect with ultrasound.
878
00:41:30,589 --> 00:41:33,581
We always thought that we have
to make medical images
879
00:41:33,592 --> 00:41:36,516
sharper and sharper,
ever more precise,
880
00:41:36,528 --> 00:41:38,690
ever more microscopic
in their resolution,
881
00:41:38,697 --> 00:41:41,462
to find out the information
882
00:41:41,466 --> 00:41:43,696
about the structure
that's there.
883
00:41:43,702 --> 00:41:45,466
What's exciting about this
884
00:41:45,470 --> 00:41:46,631
is it's giving us
microscopic information
885
00:41:46,638 --> 00:41:52,566
without us actually having
to look through a microscope.
886
00:41:52,577 --> 00:41:55,638
We think that this fractal
approach may be helpful
887
00:41:55,647 --> 00:41:58,708
in distinguishing benign
from malignant lesions
888
00:41:58,717 --> 00:42:02,620
in a way that hasn't been
possible up to now.
889
00:42:02,621 --> 00:42:05,647
NARRATOR: It may take
years before fractals
890
00:42:05,657 --> 00:42:07,682
can help doctors predict cancer.
891
00:42:07,692 --> 00:42:11,492
But they are already offering
clues to one of biology's
892
00:42:11,496 --> 00:42:12,691
more tantalizing mysteries:
893
00:42:12,697 --> 00:42:17,601
why big animals use energy
more efficiently
894
00:42:17,602 --> 00:42:20,526
than little ones.
895
00:42:20,539 --> 00:42:22,496
That's a question
that fascinates
896
00:42:22,507 --> 00:42:23,497
biologists James Brown
897
00:42:23,508 --> 00:42:26,534
and Brian Enquist
and physicist Geoffrey West.
898
00:42:26,545 --> 00:42:28,639
GEOFFREY WEST:
There is an extraordinary
899
00:42:28,647 --> 00:42:32,641
economy of scale
as you increase in size.
900
00:42:32,651 --> 00:42:35,643
NARRATOR:
An elephant, for example,
901
00:42:35,654 --> 00:42:39,579
is 200,000 times
heavier than a mouse,
902
00:42:39,591 --> 00:42:42,720
but uses only
about 10,000 times more energy
903
00:42:42,727 --> 00:42:45,651
in the form of calories
it consumes.
904
00:42:45,664 --> 00:42:50,568
WEST: The bigger you are, you
actually need less energy
905
00:42:50,569 --> 00:42:54,631
per gram of tissue
to stay alive.
906
00:42:54,639 --> 00:42:58,542
That is an amazing fact.
907
00:42:58,543 --> 00:43:00,534
NARRATOR: And even more
amazing is the fact
908
00:43:00,545 --> 00:43:04,539
that this relationship
between the mass and energy use
909
00:43:04,549 --> 00:43:07,507
of any living thing
is governed by a strict
910
00:43:07,519 --> 00:43:09,510
mathematical formula.
911
00:43:09,521 --> 00:43:11,580
JAMES BROWN:
So far as we know, that law
912
00:43:11,590 --> 00:43:17,552
is universal, or almost
universal, across all of life.
913
00:43:17,562 --> 00:43:20,520
So it operates
from the tiniest bacteria
914
00:43:20,532 --> 00:43:25,493
to whales and Sequoia trees.
915
00:43:25,503 --> 00:43:26,595
NARRATOR:
But even though this law
916
00:43:26,605 --> 00:43:28,630
had been discovered back
in the 1930s,
917
00:43:28,640 --> 00:43:31,598
no one had been able
to explain it.
918
00:43:31,610 --> 00:43:33,601
BROWN:
We had this idea that
919
00:43:33,612 --> 00:43:36,604
it probably had something to do
with how resources
920
00:43:36,615 --> 00:43:39,539
are distributed within
the bodies of organisms
921
00:43:39,551 --> 00:43:40,712
as they varied in size.
