1
00:00:00,300 --> 00:00:02,800
There's been a heist
at the Royal Institution,

2
00:00:02,800 --> 00:00:05,500
and a man-made diamond was stolen.

3
00:00:05,500 --> 00:00:07,520
That's the moment caught

4
00:00:07,520 --> 00:00:11,280
on the Royal Institution's own
CCTV cameras.

5
00:00:11,280 --> 00:00:13,680
Tonight's audience,
when you were queuing there,

6
00:00:13,680 --> 00:00:16,640
you were witnesses to a crime,

7
00:00:16,640 --> 00:00:19,000
and the thief has escaped.

8
00:00:57,200 --> 00:01:01,160
APPLAUSE

9
00:01:09,320 --> 00:01:14,200
This is the third and final
of this year's Christmas lectures

10
00:01:14,200 --> 00:01:15,640
from the Royal Institution.

11
00:01:15,640 --> 00:01:19,920
And these lectures have explored
the principles of forensic science.

12
00:01:19,920 --> 00:01:22,720
And through the course
of this final lecture,

13
00:01:22,720 --> 00:01:26,520
we're going to try to identify
and to bring to justice

14
00:01:26,520 --> 00:01:29,960
a living body, a jewel thief.

15
00:01:31,240 --> 00:01:34,520
Here in the court
of the Royal Institution

16
00:01:34,520 --> 00:01:38,480
we will put on trial a suspect
accused of stealing

17
00:01:38,480 --> 00:01:41,520
the precious man-made diamond.

18
00:01:41,520 --> 00:01:46,800
So the sole purpose of forensic
evidence is to assist the jury

19
00:01:46,800 --> 00:01:50,320
in their decision-making
in the courtroom.

20
00:01:50,320 --> 00:01:54,640
And at the end of this lecture,
you will be the jury,

21
00:01:54,640 --> 00:01:58,440
and you will be deciding
whether the suspect we have

22
00:01:58,440 --> 00:02:02,760
is guilty of the crime
for which they have been accused.

23
00:02:02,760 --> 00:02:06,080
Now, as a forensic anthropologist,
I often appear in court

24
00:02:06,080 --> 00:02:07,800
as an expert witness.

25
00:02:07,800 --> 00:02:12,600
And standing here in a witness box,
as an expert witness, like myself,

26
00:02:12,600 --> 00:02:17,440
we can offer opinions based solely
on our professional judgment.

27
00:02:17,440 --> 00:02:20,240
And what do we do?
We take it very seriously.

28
00:02:20,240 --> 00:02:22,640
We swear to tell the truth,

29
00:02:22,640 --> 00:02:26,360
the whole truth
and nothing but the truth.

30
00:02:26,360 --> 00:02:29,240
They're really important words
in there.

31
00:02:29,240 --> 00:02:33,600
And some of the most influential
figures in any courtroom

32
00:02:33,600 --> 00:02:38,320
are the lawyers. And tonight,
our defence lawyer is probably -

33
00:02:38,320 --> 00:02:41,160
no, he is -
one of the most respected,

34
00:02:41,160 --> 00:02:46,400
and definitely one of the most
feared advocates in Scotland.

35
00:02:46,400 --> 00:02:50,080
Would you please welcome
Mr Donald Findlay KC?

36
00:02:50,080 --> 00:02:52,920
APPLAUSE

37
00:02:57,840 --> 00:03:01,440
I would say I've had the pleasure,
but it was not a pleasure,

38
00:03:01,440 --> 00:03:04,920
of being cross-examined
by Mr Findlay.

39
00:03:04,920 --> 00:03:08,200
But what I'm going to ask you to do,
if I may, Mr Findlay, is to talk

40
00:03:08,200 --> 00:03:11,640
to what is our future jury,
and what is their role

41
00:03:11,640 --> 00:03:14,200
and their responsibility,
because this is one of the most

42
00:03:14,200 --> 00:03:18,280
important, serious things that you
can do as a member of the public.

