TED2011
Rajesh Rao: A Rosetta Stone for a lost language
Rajesh Rao: En Rosettasten for Indusskriften
Filmed:
Readability: 4.2
2,103,451 views
Rajesh Rao er fascineret af "alle krydsords moder": Hvordan skal det 4000 år gamle Indusskrift decifreres ?. Hos TED 2011 fortæller han, hvordan han bruger moderne computer-teknikker til at læse Indus-sproget - den afgørende nøgle til at forstå denne gamle civilisation.
Rajesh Rao - Computational neuroscientist
Rajesh Rao seeks to understand the human brain through computational modeling, on two fronts: developing computer models of our minds, and using tech to decipher the 4,000-year-old lost script of the Indus Valley civilization. Full bio
Rajesh Rao seeks to understand the human brain through computational modeling, on two fronts: developing computer models of our minds, and using tech to decipher the 4,000-year-old lost script of the Indus Valley civilization. Full bio
Double-click the English transcript below to play the video.
00:15
I'd like to begin with a thought experiment.
0
0
3000
Jeg vil begynde med et tankeeksperiment.
00:19
Imagine that it's 4,000 years into the future.
1
4000
3000
Vi er 4000 år ude i fremtiden.
00:22
Civilization as we know it
2
7000
2000
Civilisationen, som vi kender den,
00:24
has ceased to exist --
3
9000
2000
findes ikke mere
00:26
no books,
4
11000
2000
ingen bøger,
00:28
no electronic devices,
5
13000
3000
ingen elektroniske dimser,
00:31
no Facebook or Twitter.
6
16000
3000
ingen Facebook og Twitter.
00:34
All knowledge of the English language and the English alphabet
7
19000
3000
Al viden om det engelske sprog
00:37
has been lost.
8
22000
2000
og alfabetet, er gået tabt.
00:39
Now imagine archeologists
9
24000
2000
Forestil jer nu at arkæologerne
00:41
digging through the rubble of one of our cities.
10
26000
2000
graver i ruinerne fra vore byer.
00:43
What might they find?
11
28000
2000
Hvad finder de?
00:45
Well perhaps some rectangular pieces of plastic
12
30000
3000
Måske nogle firkantede stykker plastik,
00:48
with strange symbols on them.
13
33000
3000
med mærkelige symboler på dem.
00:51
Perhaps some circular pieces of metal.
14
36000
3000
Måske nogle runde stykker metal.
00:54
Maybe some cylindrical containers
15
39000
2000
Måske nogle cylinderformede beholdere,
00:56
with some symbols on them.
16
41000
2000
med symboler på dem.
00:58
And perhaps one archeologist becomes an instant celebrity
17
43000
3000
Og måske bliver en arkæolog berømt
01:01
when she discovers --
18
46000
2000
da hun i bjergene,
01:03
buried in the hills somewhere in North America --
19
48000
2000
et sted i Nord Amerika, opdager
01:05
massive versions of these same symbols.
20
50000
3000
massive udgaver af de samme symboler. (Latter)
01:10
Now let's ask ourselves,
21
55000
2000
Lad os nu spørge os:
01:12
what could such artifacts say about us
22
57000
3000
hvad kunne sådanne fund fortælle
01:15
to people 4,000 years into the future?
23
60000
3000
om os om 4000 år ?
01:18
This is no hypothetical question.
24
63000
2000
Det er ikke noget hypotetisk spørgsmål.
01:20
In fact, this is exactly the kind of question we're faced with
25
65000
3000
Det er præcis det, vi er konfronteret med,
01:23
when we try to understand the Indus Valley civilization,
26
68000
3000
når vi prøver at forstå Indusdalens civilisation
01:26
which existed 4,000 years ago.
27
71000
2000
som fandtes for 4000 år siden.
01:28
The Indus civilization was roughly contemporaneous
28
73000
3000
Induscivilisationen var omtrentlig samtidig
01:31
with the much better known Egyptian and the Mesopotamian civilizations,
29
76000
3000
med de langt mere kendte Ægyptiske
og Mesopotamiske civilisationer,
og Mesopotamiske civilisationer,
01:34
but it was actually much larger than either of these two civilizations.
30
79000
3000
men det var langt større end de to.
01:37
It occupied the area
31
82000
2000
Det bredte sig over
01:39
of approximately one million square kilometers,
32
84000
2000
omtrentlig 1 million km²,
01:41
covering what is now Pakistan,
33
86000
2000
og dækkede det nuværende Pakistan,
01:43
Northwestern India
34
88000
2000
Nordvest Indien
01:45
and parts of Afghanistan and Iran.
35
90000
2000
og dele af Afghanistan og Iran.
01:47
Given that it was such a vast civilization,
36
92000
2000
Eftersom at det var en kæmpe civilisation
01:49
you might expect to find really powerful rulers, kings,
37
94000
4000
kunne man forvente at finde rigtig magtfulde
herskere, konger
herskere, konger
01:53
and huge monuments glorifying these powerful kings.
38
98000
3000
og store monumenter, som glorificerede
disse magtfulde konger.
disse magtfulde konger.
01:56
In fact,
39
101000
2000
Men,
01:58
what archeologists have found is none of that.
40
103000
2000
arkæologerne fandt intet af dette.
02:00
They've found small objects such as these.
41
105000
3000
De fandt små objekter som disse.
02:03
Here's an example of one of these objects.
42
108000
3000
Her er et eksempel på eet af disse objekter.
02:06
Well obviously this is a replica.
43
111000
2000
Dette er selvfølgelig en replika.
02:08
But who is this person?
44
113000
3000
Men hvem er denne person ?
02:11
A king? A god?
45
116000
2000
En konge ? En gud ?
02:13
A priest?
46
118000
2000
En præst ?
02:15
Or perhaps an ordinary person
47
120000
2000
Eller måske et almindeligt menneske
02:17
like you or me?
48
122000
2000
som dig og mig ?
