1
00:00:00,760 --> 00:00:03,456
On January 26, 2013,

2
00:00:03,480 --> 00:00:06,976
a band of al-Qaeda militants
entered the ancient city of Timbuktu

3
00:00:07,000 --> 00:00:09,000
on the southern edge of the Sahara Desert.

4
00:00:09,640 --> 00:00:14,256
There, they set fire to a medieval library
of 30,000 manuscripts

5
00:00:14,280 --> 00:00:17,056
written in Arabic
and several African languages

6
00:00:17,080 --> 00:00:22,816
and ranging in subject from astronomy
to geography, history to medicine,

7
00:00:22,840 --> 00:00:24,536
including one book which records

8
00:00:24,560 --> 00:00:28,480
perhaps the first treatment
for male erectile dysfunction.

9
00:00:29,760 --> 00:00:31,056
Unknown in the West,

10
00:00:31,080 --> 00:00:34,496
this was the collected wisdom
of an entire continent,

11
00:00:34,520 --> 00:00:38,600
the voice of Africa at a time when Africa
was thought not to have a voice at all.

12
00:00:39,480 --> 00:00:41,976
The mayor of Bamako,
who witnessed the event,

13
00:00:42,000 --> 00:00:43,936
called the burning of the manuscripts

14
00:00:43,960 --> 00:00:46,200
"a crime against world cultural heritage."

15
00:00:46,880 --> 00:00:48,416
And he was right --

16
00:00:48,440 --> 00:00:51,880
or he would have been, if it weren't
for the fact that he was also lying.

17
00:00:52,560 --> 00:00:55,296
In fact, just before,

18
00:00:55,320 --> 00:00:59,456
African scholars had collected
a random assortment of old books

19
00:00:59,480 --> 00:01:01,976
and left them out
for the terrorists to burn.

20
00:01:02,000 --> 00:01:04,936
Today, the collection
lies hidden in Bamako,

21
00:01:04,959 --> 00:01:06,336
the capital of Mali,

22
00:01:06,360 --> 00:01:08,456
moldering in the high humidity.

23
00:01:08,480 --> 00:01:10,336
What was rescued by ruse

24
00:01:10,360 --> 00:01:12,096
is now once again in jeopardy,

25
00:01:12,120 --> 00:01:13,360
this time by climate.

26
00:01:14,200 --> 00:01:16,616
But Africa, and the far-flung
corners of the world,

27
00:01:16,640 --> 00:01:18,936
are not the only places,
or even the main places

28
00:01:18,960 --> 00:01:23,256
in which manuscripts that could change
the history of world culture

29
00:01:23,280 --> 00:01:24,800
are in jeopardy.

30
00:01:25,920 --> 00:01:30,656
Several years ago, I conducted
a survey of European research libraries

31
00:01:30,680 --> 00:01:32,936
and discovered that,
at the barest minimum,

32
00:01:32,960 --> 00:01:36,176
there are 60,000 manuscripts

33
00:01:36,200 --> 00:01:37,816
pre-1500

34
00:01:37,840 --> 00:01:40,776
that are illegible
because of water damage,

35
00:01:40,800 --> 00:01:44,776
fading, mold and chemical reagents.

36
00:01:44,800 --> 00:01:47,776
The real number is likely double that,

37
00:01:47,800 --> 00:01:49,976
and that doesn't even count

38
00:01:50,000 --> 00:01:52,616
Renaissance manuscripts
and modern manuscripts

39
00:01:52,640 --> 00:01:55,760
and cultural heritage
objects such as maps.

40
00:01:57,960 --> 00:02:00,416
What if there were a technology

41
00:02:00,440 --> 00:02:05,776
that could recover
these lost and unknown works?

42
00:02:05,800 --> 00:02:10,216
Imagine worldwide
how a trove of hundreds of thousands

43
00:02:10,240 --> 00:02:13,176
of previously unknown texts

44
00:02:13,200 --> 00:02:16,240
could radically transform
our knowledge of the past.

45
00:02:18,280 --> 00:02:22,296
Imagine what unknown classics
we would discover

46
00:02:22,320 --> 00:02:25,656
which would rewrite the canons
of literature, history,

47
00:02:25,680 --> 00:02:27,120
philosophy, music --

48
00:02:27,840 --> 00:02:31,376
or, more provocatively, that could
rewrite our cultural identities,

49
00:02:31,400 --> 00:02:34,760
building new bridges
between people and culture.