43
00:03:18,280 --> 00:03:23,800
In this country, if there is an
allegation that a serious crime

44
00:03:23,800 --> 00:03:28,480
has been committed, and someone
is arrested and charged,

45
00:03:28,480 --> 00:03:31,040
they are entitled
to have that case heard,

46
00:03:31,040 --> 00:03:33,320
not by a judge, not by lawyers,

47
00:03:33,320 --> 00:03:37,520
but by a jury made up
of members of the public.

48
00:03:37,520 --> 00:03:42,680
Tonight, you are our jury,
and that's quite a responsibility.

49
00:03:43,960 --> 00:03:46,960
In order to prove guilt
in any criminal case,

50
00:03:46,960 --> 00:03:50,240
the prosecution leads evidence

51
00:03:50,240 --> 00:03:55,040
to try to establish guilt
beyond a reasonable doubt.

52
00:03:55,040 --> 00:03:57,400
If you have a doubt,

53
00:03:57,400 --> 00:04:00,320
which is a kind of doubt
you would have in your own lives

54
00:04:00,320 --> 00:04:04,240
about an important decision
you were going to take,

55
00:04:04,240 --> 00:04:08,920
then you give the benefit of
that doubt to the person accused.

56
00:04:08,920 --> 00:04:13,000
The prosecution has to prove
its case on the evidence.

57
00:04:13,000 --> 00:04:14,880
It's not about personalities.

58
00:04:14,880 --> 00:04:18,160
It's not about which witnesses you
like more than any other witnesses.

59
00:04:18,160 --> 00:04:21,000
It's about the evidence
and the quality of the evidence

60
00:04:21,000 --> 00:04:22,920
that you find proved.

61
00:04:22,920 --> 00:04:26,000
And it's my job to make it
difficult for the prosecution

62
00:04:26,000 --> 00:04:28,200
to try to prove its case.

63
00:04:28,200 --> 00:04:31,880
If, at the end of it all, you are
left with a reasonable doubt,

64
00:04:31,880 --> 00:04:36,800
then you are bound by law, and bound
by the rules of this evening,

65
00:04:36,800 --> 00:04:39,600
to acquit the person
who is on trial.

66
00:04:39,600 --> 00:04:40,920
Thank you very much indeed.

67
00:04:40,920 --> 00:04:44,200
Mr Findlay is going to join us
and sit through the proceedings,

68
00:04:44,200 --> 00:04:46,720
and you will see him come forward
at different times

69
00:04:46,720 --> 00:04:48,720
to talk to expert witnesses.

70
00:04:48,720 --> 00:04:52,600
I want you to remember that science
isn't black and white,

71
00:04:52,600 --> 00:04:55,160
it is a variety of shades of grey,

72
00:04:55,160 --> 00:04:59,080
and scientists need to know how far
they can go with the evidence.

73
00:04:59,080 --> 00:05:00,800
And when you're an expert witness,

74
00:05:00,800 --> 00:05:04,360
sometimes it feels like the expert
witness is on trial as well.

75
00:05:04,360 --> 00:05:07,840
So it's a very uncomfortable place
for science to be.

76
00:05:07,840 --> 00:05:10,960
But before we start anything,

77
00:05:10,960 --> 00:05:12,760
we need a suspect,

78
00:05:12,760 --> 00:05:18,120
and we're going to need some help
in terms of identifying our suspect.

79
00:05:18,120 --> 00:05:22,160
Now, you'll have seen these sort
of wanted posters before,

80
00:05:22,160 --> 00:05:25,080
and what we're going to want to do
is we're going to fill in

81
00:05:25,080 --> 00:05:27,360
the silhouette,
because who is this person?

82
00:05:27,360 --> 00:05:31,240
And we're going to start by
doing it the old-fashioned way,

83
00:05:31,240 --> 00:05:34,600
perhaps the way you might have
seen it done in films.

84
00:05:34,600 --> 00:05:40,080
If you could all look at the CCTV
footage, which is of the burglary

85
00:05:40,080 --> 00:05:42,560
that occurred earlier this evening,

86
00:05:42,560 --> 00:05:45,680
and it's over so quickly, isn't it?