02:19
We don't know.
49
124000
2000
Vi ved det ikke.
02:21
But the Indus people also left behind artifacts with writing on them.
50
126000
3000
Men Indusfolket efterlod også genstande med tekst på.
02:24
Well no, not pieces of plastic,
51
129000
2000
Nej, ikke plastik-kort,
02:26
but stone seals, copper tablets,
52
131000
3000
men stensegl, kobberstykker
02:29
pottery and, surprisingly,
53
134000
2000
krukker og - meget overraskende -
02:31
one large sign board,
54
136000
2000
en stor tavle med tegn,
02:33
which was found buried near the gate of a city.
55
138000
2000
som blev fundet nær porten til en by.
02:35
Now we don't know if it says Hollywood,
56
140000
2000
Vi ved ikke om der står Hollywood,
02:37
or even Bollywood for that matter.
57
142000
2000
eller Bollywood for den sags skyld.
02:39
In fact, we don't even know
58
144000
2000
Vi ved ikke engang
02:41
what any of these objects say,
59
146000
2000
hvad et eneste af disse objekter fortæller,
02:43
and that's because the Indus script is undeciphered.
60
148000
3000
fordi Indusskriften er ikke afkodet.
02:46
We don't know what any of these symbols mean.
61
151000
2000
Vi ved ikke hvad disse symboler betyder.
02:48
The symbols are most commonly found on seals.
62
153000
3000
Symbolerne findes hyppigst på segl.
02:51
So you see up there one such object.
63
156000
2000
og deroppe ses et sådant objekt.
02:53
It's the square object with the unicorn-like animal on it.
64
158000
3000
Det er det firkantede med en enhjørning på det.
02:56
Now that's a magnificent piece of art.
65
161000
2000
Det er et flot stykke kunst.
02:58
So how big do you think that is?
66
163000
2000
Hvor stor tror I det er ?
03:00
Perhaps that big?
67
165000
2000
Måske så stor ?
03:02
Or maybe that big?
68
167000
2000
Eller måske så stor ?
03:04
Well let me show you.
69
169000
2000
Lad mig vise det.
03:07
Here's a replica of one such seal.
70
172000
3000
Her er et kopi af sådant segl.
03:10
It's only about one inch by one inch in size --
71
175000
2000
Det er kun en tomme på hver led
03:12
pretty tiny.
72
177000
2000
-- ret lille.
03:14
So what were these used for?
73
179000
2000
Hvad bruges de til ?
03:16
We know that these were used for stamping clay tags
74
181000
3000
Vi ved, at de blev brugt til at præge leretiketter,
03:19
that were attached to bundles of goods that were sent from one place to the other.
75
184000
3000
som blev fastgjort til varer, som blev sendt
fra eet sted til et andet.
fra eet sted til et andet.
03:22
So you know those packing slips you get on your FedEx boxes?
76
187000
3000
Ligesom de pakke-sedler, der er på FedEx-kasser.
03:25
These were used to make those kinds of packing slips.
77
190000
3000
De blev brugt til at lave disse pakkesedler.
03:28
You might wonder what these objects contain
78
193000
3000
Du spørger måske, hvad der står
03:31
in terms of their text.
79
196000
2000
på dem.
03:33
Perhaps they're the name of the sender
80
198000
2000
Måske står der afsenderens navn
03:35
or some information about the goods
81
200000
2000
eller information om varerne,
03:37
that are being sent from one place to the other -- we don't know.
82
202000
3000
som blev sendt fra eet sted til et andet.
Vi ved det ikke.
Vi ved det ikke.
03:40
We need to decipher the script to answer that question.
83
205000
2000
Vi er nødt til at decifrere teksten,
for at besvare det spørgsmål.
for at besvare det spørgsmål.
03:42
Deciphering the script
84
207000
2000
Decifreringen af teksten
03:44
is not just an intellectual puzzle;
85
209000
2000
er ikke bare en intellektuelt gåde..
03:46
it's actually become a question
86
211000
2000
Det er faktisk blevet et spørgsmål,
03:48
that's become deeply intertwined
87
213000
2000
som er blevet stærkt sammenflettet
03:50
with the politics and the cultural history of South Asia.
88
215000
3000
med politik og Sydasiens kulturhistorie.
03:53
In fact, the script has become a battleground of sorts
89
218000
3000
Skriften er faktisk blevet en kampplads,
03:56
between three different groups of people.
90
221000
2000
for tre forskellige grupper.
03:58
First, there's a group of people
91
223000
2000
Først er der den gruppe
04:00
who are very passionate in their belief
92
225000
2000
som er meget lidenskabelig i deres tro på
04:02
that the Indus script
93
227000
2000
at Indusskriften
04:04
does not represent a language at all.
94
229000
2000
ikke repræsentere noget sprog overhovedet.
04:06
These people believe that the symbols
95
231000
2000
De mener, at symbolerne
04:08
are very similar to the kind of symbols you find on traffic signs
96
233000
3000
meget ligner de symboler man ser på trafikskilte
04:11
or the emblems you find on shields.
97
236000
3000
og emblemer på skjolde.
04:14
There's a second group of people
98
239000
2000
Der er en anden gruppe,
04:16
who believe that the Indus script represents an Indo-European language.
99
241000
3000
som mener, at Indus-skriften repræsenterer
et Indoeuropæisk sprog.
et Indoeuropæisk sprog.
04:19
If you look at a map of India today,
100
244000
2000
Hvis man ser på et kort over Indien i dag,
04:21
you'll see that most of the languages spoken in North India
101
246000
3000
så ser man, at de fleste sprog, som tales i Nordindien,
04:24
belong to the Indo-European language family.
102
249000
3000
tilhører den Indoeuropæiske sprogfamilie.
04:27
So some people believe that the Indus script
103
252000
2000
Så nogle mener altså at Indusskriften
04:29
represents an ancient Indo-European language such as Sanskrit.