50
00:02:35,520 --> 00:02:38,056
These are the questions
that transformed me

51
00:02:38,080 --> 00:02:40,896
from a medieval scholar,
a reader of texts,

52
00:02:40,920 --> 00:02:42,720
into a textual scientist.

53
00:02:44,120 --> 00:02:46,416
What an unsatisfying word "reader" is.

54
00:02:46,440 --> 00:02:49,016
For me, it conjures up
images of passivity,

55
00:02:49,040 --> 00:02:51,456
of someone sitting idly in an armchair

56
00:02:51,480 --> 00:02:53,776
waiting for knowledge to come to him

57
00:02:53,800 --> 00:02:55,496
in a neat little parcel.

58
00:02:55,520 --> 00:02:58,536
How much better to be
a participant in the past,

59
00:02:58,560 --> 00:03:01,736
an adventurer in an undiscovered country,

60
00:03:01,760 --> 00:03:04,160
searching for the hidden text.

61
00:03:05,360 --> 00:03:07,800
As an academic, I was a mere reader.

62
00:03:08,560 --> 00:03:10,976
I read and taught the same classics

63
00:03:11,000 --> 00:03:14,136
that people had been reading
and teaching for hundreds of years --

64
00:03:14,160 --> 00:03:17,176
Virgil, Ovid, Chaucer, Petrarch --

65
00:03:17,200 --> 00:03:19,536
and with every scholarly article
that I published

66
00:03:19,560 --> 00:03:22,760
I added to human knowledge
in ever-diminishing slivers of insight.

67
00:03:24,760 --> 00:03:26,336
What I wanted to be

68
00:03:26,360 --> 00:03:28,456
was an archaeologist of the past,

69
00:03:28,480 --> 00:03:30,056
a discoverer of literature,

70
00:03:30,080 --> 00:03:31,976
an Indiana Jones without the whip --

71
00:03:32,000 --> 00:03:33,496
or, actually, with the whip.

72
00:03:33,520 --> 00:03:34,736
(Laughter)

73
00:03:34,760 --> 00:03:38,456
And I wanted it not just for myself
but I wanted it for my students as well.

74
00:03:38,480 --> 00:03:42,376
And so six years ago,
I changed the direction of my career.

75
00:03:42,400 --> 00:03:45,416
At the time, I was working
on "The Chess of Love,"

76
00:03:45,440 --> 00:03:48,296
the last important long poem
of the European Middle Ages

77
00:03:48,320 --> 00:03:49,936
never to have been edited.

78
00:03:49,960 --> 00:03:52,976
And it wasn't edited because
it existed in only one manuscript

79
00:03:53,000 --> 00:03:56,256
which was so badly damaged
during the firebombing of Dresden

80
00:03:56,280 --> 00:03:57,496
in World War II

81
00:03:57,520 --> 00:04:00,400
that generations of scholars
had pronounced it lost.

82
00:04:01,400 --> 00:04:04,856
For five years, I had been working
with an ultraviolet lamp

83
00:04:04,880 --> 00:04:06,776
trying to recover traces of the writing

84
00:04:06,800 --> 00:04:09,216
and I'd gone about as far
as technology at the time

85
00:04:09,240 --> 00:04:10,616
could actually take me.

86
00:04:10,640 --> 00:04:12,736
And so I did what many people do.

87
00:04:12,760 --> 00:04:14,776
I went online,

88
00:04:14,800 --> 00:04:16,536
and there I learned about

89
00:04:16,560 --> 00:04:21,136
how multispectral imaging had been used
to recover two lost treatises

90
00:04:21,160 --> 00:04:24,056
of the famed Greek
mathematician Archimedes

91
00:04:24,080 --> 00:04:25,616
from a 13th-century palimpsest.

92
00:04:25,640 --> 00:04:28,920
A palimpsest is a manuscript
which has been erased and overwritten.

93
00:04:30,200 --> 00:04:31,456
And so, out of the blue,

94
00:04:31,480 --> 00:04:34,536
I decided to write
to the lead imaging scientist

95
00:04:34,560 --> 00:04:36,776
on the Archimedes palimpsest project,

96
00:04:36,800 --> 00:04:38,296
Professor Roger Easton,

97
00:04:38,320 --> 00:04:40,136
with a plan and a plea.

98
00:04:40,160 --> 00:04:42,200
And to my surprise,
he actually wrote back.