87
00:05:45,680 --> 00:05:48,160
So we're going to be joined
by a psychologist

88
00:05:48,160 --> 00:05:50,200
from the University of
Central Lancashire

89
00:05:50,200 --> 00:05:53,840
whose team is working to establish
the best procedures

90
00:05:53,840 --> 00:05:58,000
to extract the memories of faces
from eyewitnesses.

91
00:05:58,000 --> 00:06:00,720
So would you please welcome
Professor Charlie Froud?

92
00:06:00,720 --> 00:06:03,160
APPLAUSE

93
00:06:05,160 --> 00:06:06,520
Charlie, what we're going to do

94
00:06:06,520 --> 00:06:08,960
is we're going to do it the
old-fashioned way. We are, yes.

95
00:06:08,960 --> 00:06:10,600
Not what we would necessarily
do today,

96
00:06:10,600 --> 00:06:13,440
in fact definitely not what we do
today. Yes, that's right.

97
00:06:13,440 --> 00:06:17,240
So we're going to try to build
a picture of the thief.

98
00:06:17,240 --> 00:06:19,160
What we have is some features.

99
00:06:19,160 --> 00:06:22,160
We have some eyes, as you can see,
some noses and some mouths.

100
00:06:22,160 --> 00:06:25,200
And what we're going to do
is I'm going to ask you to tell me

101
00:06:25,200 --> 00:06:27,240
which picture you think is the best.

102
00:06:27,240 --> 00:06:28,520
So we have some eyes.

103
00:06:28,520 --> 00:06:31,080
So what I'd like you to do
is I'd like you to call out

104
00:06:31,080 --> 00:06:34,920
if you think, as I point,
you think it's the thief.

105
00:06:34,920 --> 00:06:37,280
What do we think about those?
Yes or no? Yes?

106
00:06:37,280 --> 00:06:38,720
Oh, very quiet.

107
00:06:38,720 --> 00:06:40,160
No - oh, no. Not that one.

108
00:06:40,160 --> 00:06:41,360
What about these?

109
00:06:41,360 --> 00:06:42,680
AUDIENCE: Yes.

110
00:06:42,680 --> 00:06:44,800
That's a little bit better.
And these eyes?

111
00:06:44,800 --> 00:06:46,520
AUDIENCE: Yes.

112
00:06:46,520 --> 00:06:48,560
Maybe about the same.
And these eyes?

113
00:06:48,560 --> 00:06:49,720
AUDIENCE: Yeah.

114
00:06:49,720 --> 00:06:51,400
Bottom set of eyes?

115
00:06:51,400 --> 00:06:52,920
AUDIENCE: No.

116
00:06:52,920 --> 00:06:55,720
No. So maybe it's between these two.

117
00:06:55,720 --> 00:06:58,560
So who thinks - we have to have
another guess -

118
00:06:58,560 --> 00:07:00,440
who thinks these pair of eyes?

119
00:07:00,440 --> 00:07:01,840
AUDIENCE: Yes.

120
00:07:01,840 --> 00:07:03,120
And these?

121
00:07:03,120 --> 00:07:04,280
AUDIENCE: Yes.

122
00:07:04,280 --> 00:07:06,600
It's about the same. Second one.
I'm going to adjudicate.

123
00:07:06,600 --> 00:07:08,480
So let's pick the second one then.

124
00:07:08,480 --> 00:07:10,600
So we've got some eyes
for our photo fit.

125
00:07:10,600 --> 00:07:13,760
They can go in there, and let's
look at some noses, then.

126
00:07:13,760 --> 00:07:16,200
We'll do the same thing.
What about this first nose?

127
00:07:16,200 --> 00:07:17,720
AUDIENCE: No.

128
00:07:17,720 --> 00:07:19,240
No. This nose?

129
00:07:19,240 --> 00:07:20,400
AUDIENCE: No.

130
00:07:20,400 --> 00:07:21,920
No. This nose?

131
00:07:21,920 --> 00:07:23,200
AUDIENCE: No.