104
254000
3000
repræsenterer en gammelt Indoeuropæisk sprog
såsom Sanskrit.
såsom Sanskrit.
04:32
There's a last group of people
105
257000
2000
Den sidste gruppe
04:34
who believe that the Indus people
106
259000
3000
mener at Indusfolket
04:37
were the ancestors of people living in South India today.
107
262000
3000
var forfædrene til befolkningen i nutidens Sydindien.
04:40
These people believe that the Indus script
108
265000
2000
De mener at Indusskriften
04:42
represents an ancient form
109
267000
2000
repræsenterer en gammel form for
04:44
of the Dravidian language family,
110
269000
2000
den Dravidianske sprogfamilie
04:46
which is the language family spoken in much of South India today.
111
271000
3000
som er den sprogfamilie, som tales i store dele af Syd Indien i dag.
04:49
And the proponents of this theory
112
274000
2000
Og fortalerne for denne teori
04:51
point to that small pocket of Dravidian-speaking people in the North,
113
276000
3000
peger på den lille lomme
af Dravidiantalende folk.
af Dravidiantalende folk.
04:54
actually near Afghanistan,
114
279000
2000
nær Afghanistan.
04:56
and they say that perhaps, sometime in the past,
115
281000
3000
De siger at man måske engang
04:59
Dravidian languages were spoken all over India
116
284000
3000
talte Dravidianske sprog over hele Indien,
05:02
and that this suggests
117
287000
2000
og det antyder
05:04
that the Indus civilization is perhaps also Dravidian.
118
289000
3000
at Indus-civilisationen måske også er Dravidiansk.
05:07
Which of these hypotheses can be true?
119
292000
3000
Hvilken af disse hypoteser er rigtig ?
05:10
We don't know, but perhaps if you deciphered the script,
120
295000
2000
Vi ved det ikke, men hvis man kunne dechifrere skriften,
05:12
you would be able to answer this question.
121
297000
2000
så ville man måske kunne besvaret spørgsmålet.
05:14
But deciphering the script is a very challenging task.
122
299000
2000
Men dechifreringen af skriften
er en meget udfordrende opgave.
er en meget udfordrende opgave.
05:16
First, there's no Rosetta Stone.
123
301000
2000
For det første er der ikke nogen "Rosetta sten".
05:18
I don't mean the software;
124
303000
2000
Jeg tænker ikke på softwaren.
05:20
I mean an ancient artifact
125
305000
2000
Jeg tænker på en historisk genstand,
05:22
that contains in the same text
126
307000
2000
som indeholder samme tekst
05:24
both a known text and an unknown text.
127
309000
3000
på både et kendt og et ukendt sprog.
05:27
We don't have such an artifact for the Indus script.
128
312000
3000
Vi har ikke sådan en genstand for Indusskriften.
05:30
And furthermore, we don't even know what language they spoke.
129
315000
3000
Og vi ved ikke engang hvilket sprog de talte.
05:33
And to make matters even worse,
130
318000
2000
Og for at gøre sagen endnu værre,
05:35
most of the text that we have are extremely short.
131
320000
2000
så er de fleste tekster ekstremt korte.
05:37
So as I showed you, they're usually found on these seals
132
322000
2000
Som jeg viste jer, så findes de
som regel på disse segl,
som regel på disse segl,
05:39
that are very, very tiny.
133
324000
2000
som er meget meget små.
05:41
And so given these formidable obstacles,
134
326000
2000
I betragtning af disse formidable forhindringer,
05:43
one might wonder and worry
135
328000
2000
kan man spørge
05:45
whether one will ever be able to decipher the Indus script.
136
330000
3000
om man overhovedet vil blive i stand til at dechifrere
Indus-skriften
Indus-skriften
05:48
In the rest of my talk,
137
333000
2000
I resten af mit foredrag
05:50
I'd like to tell you about how I learned to stop worrying
138
335000
2000
vil jeg fortælle om, hvordan jeg lærte
at overvinde denne frygt
at overvinde denne frygt
05:52
and love the challenge posed by the Indus script.
139
337000
2000
og i stedet holde af udfordringen
som Indusskriften gav.
som Indusskriften gav.
05:54
I've always been fascinated by the Indus script
140
339000
3000
Jeg har altid været fascineret af Indusskriften
05:57
ever since I read about it in a middle school textbook.
141
342000
2000
siden jeg læste om den i skolebog.
05:59
And why was I fascinated?
142
344000
2000
Hvorfor var jeg fascineret ?
06:01
Well it's the last major undeciphered script in the ancient world.
143
346000
4000
Det er den sidste større ikke-decifrerede skrift
i den forhistoriske verden.
i den forhistoriske verden.
06:05
My career path led me to become a computational neuroscientist,
144
350000
3000
Min karrieres vej førte mig til at blive
computer-neuro-forsker
computer-neuro-forsker
06:08
so in my day job,
145
353000
2000
I mit daglige arbejde
06:10
I create computer models of the brain
146
355000
2000
laver jeg computermodeller af hjernen
06:12
to try to understand how the brain makes predictions,
147
357000
3000
for at forstå hvordan hjernen forudsiger ting,
06:15
how the brain makes decisions,
148
360000
2000
hvordan hjernen tager beslutninger,
06:17
how the brain learns and so on.
149
362000
2000
hvordan hjernen lærer og så videre.
06:19
But in 2007, my path crossed again with the Indus script.
150
364000
3000
Men i 2007, krydsedes min vej igen med Indusskriften.
06:22
That's when I was in India,
151
367000
2000
Det var da jeg var i Indien,
06:24
and I had the wonderful opportunity
152
369000
2000
og jeg havde den dejlige mulighed
06:26
to meet with some Indian scientists
153
371000
2000
for at mødes med nogle indiske forskere,
06:28
who were using computer models to try to analyze the script.
154
373000
3000
som brugte computere til at analysere skriften.