922
00:43:40,719 --> 00:43:43,552
We took this big leap and said,
923
00:43:43,555 --> 00:43:45,649
"All of life in some way
924
00:43:45,657 --> 00:43:49,685
"is sustained
by these underlying
925
00:43:49,694 --> 00:43:54,564
"networks that are
transporting oxygen,
926
00:43:54,566 --> 00:43:58,662
resources, metabolites
that are feeding cells."
927
00:43:58,670 --> 00:43:59,728
Circulatory systems
928
00:43:59,738 --> 00:44:01,570
and respiratory systems
929
00:44:01,573 --> 00:44:02,699
and renal systems
930
00:44:02,707 --> 00:44:04,664
and neural systems.
931
00:44:04,676 --> 00:44:08,704
It was obvious that fractals
were staring us in the face.
932
00:44:08,713 --> 00:44:12,672
NARRATOR: If all these
biological networks are fractal,
933
00:44:12,684 --> 00:44:16,587
it means they obey
some simple mathematical rules,
934
00:44:16,588 --> 00:44:19,717
which can lead to new insights
into how they work.
935
00:44:19,724 --> 00:44:21,749
BROWN: If you think
about it for a minute,
936
00:44:21,760 --> 00:44:23,751
it would be incredibly
inefficient
937
00:44:23,762 --> 00:44:24,752
to have a set of blueprints
938
00:44:24,763 --> 00:44:28,666
for every single stage
of increasing size.
939
00:44:28,667 --> 00:44:30,658
But if you have a fractal code,
940
00:44:30,669 --> 00:44:33,593
a code that says when to branch
941
00:44:33,605 --> 00:44:37,508
as you get bigger and bigger,
then, uh,
942
00:44:37,509 --> 00:44:39,637
a very simple genetic code
can produce
943
00:44:39,644 --> 00:44:43,638
what looks like
a complicated organism.
944
00:44:43,648 --> 00:44:47,710
Evolution by natural selection
has hit upon a design
945
00:44:47,719 --> 00:44:53,715
that appears to give
the most bang for the buck.
946
00:44:53,725 --> 00:44:57,628
NARRATOR:
In 1997, West, Brown
947
00:44:57,629 --> 00:44:59,654
and Enquist announced
their controversial theory
948
00:44:59,664 --> 00:45:03,726
that fractals hold the key
to the mysterious relationship
949
00:45:03,735 --> 00:45:06,636
between mass and energy use
in animals.
950
00:45:06,638 --> 00:45:11,565
Now, they are putting their
theory to a bold new test:
951
00:45:11,576 --> 00:45:12,668
an experiment to help determine
952
00:45:12,677 --> 00:45:14,668
if the fractal structure
of a single tree
953
00:45:14,679 --> 00:45:18,707
can predict how an entire
rain forest works.
954
00:45:18,717 --> 00:45:26,716
Measurements
of its trunk...
955
00:45:26,725 --> 00:45:30,593
NARRATOR: Enquist has
traveled to Costa Rica--
956
00:45:30,595 --> 00:45:32,529
to Guanacaste province,
957
00:45:32,530 --> 00:45:36,592
in the northwestern part
of the country.
958
00:45:36,601 --> 00:45:40,560
The government has set aside
more than 300,000 acres
959
00:45:40,572 --> 00:45:46,705
in Guanacaste
as a conservation area.
960
00:45:46,711 --> 00:45:49,635
This rain forest,
like others around the world,
961
00:45:49,647 --> 00:45:52,673
plays a vital role in regulating
the earth's climate,
962
00:45:52,684 --> 00:45:57,554
by removing carbon dioxide
from the atmosphere.
963
00:45:57,555 --> 00:45:58,681
If you look at the forest,
it basically breathes.
964
00:45:58,690 --> 00:46:02,615
And if we understand the total
amount of carbon dioxide,
965
00:46:02,627 --> 00:46:05,619
that's coming into, uh,
these trees within this forest
966
00:46:05,630 --> 00:46:08,691
we can then better
understand how, uh,
967
00:46:08,700 --> 00:46:11,601
this forest then ultimately
regulates the total amount
968
00:46:11,603 --> 00:46:13,731
of carbon dioxide
in our atmosphere.