132
00:07:23,200 --> 00:07:24,840
Oh, you don't like noses!

133
00:07:24,840 --> 00:07:26,520
Maybe a yes in there with this nose.

134
00:07:26,520 --> 00:07:28,320
AUDIENCE: Yes.

135
00:07:28,320 --> 00:07:30,400
Oh, that's interesting.
And this nose?

136
00:07:30,400 --> 00:07:31,840
AUDIENCE: Yes.

137
00:07:31,840 --> 00:07:35,880
I think it's number four.
Number four? Yes, I think so.

138
00:07:35,880 --> 00:07:37,000
There we go.

139
00:07:37,000 --> 00:07:39,600
So let's do the mouths, then,
same again. This first mouth.

140
00:07:39,600 --> 00:07:40,960
AUDIENCE: No.

141
00:07:40,960 --> 00:07:42,080
This one?

142
00:07:42,080 --> 00:07:43,520
AUDIENCE MURMUR UNCERTAINLY

143
00:07:43,520 --> 00:07:44,640
Perhaps a little bit.

144
00:07:44,640 --> 00:07:46,840
This one?

145
00:07:46,840 --> 00:07:48,760
A bit more perhaps for this one.
This one?

146
00:07:48,760 --> 00:07:50,120
AUDIENCE: Yes.

147
00:07:50,120 --> 00:07:51,320
Ah, maybe. And this one?

148
00:07:51,320 --> 00:07:52,800
AUDIENCE: No.

149
00:07:52,800 --> 00:07:55,040
Oh! So we're going to pick
this mouth. OK.

150
00:07:55,040 --> 00:07:59,240
So this is who you think our suspect
might be.

151
00:07:59,240 --> 00:08:01,680
If we imagine there are glasses
all the way around, of course.

152
00:08:01,680 --> 00:08:03,320
We have to. We need to do that.

153
00:08:03,320 --> 00:08:06,960
But what would we have done
with that image in the past?

154
00:08:06,960 --> 00:08:11,000
So in the past, the police
would show this photo fit to other

155
00:08:11,000 --> 00:08:14,600
police officers and members
of the public for them to recognise.

156
00:08:14,600 --> 00:08:17,520
And the truth is that the technique
doesn't work very well.

157
00:08:17,520 --> 00:08:20,800
So now the field has been
transformed in the last 20 years.

158
00:08:20,800 --> 00:08:23,960
But you spoke to one
of the witnesses tonight.

159
00:08:23,960 --> 00:08:25,240
Now talk us through.

160
00:08:25,240 --> 00:08:28,400
I'm interviewing our young witness
here, and the witness

161
00:08:28,400 --> 00:08:32,360
is selecting faces from the screen,
the best matches for the face,

162
00:08:32,360 --> 00:08:35,440
really in terms of the region
around the eyes, because we know

163
00:08:35,440 --> 00:08:36,960
this is really quite important,

164
00:08:36,960 --> 00:08:39,280
and those choices
were combined together.

165
00:08:39,280 --> 00:08:43,080
And then what she did was she uses
some sliding scales to enhance

166
00:08:43,080 --> 00:08:45,440
the age and the health
of the person.

167
00:08:45,440 --> 00:08:48,680
And then at the end,
she added on some hair,

168
00:08:48,680 --> 00:08:50,680
and actually a hooded top,

169
00:08:50,680 --> 00:08:52,760
as she remembered
what this man looked like.

170
00:08:52,760 --> 00:08:55,640
And the final image, I think,
is up in front of us.

171
00:08:55,640 --> 00:08:59,200
And how many faces can we remember,
do you think?

172
00:08:59,200 --> 00:09:02,760
Recent research suggests
that we remember a lot, Sue.

173
00:09:02,760 --> 00:09:05,480
About 5,000 faces.

174
00:09:05,480 --> 00:09:07,720
So it's a really large memory
we have.

175
00:09:07,720 --> 00:09:10,480
So what we've got here
is we've got an eyewitness

176
00:09:10,480 --> 00:09:14,920
working with you that feels that
this is the best likeness.