06:31
And so it was then that I realized
155
376000
2000
Da erkendte jeg
06:33
there was an opportunity for me to collaborate with these scientists,
156
378000
3000
at der var en mulighed for at samarbejde
med disse forskere,
med disse forskere,
06:36
and so I jumped at that opportunity.
157
381000
2000
så jeg tog chancen.
06:38
And I'd like to describe some of the results that we have found.
158
383000
2000
Jeg vil fortælle nogle af de resultater som vi har fået,
06:40
Or better yet, let's all collectively decipher.
159
385000
3000
eller snarere: lad os alle dechifrere sammen.
06:43
Are you ready?
160
388000
2000
Er I klar ?
06:45
The first thing that you need to do when you have an undeciphered script
161
390000
3000
Det første, som man skal gøre,
når man har en ikke-decifreret tekst
når man har en ikke-decifreret tekst
06:48
is try to figure out the direction of writing.
162
393000
2000
er at finde ud af skriveretningen.
06:50
Here are two texts that contain some symbols on them.
163
395000
3000
Her er to tekster som indeholder nogle symboler på dem.
06:53
Can you tell me
164
398000
2000
Kan I fortælle mig
06:55
if the direction of writing is right to left or left to right?
165
400000
3000
om retningen er højre-mod-venstre
eller venstre-mod-højre ?
eller venstre-mod-højre ?
06:58
I'll give you a couple of seconds.
166
403000
3000
I får nogle sekunder.
07:01
Okay. Right to left, how many? Okay.
167
406000
3000
OK. Højre-mod-venstre. Hvor mange ? OK.
07:04
Okay. Left to right?
168
409000
2000
Venstre-mod-højre ?
07:06
Oh, it's almost 50/50. Okay.
169
411000
2000
Oh - det er næsten 50/50. OK
07:08
The answer is:
170
413000
2000
Svaret er:
07:10
if you look at the left-hand side of the two texts,
171
415000
2000
hvis I ser på den venstre side af de to tekster,
07:12
you'll notice that there's a cramping of signs,
172
417000
3000
ser I at der en sammenklemning af tegnene,
07:15
and it seems like 4,000 years ago,
173
420000
2000
og det lader til at for 4000 år siden,
07:17
when the scribe was writing from right to left,
174
422000
2000
da skriveren skrev fra højre mod venstre,
07:19
they ran out of space.
175
424000
2000
at han løb tør for plads.
07:21
And so they had to cram the sign.
176
426000
2000
Så de var nødt til at klemme tegnene.
07:23
One of the signs is also below the text on the top.
177
428000
2000
Een af tegnene er også under teksten øverst.
07:25
This suggests the direction of writing
178
430000
2000
Det lader formode, at skriveretningen
07:27
was probably from right to left,
179
432000
2000
sikkert var fra højre mod venstre,
07:29
and so that's one of the first things we know,
180
434000
2000
Det er en af de første ting vi ved,
07:31
that directionality is a very key aspect of linguistic scripts.
181
436000
3000
at retningen er et væsentligt aspekt
ved lingvistiske skrifter.
ved lingvistiske skrifter.
07:34
And the Indus script now has
182
439000
2000
Og Indusskriften har altså
07:36
this particular property.
183
441000
2000
denne egenskab.
07:38
What other properties of language does the script show?
184
443000
2000
Hvilke andre sprogegenskaber udviser skriften ?
07:40
Languages contain patterns.
185
445000
2000
Sprog indeholder mønstre.
07:42
If I give you the letter Q
186
447000
2000
Hvis jeg giver jer bogstavet Q,
07:44
and ask you to predict the next letter, what do you think that would be?
187
449000
3000
og beder jer forudsige hvad det næste bogstav ville være,
hvad tror I så, det ville være?
hvad tror I så, det ville være?
07:47
Most of you said U, which is right.
188
452000
2000
De fleste af jer sagde U, hvilket er rigtigt.
07:49
Now if I asked you to predict one more letter,
189
454000
2000
Men hvis jeg bad jer om at forudsige eet bogstag til,
07:51
what do you think that would be?
190
456000
2000
hvad tror I så det ville være ?
07:53
Now there's several thoughts. There's E. It could be I. It could be A,
191
458000
3000
Der er flere muligheder Det kunne være E.
Det kunne være I. Det kunne være A,
Det kunne være I. Det kunne være A,
07:56
but certainly not B, C or D, right?
192
461000
3000
men bestemt ikke B, C eller D - vel ?
07:59
The Indus script also exhibits similar kinds of patterns.
193
464000
3000
Indusskriften udviser også lignende mønstre.
08:02
There's a lot of text that start with this diamond-shaped symbol.
194
467000
3000
En stor mængde tekst begynder med
dette ruder-formet symbol.
dette ruder-formet symbol.
08:05
And this in turn tends to be followed
195
470000
2000
Og det følges hyppigt af
08:07
by this quotation marks-like symbol.
196
472000
2000
dette spørgsmålstegn lignende symbol.
08:09
And this is very similar to a Q and U example.
197
474000
2000
Det er helt ligesom Q og U eksemplet.
08:11
This symbol can in turn be followed
198
476000
2000
Dette symbol kan så igen følges
08:13
by these fish-like symbols and some other signs,
199
478000
3000
af disse fiskelignende symboler og nogle andre,
08:16
but never by these other signs at the bottom.
200
481000
2000
men aldrig af disse andre tegn i bunden.
08:18
And furthermore, there's some signs
201
483000
2000
Ydermere er der nogle tegn,
08:20
that really prefer the end of texts,
202
485000
2000
som "foretrækker" slutningen af en tekst,
08:22
such as this jar-shaped sign,
203
487000
2000
såsom dette krukkeformede tegn,
08:24
and this sign, in fact, happens to be
204
489000
2000
og dette tegn viser sig at være
08:26
the most frequently occurring sign in the script.
205
491000
2000
det mest hyppigt forekommende tegn i skriften.