969
00:46:13,738 --> 00:46:17,697
NARRATOR: With carbon dioxide
levels around the world rising,
970
00:46:17,709 --> 00:46:21,737
how much (:02 can rain forests
like this one absorb,
971
00:46:21,746 --> 00:46:25,649
and how important is their role
in protecting us
972
00:46:25,650 --> 00:46:28,745
from further global warming?
973
00:46:28,753 --> 00:46:31,552
Enquist and a team
of US scientists
974
00:46:31,556 --> 00:46:35,720
think that fractal geometry
may help answer these questions.
975
00:46:35,727 --> 00:46:36,626
...baseline.
976
00:46:36,628 --> 00:46:38,687
Let's try to get the height
of the tree measured.
977
00:46:38,696 --> 00:46:41,620
NARRATOR:
They are going to start by doing
978
00:46:41,633 --> 00:46:42,623
just about the last thing
979
00:46:42,634 --> 00:46:47,629
you'd think a scientist would do
here: cut down a balsa tree.
980
00:46:47,639 --> 00:46:48,731
It's dying anyway,
981
00:46:48,740 --> 00:46:51,698
and they have the permission
of the authorities.
982
00:46:51,709 --> 00:46:52,642
So Christina,
983
00:46:52,644 --> 00:46:54,635
as soon as you know the height
of that tree,
984
00:46:54,646 --> 00:46:57,707
we can actually figure out
the approximate angle
985
00:46:57,715 --> 00:46:59,581
that we need
to take it down on.
986
00:46:59,584 --> 00:47:02,713
NARRATOR: Hooking a guide
line on a high branch
987
00:47:02,720 --> 00:47:05,746
helps insure the tree will land
where they want it to.
988
00:47:05,757 --> 00:47:07,714
Yay!
989
00:47:07,725 --> 00:47:09,659
Good work.
990
00:47:09,661 --> 00:47:10,753
Very good.
991
00:47:10,762 --> 00:47:12,719
Very nice.
992
00:47:12,730 --> 00:47:17,730
(mechanical whirring)
993
00:47:33,651 --> 00:47:34,743
Nice.
994
00:47:34,752 --> 00:47:36,686
Well done.
995
00:47:36,688 --> 00:47:38,645
Jose, perfecto!
\AIsta bien?
996
00:47:38,656 --> 00:47:41,648
NARRATOR:
Enquist and his colleagues
997
00:47:41,659 --> 00:47:42,785
then measure
the width and length
998
00:47:42,794 --> 00:47:48,665
of the branches to quantify
the tree's fractal structure.
999
00:47:48,666 --> 00:47:50,760
Eight.
1000
00:47:50,768 --> 00:47:53,760
10.06.
1001
00:47:53,771 --> 00:47:58,698
No, that's eight.
1002
00:47:58,710 --> 00:48:01,611
6.3.
.03.
1003
00:48:01,613 --> 00:48:02,705
6.0.
1004
00:48:02,714 --> 00:48:03,545
Eight.
1005
00:48:03,548 --> 00:48:04,743
Seven on the nose.
1006
00:48:04,749 --> 00:48:08,617
NARRATOR: They also measure how
much carbon a single leaf contains,
1007
00:48:08,620 --> 00:48:10,714
which should allow them
to figure out
1008
00:48:10,722 --> 00:48:12,781
what the whole tree can absorb.
1009
00:48:12,790 --> 00:48:15,589
So if we know the amount
of carbon dioxide
1010
00:48:15,593 --> 00:48:17,652
that one leaf is able
to take in,
1011
00:48:17,662 --> 00:48:20,620
then hopefully using
the fractal branching rule
1012
00:48:20,632 --> 00:48:22,691
we can know
how much carbon dioxide
1013
00:48:22,700 --> 00:48:24,657
the entire tree is taking in.
1014
00:48:24,669 --> 00:48:27,570
NARRATOR:
Their next step is to move
1015
00:48:27,572 --> 00:48:33,705
from the tree
to the whole forest.
1016
00:48:33,711 --> 00:48:36,669
All right,
this is good.