177
00:09:14,920 --> 00:09:18,120
We have all these witnesses
in here

178
00:09:18,120 --> 00:09:20,400
who've looked at the CCTV footage,

179
00:09:20,400 --> 00:09:23,800
and they think it looks like
this individual.

180
00:09:23,800 --> 00:09:27,080
What do we have in terms
of any similarities?

181
00:09:27,080 --> 00:09:30,640
I mean, the photo fit was made
in a very simple way, wasn't it?

182
00:09:30,640 --> 00:09:34,520
We had five features that we picked,
so there wasn't a lot of choice,

183
00:09:34,520 --> 00:09:37,200
so we wouldn't expect
a great likeness.

184
00:09:37,200 --> 00:09:39,560
I mean, this is the latest system
that the police

185
00:09:39,560 --> 00:09:41,840
have been using in this country
and abroad

186
00:09:41,840 --> 00:09:44,080
to evolve the likeness of the face.

187
00:09:44,080 --> 00:09:47,240
So where is our eyewitness
who did that piece of work for us?

188
00:09:47,240 --> 00:09:50,360
Thank you very much indeed
for all your hard work.

189
00:09:50,360 --> 00:09:51,920
Well, we've got good news.

190
00:09:51,920 --> 00:09:55,680
We've got good news because
the EvoFIT has done its job,

191
00:09:55,680 --> 00:09:59,200
and the police have received a call
from somebody claiming

192
00:09:59,200 --> 00:10:03,880
that they know who the individual
in the picture might be.

193
00:10:03,880 --> 00:10:06,640
So we have a suspect, Charlie,

194
00:10:06,640 --> 00:10:11,560
and maybe we're ready to start
thinking about a trial? Possibly?

195
00:10:11,560 --> 00:10:13,800
ANNOUNCEMENT: Call for Professor
Charlie Froud.

196
00:10:17,320 --> 00:10:21,240
Professor, you say to the members
of the jury

197
00:10:21,240 --> 00:10:27,920
that they individually might be able
to recognise 5,000 faces?

198
00:10:27,920 --> 00:10:29,760
This is what the research
would suggest.

199
00:10:29,760 --> 00:10:32,560
Some people will recognise less,
some people more, but on average

200
00:10:32,560 --> 00:10:34,520
for faces that we're familiar with,

201
00:10:34,520 --> 00:10:37,160
we can, we can recognise
about 5,000.

202
00:10:37,160 --> 00:10:39,640
Very grateful to you, Professor,
but what I'm interested in,

203
00:10:39,640 --> 00:10:43,160
and the jury will want to bear
in mind, is that in nearly half

204
00:10:43,160 --> 00:10:45,960
of the instances of an EvoFIT,

205
00:10:45,960 --> 00:10:50,920
it produced no identification result
of any kind. Isn't that right?

206
00:10:50,920 --> 00:10:53,280
Almost half. That's true.

207
00:10:53,280 --> 00:10:57,520
So about 60% of the time
somebody is able to say a name

208
00:10:57,520 --> 00:10:59,840
for an EvoFIT image.

209
00:10:59,840 --> 00:11:03,280
And so if we show that image
to lots of people, it increases

210
00:11:03,280 --> 00:11:06,240
the chance of that face
being recognised.

211
00:11:06,240 --> 00:11:09,000
Given that nearly half
won't recognise it,

212
00:11:09,000 --> 00:11:12,040
it also increases the chances of
people not recognising it.

213
00:11:12,040 --> 00:11:14,840
That's also true.
Thank you very much, Professor.

214
00:11:14,840 --> 00:11:17,920
ANNOUNCEMENT: Professor Froud,
you may step down.

215
00:11:17,920 --> 00:11:20,400
Charlie, thank you very much indeed.
Thank you. Thank you.

216
00:11:21,880 --> 00:11:24,600
APPLAUSE

217
00:11:26,960 --> 00:11:30,040
Let's carry on, because there's
another way in which we can

218
00:11:30,040 --> 00:11:33,760
identify somebody, and that's
the good old-fashioned line-up.