08:28
Given such patterns, here was our idea.
206
493000
3000
Sådanne mønstre bragte os på end ide.
08:31
The idea was to use a computer
207
496000
2000
Ideen var at bruge en computer
08:33
to learn these patterns,
208
498000
2000
til at lære disse mønstre,
08:35
and so we gave the computer the existing texts.
209
500000
3000
og derfor gav vi computeren de eksisterende tekster.
08:38
And the computer learned a statistical model
210
503000
2000
Og computeren lærte en statistisk model
08:40
of which symbols tend to occur together
211
505000
2000
over hvilke symboler der optræder sammen
08:42
and which symbols tend to follow each other.
212
507000
2000
og hvilke symboler som følger efter hinanden.
08:44
Given the computer model,
213
509000
2000
Med denne computermodel,
08:46
we can test the model by essentially quizzing it.
214
511000
3000
kan vi afprøve modellen ved simpelt at spørge den.
08:49
So we could deliberately erase some symbols,
215
514000
2000
Vi kunne fjerne nogle symboler med vilje,
08:51
and we can ask it to predict the missing symbols.
216
516000
3000
og vi kunne bede den om at forudsige de manglende symboler.
08:54
Here are some examples.
217
519000
3000
Her er nogle eksempler.
09:00
You may regard this
218
525000
2000
Man kan betragte dette
09:02
as perhaps the most ancient game
219
527000
2000
som måske det ældste forhistoriske
09:04
of Wheel of Fortune.
220
529000
3000
lykkehjul.
09:08
What we found
221
533000
2000
Vi fandt
09:10
was that the computer was successful in 75 percent of the cases
222
535000
2000
at computeren var korrekt i 75% af tilfældene
09:12
in predicting the correct symbol.
223
537000
2000
ved forudsigelse af det korrekte symbol.
09:14
In the rest of the cases,
224
539000
2000
I resten af tilfældene,
09:16
typically the second best guess or third best guess was the right answer.
225
541000
3000
var det næstbedste eller det tredjebedste gæt
det rigtige svar.
det rigtige svar.
09:19
There's also practical use
226
544000
2000
Der er også en praktisk anvendelse
09:21
for this particular procedure.
227
546000
2000
af denne fremgangsmåde.
09:23
There's a lot of these texts that are damaged.
228
548000
2000
Der findes en masse tekster, som er beskadiget.
09:25
Here's an example of one such text.
229
550000
2000
Her er et eksempel.
09:27
And we can use the computer model now to try to complete this text
230
552000
3000
Vi kan bruge computer-modellen til at forsøge
at gøre teksten komplet,
at gøre teksten komplet,
09:30
and make a best guess prediction.
231
555000
2000
ved at fremkomme ved det bedste gæt.
09:32
Here's an example of a symbol that was predicted.
232
557000
3000
Her er et eksempel på, et symbol som blev forudsagt.
09:35
And this could be really useful as we try to decipher the script
233
560000
2000
Det kunne være rigtig brugbart
når vi forsøger at decifrere teksten
når vi forsøger at decifrere teksten
09:37
by generating more data that we can analyze.
234
562000
3000
ved at frembringe mere data, som vi kan analysere.
09:40
Now here's one other thing you can do with the computer model.
235
565000
3000
Her er noget andet
som man kan gøre med computer-modellen.
som man kan gøre med computer-modellen.
09:43
So imagine a monkey
236
568000
2000
Forestil jer en abe
09:45
sitting at a keyboard.
237
570000
2000
som sidder på et tastatur.
09:47
I think you might get a random jumble of letters that looks like this.
238
572000
3000
Man vil få en tilfældig smøre af bogstaver,
som kunne se sådan ud.
som kunne se sådan ud.
09:50
Such a random jumble of letters
239
575000
2000
Sådan en rodet tekst
09:52
is said to have a very high entropy.
240
577000
2000
siges at have en høj entropi.
09:54
This is a physics and information theory term.
241
579000
2000
Det er et fysik- og informationsteoretisk udtryk.
09:56
But just imagine it's a really random jumble of letters.
242
581000
3000
Men forestil jer bare en rodet bunke bogstaver.
09:59
How many of you have ever spilled coffee on a keyboard?
243
584000
4000
Hvor mange af jer har hældt kaffen udover tastaturet?
10:03
You might have encountered the stuck-key problem --
244
588000
2000
Så har I måske set problemet med den fastsiddende tast.
10:05
so basically the same symbol being repeated over and over again.
245
590000
3000
Det samme tegn bare gentaget igen og igen.
10:08
This kind of a sequence is said to have a very low entropy
246
593000
3000
Den slags sekvenser siges at have en lav entropi
10:11
because there's no variation at all.
247
596000
2000
fordi der ikke er nogen variation overhovedet.
10:13
Language, on the other hand, has an intermediate level of entropy;
248
598000
3000
Sprog har, på den anden side, et mellemniveau af entropi.
10:16
it's neither too rigid,
249
601000
2000
Den er hverken for monoton,
10:18
nor is it too random.
250
603000
2000
eller for tilfældig
10:20
What about the Indus script?
251
605000
2000
Og Indusskriften ?
10:22
Here's a graph that plots the entropies of a whole bunch of sequences.
252
607000
4000
Her er en graf som viser entropien for en bunke sekvenser.
10:26
At the very top you find the uniformly random sequence,
253
611000
2000
Øverst ser du den tilfældigt dannede sekvens
10:28
which is a random jumble of letters --
254
613000
2000
som er et tilfældigt virvar af bogstaver,
10:30
and interestingly, we also find
255
615000
2000
og interessant nok, finder vi også
10:32
the DNA sequence from the human genome and instrumental music.
256
617000
3000
menneskets DNA-sekvens og musik.
10:35
And both of these are very, very flexible,
257
620000
2000
De er begge meget foranderlige,
10:37
which is why you find them in the very high range.
258
622000
2000
hvorfor man ser dem øverst i feltet.