99
00:04:44,000 --> 00:04:47,656
With his help, I was able
to win a grant from the US government

100
00:04:47,680 --> 00:04:51,696
to build a transportable,
multispectral imaging lab,

101
00:04:51,720 --> 00:04:56,536
And with this lab, I transformed
what was a charred and faded mess

102
00:04:56,560 --> 00:04:58,560
into a new medieval classic.

103
00:04:59,360 --> 00:05:02,136
So how does multispectral
imaging actually work?

104
00:05:02,160 --> 00:05:04,616
Well, the idea
behind multispectral imaging

105
00:05:04,640 --> 00:05:08,936
is something that anyone who is familiar
with infrared night vision goggles

106
00:05:08,960 --> 00:05:10,336
will immediately appreciate:

107
00:05:10,360 --> 00:05:12,856
that what we can see
in the visible spectrum of light

108
00:05:12,880 --> 00:05:15,200
is only a tiny fraction
of what's actually there.

109
00:05:15,720 --> 00:05:17,920
The same is true with invisible writing.

110
00:05:19,280 --> 00:05:23,456
Our system uses 12 wavelengths of light

111
00:05:23,480 --> 00:05:25,976
between the ultraviolet and the infrared,

112
00:05:26,000 --> 00:05:29,056
and these are shown down
onto the manuscript from above

113
00:05:29,080 --> 00:05:30,696
from banks of LEDs,

114
00:05:30,720 --> 00:05:32,576
and another multispectral light source

115
00:05:32,600 --> 00:05:35,553
which comes up through
the individual leaves of the manuscript.

116
00:05:35,577 --> 00:05:40,096
Up to 35 images per sequence
per leaf are imaged this way

117
00:05:40,120 --> 00:05:42,740
using a high-powered digital camera
equipped with a lens

118
00:05:42,764 --> 00:05:44,736
which is made out of quartz.

119
00:05:44,760 --> 00:05:46,856
There are about five
of these in the world.

120
00:05:46,880 --> 00:05:48,696
And once we capture these images,

121
00:05:48,720 --> 00:05:50,776
we feed them through
statistical algorithms

122
00:05:50,800 --> 00:05:53,296
to further enhance and clarify them,

123
00:05:53,320 --> 00:05:56,696
using software which was originally
designed for satellite images

124
00:05:56,720 --> 00:06:00,096
and used by people
like geospatial scientists

125
00:06:00,120 --> 00:06:01,320
and the CIA.

126
00:06:02,040 --> 00:06:04,336
The results can be spectacular.

127
00:06:04,360 --> 00:06:06,536
You may already have heard
of what's been done

128
00:06:06,560 --> 00:06:07,936
for the Dead Sea Scrolls,

129
00:06:07,960 --> 00:06:09,520
which are slowly gelatinizing.

130
00:06:10,480 --> 00:06:14,136
Using infrared, we've been able
to read even the darkest corners

131
00:06:14,160 --> 00:06:15,880
of the Dead Sea Scrolls.

132
00:06:16,880 --> 00:06:18,336
You may not be aware, however,

133
00:06:18,360 --> 00:06:21,096
of other Biblical texts
that are in jeopardy.

134
00:06:21,120 --> 00:06:24,536
Here, for example,
is a leaf from a manuscript

135
00:06:24,560 --> 00:06:26,336
that we imaged,

136
00:06:26,360 --> 00:06:30,240
which is perhaps the most valuable
Christian Bible in the world.

137
00:06:30,880 --> 00:06:36,656
The Codex Vercellensis is the oldest
translation of the Gospels into Latin,

138
00:06:36,680 --> 00:06:39,280
and it dates from the first half
of the fourth century.

139
00:06:40,560 --> 00:06:42,816
This is the closest we can come

140
00:06:42,840 --> 00:06:46,776
to the Bible at the time
of the foundation of Christendom

141
00:06:46,800 --> 00:06:48,416
under Emperor Constantine,

142
00:06:48,440 --> 00:06:50,976
and at the time also
of the Council of Nicaea,

143
00:06:51,000 --> 00:06:54,496
when the basic creed of Christianity
was being agreed upon.

144
00:06:54,520 --> 00:06:57,776
This manuscript, unfortunately,
has been very badly damaged,

145
00:06:57,800 --> 00:07:00,016
and it's damaged because for centuries

146
00:07:00,040 --> 00:07:02,496
it had been used and handled

147
00:07:02,520 --> 00:07:05,216
in swearing in ceremonies in the church.