219
00:11:33,760 --> 00:11:36,400
So we're going to see whether you
as an audience can identify -

220
00:11:36,400 --> 00:11:38,920
we're giving you another go
at this.

221
00:11:38,920 --> 00:11:41,680
And to run this line-up
I'm helped by a psychologist

222
00:11:41,680 --> 00:11:43,720
from the University of Birmingham

223
00:11:43,720 --> 00:11:45,920
who studies facial recognition.

224
00:11:45,920 --> 00:11:48,120
So please welcome
Professor Heather Flowe.

225
00:11:48,120 --> 00:11:51,280
APPLAUSE

226
00:11:52,640 --> 00:11:55,080
Thank you.

227
00:11:55,080 --> 00:11:56,880
Thank you, Heather, for joining us.

228
00:11:56,880 --> 00:11:59,120
Can you talk us through,
what is a line-up?

229
00:11:59,120 --> 00:12:00,440
What does it mean? Love to.

230
00:12:00,440 --> 00:12:04,560
So a line-up procedure is one
in which the police have apprehended

231
00:12:04,560 --> 00:12:09,520
a suspect, and that suspect
may or may not be guilty.

232
00:12:09,520 --> 00:12:14,000
So the police create a bit of a test
for the witness, whereby they place

233
00:12:14,000 --> 00:12:17,520
the suspect in with a group
of individuals who look similar

234
00:12:17,520 --> 00:12:19,480
in appearance to the suspect,

235
00:12:19,480 --> 00:12:22,640
but who are actually not guilty
of the crime.

236
00:12:22,640 --> 00:12:25,160
OK. So let's welcome them in,
please.

237
00:12:26,080 --> 00:12:28,800
APPLAUSE

238
00:12:31,600 --> 00:12:33,240
All right, wonderful.

239
00:12:33,240 --> 00:12:35,480
I don't know about you,
but I think they're all guilty.

240
00:12:35,480 --> 00:12:37,560
But there we go. Look a bit shady.

241
00:12:37,560 --> 00:12:39,560
All right, so drawing from
your memories,

242
00:12:39,560 --> 00:12:41,040
we want you to help us out.

243
00:12:41,040 --> 00:12:45,440
Tell us which one of these
individuals is the culprit.

244
00:12:45,440 --> 00:12:49,880
So how many of you think
number one is the culprit?

245
00:12:51,600 --> 00:12:53,680
OK. What about number two?

246
00:12:55,320 --> 00:12:57,120
OK, one or two. OK.

247
00:12:57,120 --> 00:12:58,560
And what about number three?

248
00:12:59,960 --> 00:13:02,720
Well, again, just one.
Oh, one or two? OK.

249
00:13:02,720 --> 00:13:04,880
Number four, put your hand up.

250
00:13:04,880 --> 00:13:08,080
OK. And if you think it's five,
put your hand up.

251
00:13:08,080 --> 00:13:12,480
Oh, I think four's just got it,
would be my thinking on it.

252
00:13:12,480 --> 00:13:15,720
I agree. Number four, could you
step forward, please?

253
00:13:15,720 --> 00:13:20,040
Then I would say that our
eyewitnesses have suggested

254
00:13:20,040 --> 00:13:24,200
that you may be the person
who has committed a crime.

255
00:13:24,200 --> 00:13:27,640
And so we're going to call you back
at a later date into the courtroom.

256
00:13:27,640 --> 00:13:30,280
Thank you very much indeed
to all of our line-up.

257
00:13:30,280 --> 00:13:32,880
APPLAUSE

258
00:13:34,240 --> 00:13:37,440
So, Heather, that wasn't clear-cut,
was it, at all?

259
00:13:37,440 --> 00:13:40,080
Where we didn't feel comfortable,
there was a lot of consensus.

260
00:13:40,080 --> 00:13:41,440
What does that tell us?

261
00:13:41,440 --> 00:13:44,360
It was really interesting.
There was a difference of opinion.