10:39
At the lower end of the scale,
259
624000
2000
Nederst i feltet
10:41
you find a rigid sequence, a sequence of all A's,
260
626000
2000
ses en fastlåst sekvens. En sekvens af A'er,
10:43
and you also find a computer program,
261
628000
2000
og man ser et computerprogram
10:45
in this case in the language Fortran,
262
630000
2000
som her er i Fortran-sproget,
10:47
which obeys really strict rules.
263
632000
2000
som følger meget stramme regler.
10:49
Linguistic scripts
264
634000
2000
Lingvistiske tekster
10:51
occupy the middle range.
265
636000
2000
ligger i midterfeltet.
10:53
Now what about the Indus script?
266
638000
2000
Hvad med Industeksten ?
10:55
We found that the Indus script
267
640000
2000
Vi fandt at Industeksten
10:57
actually falls within the range of the linguistic scripts.
268
642000
2000
faktisk ligger i området med lingvistiske tekster.
10:59
When this result was first published,
269
644000
2000
Da dette resultat blev offentliggjort første gang
11:01
it was highly controversial.
270
646000
3000
var det yderst kontroversielt.
11:04
There were people who raised a hue and cry,
271
649000
3000
Der var nogle som protesterede voldsomt
11:07
and these people were the ones who believed
272
652000
2000
og det var dem, som mente
11:09
that the Indus script does not represent language.
273
654000
3000
at Indusskriften ikke repræsenterer noget sprog.
11:12
I even started to get some hate mail.
274
657000
2000
Jeg fik endda nogle hademails.
11:14
My students said
275
659000
2000
Mine studerende sagde,
11:16
that I should really seriously consider getting some protection.
276
661000
3000
at jeg skulle overveje at få beskyttelse.
11:19
Who'd have thought
277
664000
2000
Hvem ville have troet
11:21
that deciphering could be a dangerous profession?
278
666000
2000
at decifrering kunne være en farlig profession ?
11:23
What does this result really show?
279
668000
2000
Hvad viser dette resultat i virkeligheden ?
11:25
It shows that the Indus script
280
670000
2000
Det viser, at Indusskriften
11:27
shares an important property of language.
281
672000
2000
har en vigtig egenskab for et sprog.
11:29
So, as the old saying goes,
282
674000
2000
Som man plejer at sige,
11:31
if it looks like a linguistic script
283
676000
2000
hvis det ligner et lingvistisk skrift,
11:33
and it acts like a linguistic script,
284
678000
2000
og opfører sig som et lingvistisk skrift,
11:35
then perhaps we may have a linguistic script on our hands.
285
680000
3000
så har man nok også et lingvistisk skrift.
11:38
What other evidence is there
286
683000
2000
Hvilke andre beviser er der
11:40
that the script could actually encode language?
287
685000
2000
for at skriften faktisk koder for et sprog ?
11:42
Well linguistic scripts can actually encode multiple languages.
288
687000
3000
Lingvistiske skrifter kan kode adskillige sprog.
11:45
So for example, here's the same sentence written in English
289
690000
3000
For eksempel er der her den samme
tekst på engelsk
tekst på engelsk
11:48
and the same sentence written in Dutch
290
693000
2000
og den samme sætning på hollandsk,
11:50
using the same letters of the alphabet.
291
695000
2000
og som bruger det samme alfabet.
11:52
If you don't know Dutch and you only know English
292
697000
3000
Hvis I ikke kender til hollandsk
og I kun kender til engelsk
og I kun kender til engelsk
11:55
and I give you some words in Dutch,
293
700000
2000
og jeg giver jer nogle ord på hollandsk,
11:57
you'll tell me that these words contain
294
702000
2000
så vil I sige at disse ord indeholder
11:59
some very unusual patterns.
295
704000
2000
nogle meget usædvanlige mønstre.
12:01
Some things are not right,
296
706000
2000
Noget stemmer ikke,
12:03
and you'll say these words are probably not English words.
297
708000
3000
og I vil sige, at disse ord nok ikke er engelske ord.
12:06
The same thing happens in the case of the Indus script.
298
711000
2000
Det samme er tilfældet med Indusskriften.
12:08
The computer found several texts --
299
713000
2000
Computeren fandt adskillige tekster,
12:10
two of them are shown here --
300
715000
2000
to af dem vises her
12:12
that have very unusual patterns.
301
717000
2000
som meget usædvanlige mønstre.
12:14
So for example the first text:
302
719000
2000
For eksempel den første tekst:
12:16
there's a doubling of this jar-shaped sign.
303
721000
3000
krukke-tegnet står dobbelt.
12:19
This sign is the most frequently-occurring sign
304
724000
2000
Dette tegn er det hyppigste
12:21
in the Indus script,
305
726000
2000
i Indusskriften,
12:23
and it's only in this text
306
728000
2000
og det er kun i denne tekst
12:25
that it occurs as a doubling pair.
307
730000
2000
at det optræder dobbelt.
12:27
Why is that the case?
308
732000
2000
Hvorfor var det sådan ?
12:29
We went back and looked at where these particular texts were found,
309
734000
3000
Vi undersøgte, hvor disse tekster kom fra,
12:32
and it turns out that they were found
310
737000
2000
og det viste sig, at de blev fundet
12:34
very, very far away from the Indus Valley.
311
739000
2000
langt, langt væk fra Indusdalen.
12:36
They were found in present day Iraq and Iran.
312
741000
3000
De blev fundet i det nutidige Irak og Iran.
12:39
And why were they found there?
313
744000
2000
Hvorfor blev de fundet der ?
12:41
What I haven't told you is that
314
746000
2000
Jeg fik ikke fortalt
12:43
the Indus people were very, very enterprising.
315
748000
2000
at Indusfolket var meget handelsmindede.