148
00:07:05,240 --> 00:07:09,896
In fact, that purple splotch
that you see in the upper left hand corner

149
00:07:09,920 --> 00:07:14,216
is Aspergillus, which is a fungus

150
00:07:14,240 --> 00:07:17,776
which originates in the unwashed hands

151
00:07:17,800 --> 00:07:19,800
of a person with tuberculosis.

152
00:07:20,640 --> 00:07:23,696
Our imaging has enabled me
to make the first transcription

153
00:07:23,720 --> 00:07:26,320
of this manuscript in 250 years.

154
00:07:27,800 --> 00:07:31,256
Having a lab that can travel
to collections where it's needed, however,

155
00:07:31,280 --> 00:07:32,760
is only part of the solution.

156
00:07:33,480 --> 00:07:36,336
The technology is expensive and very rare,

157
00:07:36,360 --> 00:07:39,576
and the imaging and image
processing skills are esoteric.

158
00:07:39,600 --> 00:07:41,296
That means that mounting recoveries

159
00:07:41,320 --> 00:07:46,216
is beyond the reach of most researchers
and all but the wealthiest institutions.

160
00:07:46,240 --> 00:07:49,016
That's why I founded the Lazarus Project,

161
00:07:49,040 --> 00:07:50,696
a not-for-profit initiative

162
00:07:50,720 --> 00:07:54,696
to bring multispectral imaging
to individual researchers

163
00:07:54,720 --> 00:07:58,560
and smaller institutions
at little or no cost whatsoever.

164
00:07:59,560 --> 00:08:01,176
Over the past five years,

165
00:08:01,200 --> 00:08:05,056
our team of imaging scientists,
scholars and students

166
00:08:05,080 --> 00:08:07,256
has travelled to seven different countries

167
00:08:07,280 --> 00:08:11,136
and have recovered some of the world's
most valuable damaged manuscripts,

168
00:08:11,160 --> 00:08:14,160
included the Vercelli Book,
which is the oldest book of English,

169
00:08:14,184 --> 00:08:16,800
the Black Book of Carmarthen,
the oldest book of Welsh,

170
00:08:16,824 --> 00:08:20,296
and some of the most valuable
earliest Gospels

171
00:08:20,320 --> 00:08:23,200
located in what is now
the former Soviet Georgia.

172
00:08:24,600 --> 00:08:27,560
So, spectral imaging
can recover lost texts.

173
00:08:28,360 --> 00:08:33,456
More subtly, though, it can recover
a second story behind every object,

174
00:08:33,480 --> 00:08:38,176
the story of how, when
and by whom a text was created,

175
00:08:38,200 --> 00:08:41,600
and, sometimes, what the author
was thinking at the time he wrote.

176
00:08:42,600 --> 00:08:45,696
Take, for example, a draft
of the Declaration of Independence

177
00:08:45,720 --> 00:08:48,136
written in Thomas Jefferson's own hand,

178
00:08:48,160 --> 00:08:50,589
which some colleagues of mine
imaged a few years ago

179
00:08:50,613 --> 00:08:51,933
at the Library of Congress.

180
00:08:52,360 --> 00:08:55,056
Curators had noticed
that one word throughout

181
00:08:55,080 --> 00:08:57,456
had been scratched out and overwritten.

182
00:08:57,480 --> 00:08:59,560
The word overwritten was "citizens."

183
00:09:00,240 --> 00:09:02,840
Perhaps you can guess
what the word underneath was.

184
00:09:03,960 --> 00:09:05,376
"Subjects."

185
00:09:05,400 --> 00:09:08,216
There, ladies and gentlemen,
is American democracy

186
00:09:08,240 --> 00:09:10,320
evolving under the hand
of Thomas Jefferson.

187
00:09:11,360 --> 00:09:15,216
Or consider the 1491 Martellus Map,

188
00:09:15,240 --> 00:09:17,576
which we imaged
at Yale's Beinecke Library.

189
00:09:17,600 --> 00:09:20,056
This was the map
that Columbus likely consulted

190
00:09:20,080 --> 00:09:21,776
before he traveled to the New World

191
00:09:21,800 --> 00:09:24,536
and which gave him his idea
of what Asia looked like

192
00:09:24,560 --> 00:09:26,160
and where Japan was located.