262
00:13:44,360 --> 00:13:46,080
And one of the things that I did

263
00:13:46,080 --> 00:13:48,680
is I actually baited you into
picking someone.

264
00:13:48,680 --> 00:13:51,280
And Mr Findlay would find
a lot of fault with that,

265
00:13:51,280 --> 00:13:54,640
because I didn't tell you
that the perpetrator

266
00:13:54,640 --> 00:13:58,400
might actually not be present,
that it's OK not to identify anyone,

267
00:13:58,400 --> 00:14:01,640
because it's really important
not only to identify the guilty,

268
00:14:01,640 --> 00:14:05,240
but also to exonerate
innocent suspects.

269
00:14:05,240 --> 00:14:08,640
The other thing is that
all of the faces that you saw

270
00:14:08,640 --> 00:14:10,360
were facing forward.

271
00:14:10,360 --> 00:14:14,200
All you had the opportunity
to see during the line-up test

272
00:14:14,200 --> 00:14:18,200
was the frontal part of
the individual's faces, right?

273
00:14:18,200 --> 00:14:20,520
You also didn't see
those faces move.

274
00:14:20,520 --> 00:14:24,360
So you didn't get all of these rich
cues that might help you remember,

275
00:14:24,360 --> 00:14:26,560
such as other parts of the face.

276
00:14:26,560 --> 00:14:29,040
So our line-up,
we don't do that any more,

277
00:14:29,040 --> 00:14:32,000
although it looks great in
the movies as well, doesn't it?

278
00:14:32,000 --> 00:14:35,600
But you've been developing
new techniques that will help us

279
00:14:35,600 --> 00:14:40,480
perhaps get that line-up into
a better place scientifically.

280
00:14:40,480 --> 00:14:42,720
I know this is at a research stage.
It is.

281
00:14:42,720 --> 00:14:45,400
But would you like to tell us a bit
about it? I would love to tell you.

282
00:14:45,400 --> 00:14:47,960
So for about the last ten years,
my colleagues and I

283
00:14:47,960 --> 00:14:50,920
have been developing
a new line-up procedure.

284
00:14:50,920 --> 00:14:54,760
So the witness who is looking
at the faces in the interactive

285
00:14:54,760 --> 00:14:59,040
line-up procedure is able to
grab those faces with their mouse

286
00:14:59,040 --> 00:15:02,600
and rotate the faces
into any pose desired.

287
00:15:02,600 --> 00:15:06,880
And what we have found when
we test this procedure against all

288
00:15:06,880 --> 00:15:10,040
of the other line-up procedures
that are used around the world,

289
00:15:10,040 --> 00:15:12,000
and there are many different kinds,

290
00:15:12,000 --> 00:15:14,520
is that it allows witnesses

291
00:15:14,520 --> 00:15:19,080
to better be able to detect guilty
from innocent suspects.

292
00:15:19,080 --> 00:15:20,680
And that's really important.

293
00:15:20,680 --> 00:15:25,440
So we had two of our audience
participate with you.

294
00:15:25,440 --> 00:15:27,800
They've not been in here
when we did the line-up,

295
00:15:27,800 --> 00:15:30,040
so they don't know who we've chosen.
Excellent.

296
00:15:31,120 --> 00:15:33,440
APPLAUSE

297
00:15:36,520 --> 00:15:38,240
What's your name? Matthew.

298
00:15:38,240 --> 00:15:40,080
Matthew. Nice to meet you.
I'm Heather.

299
00:15:40,080 --> 00:15:42,000
Joe. Joe. Nice to meet you.

300
00:15:42,000 --> 00:15:45,160
OK, so you're about to take
a line-up test,

301
00:15:45,160 --> 00:15:47,480
and the perpetrator
that committed the heist

302
00:15:47,480 --> 00:15:51,000
might not actually be present
in this line-up.

303
00:15:51,000 --> 00:15:54,200
So what I want you to do is
to take a look at the individuals

304
00:15:54,200 --> 00:15:57,520
in the next line-up, and see
if you recognise any of them

305
00:15:57,520 --> 00:16:00,040
as the perpetrator.
So go ahead and click "Next".