12:45
They used to trade with people pretty far away from where they lived,
316
750000
3000
De handlede med folk et langvejs fra, hvor de boede,
12:48
and so in this case, they were traveling by sea
317
753000
3000
og i dette tilfælde, rejste de over havet,
12:51
all the way to Mesopotamia, present-day Iraq.
318
756000
3000
helt til Mesopotamien, det nutidige Irak.
12:54
And what seems to have happened here
319
759000
2000
Det lader til
12:56
is that the Indus traders, the merchants,
320
761000
3000
at Induskøbmændene
12:59
were using this script to write a foreign language.
321
764000
3000
brugte denne skrift til at skrive et fremmed sprog.
13:02
It's just like our English and Dutch example.
322
767000
2000
Ligesom eksemplet med engelsk og hollandsk.
13:04
And that would explain why we have these strange patterns
323
769000
2000
Det kunne forklare disse mærkelige mønstre,
13:06
that are very different from the kinds of patterns you see in the text
324
771000
3000
som er meget forskellig fra mønstrene,
13:09
that are found within the Indus Valley.
325
774000
3000
som fandtes i Indusdalen.
13:12
This suggests that the same script, the Indus script,
326
777000
2000
Det lader antyde, at Indusskriften
13:14
could be used to write different languages.
327
779000
3000
kunne bruges til at skrive forskellige sprog.
13:17
The results we have so far seem to point to the conclusion
328
782000
3000
De resultater vi har indtil nu,
synes at pege på den konklusion
synes at pege på den konklusion
13:20
that the Indus script probably does represent language.
329
785000
3000
at Indus-skriften sandsynligvis repræsenterer et sprog.
13:23
If it does represent language,
330
788000
2000
Hvis det ikke repræsenterer et sprog,
13:25
then how do we read the symbols?
331
790000
2000
hvordan skal vi så læse symbolerne ?
13:27
That's our next big challenge.
332
792000
2000
Det er vor næste store udfordring.
13:29
So you'll notice that many of the symbols
333
794000
2000
Man vil bemærke, at mange symboler
13:31
look like pictures of humans, of insects,
334
796000
2000
ligner tegninger af mennesker, insekter,
13:33
of fishes, of birds.
335
798000
3000
fisk, fugle.
13:36
Most ancient scripts
336
801000
2000
De fleste forhistoriske skrifter
13:38
use the rebus principle,
337
803000
2000
bruger rebusprincippet,
13:40
which is, using pictures to represent words.
338
805000
3000
hvor man bruger billeder til at stå for ord.
13:43
So as an example, here's a word.
339
808000
3000
For eksempel
13:46
Can you write it using pictures?
340
811000
2000
Kan I skrive det ved brug af tegninger ?
13:48
I'll give you a couple seconds.
341
813000
2000
I får et par sekunder.
13:50
Got it?
342
815000
2000
Ja ?
13:52
Okay. Great.
343
817000
2000
OK.
13:54
Here's my solution.
344
819000
2000
Her er min løsning.
13:56
You could use the picture of a bee followed by a picture of a leaf --
345
821000
2000
Man kan bruge tegningen af en bi
efterfulgt af en tegning af et blad,
efterfulgt af en tegning af et blad,
13:58
and that's "belief," right.
346
823000
2000
og så står der på engelsk - "belief"
14:00
There could be other solutions.
347
825000
2000
Der er sikkert andre løsninger.
14:02
In the case of the Indus script,
348
827000
2000
I tilfældet med Indusskriften,
14:04
the problem is the reverse.
349
829000
2000
er problemet det omvendte.
14:06
You have to figure out the sounds of each of these pictures
350
831000
3000
Man er nødt til at regne lydene ud
for hvert af disse tegninger,
for hvert af disse tegninger,
14:09
such that the entire sequence makes sense.
351
834000
2000
sådan at det hele sekvensen giver mening.
14:11
So this is just like a crossword puzzle,
352
836000
3000
Det er ligesom krydsord,
14:14
except that this is the mother of all crossword puzzles
353
839000
3000
med den undtagelse at dette
er alle krydsords moder
er alle krydsords moder
14:17
because the stakes are so high if you solve it.
354
842000
4000
fordi gevinsten er så stor, hvis man løser den.
14:21
My colleagues, Iravatham Mahadevan and Asko Parpola,
355
846000
3000
Mine kolleger Iravatham Mahadevan og Asko Parpola
14:24
have been making some headway on this particular problem.
356
849000
2000
har gjort nogle fremskridt.
14:26
And I'd like to give you a quick example of Parpola's work.
357
851000
2000
Jeg vil vise jer et eksempel på Parpola's arbejde.
14:28
Here's a really short text.
358
853000
2000
Her er en rigtig kort tekst.
14:30
It contains seven vertical strokes followed by this fish-like sign.
359
855000
3000
Det indeholder 7 lodrette streger
efterfulgt af et fisketegn.
efterfulgt af et fisketegn.
14:33
And I want to mention that these seals were used
360
858000
2000
Jeg skal nævne, at disse segl blev brugt
14:35
for stamping clay tags
361
860000
2000
til at præge lermærkater,
14:37
that were attached to bundles of goods,
362
862000
2000
som blev fastgjort til varebundter,
14:39
so it's quite likely that these tags, at least some of them,
363
864000
3000
så det er sandsynligt
at nogle af disse mærkater,
at nogle af disse mærkater,
14:42
contain names of merchants.
364
867000
2000
indeholder købmændenes navne.
14:44
And it turns out that in India
365
869000
2000
Det viser sig, at der i Indien
14:46
there's a long tradition
366
871000
2000
er en lang tradition
14:48
of names being based on horoscopes
367
873000
2000
for navne som er baseret på stjernebilleder,
14:50
and star constellations present at the time of birth.
368
875000
3000
og en stjernekonstellation står for fødselstidspunktet.
14:53
In Dravidian languages,
369
878000
2000
I dravidianske sprog,
14:55
the word for fish is "meen"
370
880000
2000
er ordet for fisk "meen"
14:57
which happens to sound just like the word for star.