1017
00:48:36,681 --> 00:48:38,581
13.2.
1018
00:48:38,583 --> 00:48:38,742
3.3.
1019
00:48:38,750 --> 00:48:40,684
ENQUIST: We're going
to census this forest.
1020
00:48:40,685 --> 00:48:42,710
We're going to be measuring
1021
00:48:42,720 --> 00:48:44,586
the diameter at the base
of the tree,
1022
00:48:44,589 --> 00:48:46,785
ranging all the way
from the largest trees down
1023
00:48:46,791 --> 00:48:48,657
to the smallest trees.
1024
00:48:48,660 --> 00:48:52,722
And in that way we can
then sample the distribution
1025
00:48:52,730 --> 00:48:54,755
of sizes within the forest.
1026
00:48:54,766 --> 00:48:58,725
It's 61.8 centimeters.
1027
00:48:58,736 --> 00:49:03,572
Even though the forest may
appear random and chaotic,
1028
00:49:03,574 --> 00:49:05,668
the team believes it
actually has a structure--
1029
00:49:05,677 --> 00:49:08,772
one that amazingly
is almost identical
1030
00:49:08,780 --> 00:49:13,741
to the fractal structure of the
tree they have just cut down.
1031
00:49:13,751 --> 00:49:17,585
BROWN:
The beautiful thing is
1032
00:49:17,588 --> 00:49:19,750
that the distribution
of the sizes
1033
00:49:19,757 --> 00:49:22,624
of individual trees
in the forest
1034
00:49:22,627 --> 00:49:25,653
appears to exactly match
the distribution
1035
00:49:25,663 --> 00:49:28,758
of the sizes
of individual branches
1036
00:49:28,766 --> 00:49:31,758
within a single tree.
1037
00:49:31,769 --> 00:49:33,703
NARRATOR:
If they're correct,
1038
00:49:33,705 --> 00:49:37,573
studying a single tree
will make it easier
1039
00:49:37,575 --> 00:49:39,634
to predict how much
carbon dioxide
1040
00:49:39,644 --> 00:49:44,741
an entire forest can absorb.
1041
00:49:44,749 --> 00:49:46,581
When they finish here,
1042
00:49:46,584 --> 00:49:48,780
they take their measurements
back to base camp,
1043
00:49:48,786 --> 00:49:51,744
where they'll see
if their ideas hold up.
1044
00:49:51,756 --> 00:49:53,781
So is this the...
this is the tree plot, right?
1045
00:49:53,791 --> 00:49:55,589
The cool thing is that,
1046
00:49:55,593 --> 00:49:58,722
if you look at the tree,
you see the same pattern
1047
00:49:58,730 --> 00:50:00,789
amongst the branches
as we see amongst the trunks
1048
00:50:00,798 --> 00:50:02,698
in the forest.
Very nice.
1049
00:50:02,700 --> 00:50:04,759
NARRATOR:
Just as they'd predicted,
1050
00:50:04,769 --> 00:50:07,670
the relative number
of big and small trees
1051
00:50:07,672 --> 00:50:09,766
closely matches
the relative number
1052
00:50:09,774 --> 00:50:12,607
of big and small branches.
1053
00:50:12,610 --> 00:50:15,602
ENQUIST: It's actually
phenomenal that it is parallel.
1054
00:50:15,613 --> 00:50:17,741
The slope of that line
for the tree appears
1055
00:50:17,749 --> 00:50:20,775
to be the same
for the forest as well.
1056
00:50:20,785 --> 00:50:22,685
So I guess it was worth
cutting up the tree.
1057
00:50:22,687 --> 00:50:25,816
It was definitely worth
cutting up the tree.
1058
00:50:25,823 --> 00:50:29,623
NARRATOR: So far, the
measurements from the field
1059
00:50:29,627 --> 00:50:31,618
appear to support
the scientist's theory
1060
00:50:31,629 --> 00:50:34,758
that a single tree
can help scientists assess
1061
00:50:34,766 --> 00:50:37,599
how much this rain forest
is helping
1062
00:50:37,602 --> 00:50:39,730
to slow down global warming.