193
00:09:27,640 --> 00:09:30,656
The problem with this map
is that its inks and pigments

194
00:09:30,680 --> 00:09:32,776
had so degraded over time

195
00:09:32,800 --> 00:09:34,936
that this large, nearly seven-foot map,

196
00:09:34,960 --> 00:09:37,040
made the world look like a giant desert.

197
00:09:37,520 --> 00:09:41,136
Until now, we had very little idea,
detailed idea, that is,

198
00:09:41,160 --> 00:09:42,816
of what Columbus knew of the world

199
00:09:42,840 --> 00:09:44,760
and how world cultures were represented.

200
00:09:45,240 --> 00:09:49,376
The main legend of the map
was entirely illegible under normal light.

201
00:09:49,400 --> 00:09:51,656
Ultraviolet did very little for it.

202
00:09:51,680 --> 00:09:53,680
Multispectral gave us everything.

203
00:09:54,640 --> 00:09:58,216
In Asia, we learned of monsters
with ears so long

204
00:09:58,240 --> 00:10:00,560
that they could cover
the creature's entire body.

205
00:10:01,040 --> 00:10:05,440
In Africa, about a snake
who could cause the ground to smoke.

206
00:10:06,680 --> 00:10:08,936
Like starlight, which can convey images

207
00:10:08,960 --> 00:10:11,536
of the way the Universe
looked in the distant past,

208
00:10:11,560 --> 00:10:15,456
so multispectral light can take us back
to the first stuttering moments

209
00:10:15,480 --> 00:10:16,760
of an object's creation.

210
00:10:17,480 --> 00:10:21,416
Through this lens, we witness
the mistakes, the changes of mind,

211
00:10:21,440 --> 00:10:24,336
the naïvetés, the uncensored thoughts,

212
00:10:24,360 --> 00:10:26,576
the imperfections of the human imagination

213
00:10:26,600 --> 00:10:29,416
that allow these hallowed objects
and their authors

214
00:10:29,440 --> 00:10:31,096
to become more real,

215
00:10:31,120 --> 00:10:33,640
that make history closer to us.

216
00:10:34,760 --> 00:10:36,040
What about the future?

217
00:10:36,480 --> 00:10:38,656
There's so much of the past,

218
00:10:38,680 --> 00:10:41,896
and so few people
with the skills to rescue it

219
00:10:41,920 --> 00:10:45,800
before these objects disappear forever.

220
00:10:46,480 --> 00:10:49,776
That's why I have begun to teach
this new hybrid discipline

221
00:10:49,800 --> 00:10:51,696
that I call "textual science."

222
00:10:51,720 --> 00:10:53,336
Textual science is a marriage

223
00:10:53,360 --> 00:10:55,736
of the traditional skills
of a literary scholar --

224
00:10:55,760 --> 00:10:58,296
the ability to read old languages
and old handwriting,

225
00:10:58,320 --> 00:11:00,016
the knowledge of how texts are made

226
00:11:00,040 --> 00:11:02,176
in order to be able
to place and date them --

227
00:11:02,200 --> 00:11:04,936
with new techniques like imaging science,

228
00:11:04,960 --> 00:11:07,536
the chemistry of inks and pigments,

229
00:11:07,560 --> 00:11:09,960
computer-aided optical
character recognition.

230
00:11:11,160 --> 00:11:13,336
Last year, a student in my class,

231
00:11:13,360 --> 00:11:14,576
a freshman,

232
00:11:14,600 --> 00:11:16,416
with a background in Latin and Greek,

233
00:11:16,440 --> 00:11:18,776
was image-processing a palimpsest

234
00:11:18,800 --> 00:11:21,520
that we had photographed
at a famous library in Rome.

235
00:11:22,240 --> 00:11:27,000
As he worked, tiny Greek writing
began to appear from behind the text.

236
00:11:28,200 --> 00:11:29,696
Everyone gathered around,

237
00:11:29,720 --> 00:11:32,376
and he read a line from a lost work

238
00:11:32,400 --> 00:11:34,720
of the Greek comic dramatist Menander.

239
00:11:35,760 --> 00:11:38,456
This was the first time
in well over a thousand years

240
00:11:38,480 --> 00:11:41,040
that those words
had been pronounced aloud.

241
00:11:41,880 --> 00:11:44,440
In that moment, he became a scholar.

242
00:11:45,440 --> 00:11:48,320
Ladies and gentlemen,
that is the future of the past.

243
00:11:48,840 --> 00:11:50,056
Thank you very much.

244
00:11:50,080 --> 00:11:53,080
(Applause)