371
882000
3000
som lyder, som ordet for stjerne.
15:00
And so seven stars
372
885000
2000
Og syv stjerner
15:02
would stand for "elu meen,"
373
887000
2000
kunne stå for "elu meen"
15:04
which is the Dravidian word
374
889000
2000
som er det davidianske ord
15:06
for the Big Dipper star constellation.
375
891000
2000
for Karlsvognen.
15:08
Similarly, there's another sequence of six stars,
376
893000
3000
På samme måde, er der en anden sekvens
på seks stjerner
på seks stjerner
15:11
and that translates to "aru meen,"
377
896000
2000
som kan oversættes to "aru meen"
15:13
which is the old Dravidian name
378
898000
2000
som er det gamle dravidianske ord
15:15
for the star constellation Pleiades.
379
900000
2000
for stjerne-konstellationen Plejaderne.
15:17
And finally, there's other combinations,
380
902000
3000
Og til sidst er der andre kombinationer,
15:20
such as this fish sign with something that looks like a roof on top of it.
381
905000
3000
såsom denne fisk med noget
som ligner et tag ovenpå den.
som ligner et tag ovenpå den.
15:23
And that could be translated into "mey meen,"
382
908000
3000
Det kunne oversættes til "mey meen",
15:26
which is the old Dravidian name for the planet Saturn.
383
911000
3000
som er det gamle dravidianske ord for Saturn.
15:29
So that was pretty exciting.
384
914000
2000
Det var ret spændende.
15:31
It looks like we're getting somewhere.
385
916000
2000
Det lader til, at vi kommer fremad.
15:33
But does this prove
386
918000
2000
Men beviser dette
15:35
that these seals contain Dravidian names
387
920000
2000
at disse segl indeholder dravidianske navne
15:37
based on planets and star constellations?
388
922000
2000
baseret på stjernekonstellationer ?
15:39
Well not yet.
389
924000
2000
Ikke endnu.
15:41
So we have no way of validating
390
926000
2000
Vi kan ikke på nogen måde bekræfte
15:43
these particular readings,
391
928000
2000
disse tolkninger,
15:45
but if more and more of these readings start making sense,
392
930000
3000
men hvis flere og flere af sådanne tolkninger
giver mening
giver mening
15:48
and if longer and longer sequences
393
933000
2000
og hvis længere og længere sekvenser
15:50
appear to be correct,
394
935000
2000
synes at være korrekt,
15:52
then we know that we are on the right track.
395
937000
2000
så ved vi, at vi er på rette spor.
15:54
Today,
396
939000
2000
I dag
15:56
we can write a word such as TED
397
941000
3000
kan vi skrive ord såsom TED
15:59
in Egyptian hieroglyphics and in cuneiform script,
398
944000
3000
med egyptiske hieroglyffer og i kileskrift,
16:02
because both of these were deciphered
399
947000
2000
fordi disse to blev dechifreret
16:04
in the 19th century.
400
949000
2000
i det 19. århundrede.
16:06
The decipherment of these two scripts
401
951000
2000
Dechifreringen af disse to skrifter
16:08
enabled these civilizations to speak to us again directly.
402
953000
3000
gjorde det muligt for disse civilisationer
at tale direkte til os.
at tale direkte til os.
16:11
The Mayans
403
956000
2000
Maya'erne
16:13
started speaking to us in the 20th century,
404
958000
2000
begyndte at tale til os i det 20. århundrede,
16:15
but the Indus civilization remains silent.
405
960000
3000
men Induscivilisationen er stadig tavs.
16:18
Why should we care?
406
963000
2000
Betyder det noget ?
16:20
The Indus civilization does not belong
407
965000
2000
Induscivilisationen tilhører ikke bare
16:22
to just the South Indians or the North Indians
408
967000
2000
sydindere eller nordindere
16:24
or the Pakistanis;
409
969000
2000
eller pakistanere.
16:26
it belongs to all of us.
410
971000
2000
Den tilhører os alle.
16:28
These are our ancestors --
411
973000
2000
De er vore forfædre
16:30
yours and mine.
412
975000
2000
dine og mine.
16:32
They were silenced
413
977000
2000
De blev gjort tavse
16:34
by an unfortunate accident of history.
414
979000
2000
af en uheldig begivenhed i historien.
16:36
If we decipher the script,
415
981000
2000
Hvis vi dechifrerer skriften
16:38
we would enable them to speak to us again.
416
983000
2000
kan vi få dem til at tale til os igen.
16:40
What would they tell us?
417
985000
3000
Hvad vil de fortælle os ?
16:43
What would we find out about them? About us?
418
988000
3000
Hvad kan vi lære om dem ? Om os ?
16:46
I can't wait to find out.
419
991000
3000
Jeg kan ikke vente.
16:49
Thank you.
420
994000
2000
Tak.
16:51
(Applause)
421
996000
4000
(Applaus)
ABOUT THE SPEAKER
Rajesh Rao - Computational neuroscientistRajesh Rao seeks to understand the human brain through computational modeling, on two fronts: developing computer models of our minds, and using tech to decipher the 4,000-year-old lost script of the Indus Valley civilization.
Why you should listen
Rajesh Rao is looking for the computational principles underlying the brain's remarkable ability to learn, process and store information -- hoping to apply this knowledge to the task of building adaptive robotic systems and artificially intelligent agents.
Some of the questions that motivate his research include: How does the brain learn efficient representations of novel objects and events occurring in the natural environment? What are the algorithms that allow useful sensorimotor routines and behaviors to be learned? What computational mechanisms allow the brain to adapt to changing circumstances and remain fault-tolerant and robust?
By investigating these questions within a computational and probabilistic framework, it is often possible to derive algorithms that not only provide functional interpretations of neurobiological properties but also suggest solutions to difficult problems in computer vision, speech, robotics and artificial intelligence.
Rajesh Rao | Speaker | TED.com