1063
00:50:39,737 --> 00:50:42,695
By analyzing the fractal
patterns within the forest,
1064
00:50:42,707 --> 00:50:44,732
that then enables us
to do something
1065
00:50:44,742 --> 00:50:47,666
that we haven't really
been able to do before.
1066
00:50:47,678 --> 00:50:49,635
Have then a mathematical basis
1067
00:50:49,647 --> 00:50:52,639
to then predict
how the forest as a whole
1068
00:50:52,650 --> 00:50:54,778
takes in carbon dioxide
and, ultimately,
1069
00:50:54,786 --> 00:50:57,687
that's important
for understanding
1070
00:50:57,688 --> 00:51:02,592
what may happen
with global climate change.
1071
00:51:02,593 --> 00:51:03,651
NARRATOR:
For generations,
1072
00:51:03,661 --> 00:51:06,619
scientists believed
that the wildness of nature
1073
00:51:06,631 --> 00:51:08,759
could not be defined
by mathematics.
1074
00:51:08,766 --> 00:51:13,670
But fractal geometry is leading
to a whole new understanding,
1075
00:51:13,671 --> 00:51:15,730
revealing an underlying order
1076
00:51:15,740 --> 00:51:20,610
governed
by simple mathematical rules.
1077
00:51:20,611 --> 00:51:23,637
What I thought of
in my hikes through forests,
1078
00:51:23,648 --> 00:51:24,774
that, you know,
it's just a bunch of trees
1079
00:51:24,782 --> 00:51:27,808
of different sizes,
big ones here, small ones there,
1080
00:51:27,819 --> 00:51:31,744
looking like it's sort of
some arbitrary chaotic mess
1081
00:51:31,756 --> 00:51:35,750
actually has
an extraordinary structure.
1082
00:51:35,760 --> 00:51:38,718
NARRATOR: A structure
that can be mapped out
1083
00:51:38,729 --> 00:51:44,623
and measured
using fractal geometry.
1084
00:51:44,635 --> 00:51:47,798
ENQUIST: What's absolutely
amazing is that you can
1085
00:51:47,805 --> 00:51:50,763
translate what you see
in the natural world
1086
00:51:50,775 --> 00:51:51,799
in the language of mathematics.
1087
00:51:51,809 --> 00:52:00,672
And I can't think of anything
more beautiful than that.
1088
00:52:00,685 --> 00:52:02,619
Math is our one
and only strategy
1089
00:52:02,620 --> 00:52:05,817
for understanding
the complexity of nature.
1090
00:52:05,823 --> 00:52:09,726
Now, fractal geometry
has given us
1091
00:52:09,727 --> 00:52:11,786
a much larger vocabulary.
1092
00:52:11,796 --> 00:52:13,787
And with the larger vocabulary
1093
00:52:13,798 --> 00:52:18,798
we can read more
of the book of nature.
1094
00:52:48,799 --> 00:52:50,665
On NOVA's "Hidden Dimension"
Web site,
1095
00:52:50,668 --> 00:52:54,798
explore the Mandelbrot set, see
a gallery of fractal images
1096
00:52:54,805 --> 00:52:56,603
and much more.
1097
00:52:56,607 --> 00:53:00,805
Find it on pbs.org.
1098
00:53:00,811 --> 00:53:06,739
Major funding for NOVA
is provided by the following:
1099
00:53:06,751 --> 00:53:11,712
Taking on the world's
toughest energy challenges.
1100
00:53:11,722 --> 00:53:16,785
And by:
1101
00:53:16,794 --> 00:53:21,794
And...
1102
00:53:26,804 --> 00:53:29,796
And by the Corporation
for Public Broadcasting
1103
00:53:29,807 --> 00:53:36,679
and by contributions
to your PBS station from:
1104
00:53:36,681 --> 00:53:39,681
Captioned by Media Access
Group at WGBH access.wgbh.org
87384
Can't find what you're looking for?
Get subtitles in any language from opensubtitles.com, and translate them here.