306
00:16:01,680 --> 00:16:04,760
So what you can do is hold down
the mouse button,

307
00:16:04,760 --> 00:16:09,200
and then use your finger to rotate
the individuals in the line-up,

308
00:16:09,200 --> 00:16:11,520
so you can see them from different
vantage points.

309
00:16:11,520 --> 00:16:15,640
You can see how the individual
facial features are moving together.

310
00:16:15,640 --> 00:16:18,960
What we found in our experiments
is if you allow witnesses

311
00:16:18,960 --> 00:16:21,800
to have control over the faces
in the line-up,

312
00:16:21,800 --> 00:16:23,600
what they most often do

313
00:16:23,600 --> 00:16:26,680
is they rotate the faces
to match the vantage point

314
00:16:26,680 --> 00:16:29,160
from which they studied
the perpetrator.

315
00:16:29,160 --> 00:16:33,000
We control this in experiments
where people either get to see

316
00:16:33,000 --> 00:16:35,120
the front of the face
when the crime is committed,

317
00:16:35,120 --> 00:16:37,920
or the side of the face
when the crime is committed,

318
00:16:37,920 --> 00:16:40,960
and we find that people rotate
the images to match that.

319
00:16:40,960 --> 00:16:43,440
So are we ready to have a selection,
do you think?

320
00:16:43,440 --> 00:16:45,720
Yeah, let's do it.
Are you all ready?

321
00:16:45,720 --> 00:16:50,000
OK. So please enter in a number
or write "not present"

322
00:16:50,000 --> 00:16:52,200
to indicate your identification
decision.

323
00:16:55,800 --> 00:16:57,120
Fantastic.

324
00:16:57,120 --> 00:17:00,080
Which number did you choose?
Number two. Number two.

325
00:17:00,080 --> 00:17:01,680
And how confident are you?

326
00:17:01,680 --> 00:17:03,440
Around 60%. 60%. OK.

327
00:17:03,440 --> 00:17:05,400
What did you choose? Number four.

328
00:17:05,400 --> 00:17:09,280
Number four. Around 10%.
Around 10% confidence.

329
00:17:09,280 --> 00:17:11,960
Thank you very much. Let's give them
a round of applause.

330
00:17:11,960 --> 00:17:14,720
APPLAUSE

331
00:17:14,720 --> 00:17:18,520
Really interesting, Heather. Yeah.
Because we have a two. Yes.

332
00:17:18,520 --> 00:17:19,880
We have a four. Yes.

333
00:17:19,880 --> 00:17:24,120
In this case, we have witnesses
who had very low confidence.

334
00:17:24,120 --> 00:17:28,560
I would say 10% is most
definitely too low of confidence

335
00:17:28,560 --> 00:17:31,880
for that individual to be able
to make an identification.

336
00:17:31,880 --> 00:17:34,240
And the one that's even at 60%,

337
00:17:34,240 --> 00:17:37,360
I don't think would really stand up
in a court of law either.

338
00:17:37,360 --> 00:17:40,640
And do we know ultimately
which the room selected?

339
00:17:40,640 --> 00:17:43,960
So the room identified the suspect.

340
00:17:45,080 --> 00:17:49,120
So we have identified a suspect,
which is Ben.

341
00:17:49,120 --> 00:17:51,160
APPLAUSE

342
00:17:54,160 --> 00:17:56,040
ANNOUNCEMENT: Call Professor
Heather Flowe.

343
00:18:02,360 --> 00:18:07,760
Professor, when it comes to asking
someone to make an identification,

344
00:18:07,760 --> 00:18:11,720
what, if anything, do you do
to test their ability

345
00:18:11,720 --> 00:18:13,280
to identify someone?

346
00:18:13,280 --> 00:18:16,720
That's outside of the purview
of an expert witness to do that.

347
00:18:16,720 --> 00:18:20,320
So no, we do not test
the individual witnesses in a case.

348
00:18:20,320 --> 00:18:23,840
And of course, some people
may be no good at that.

349
00:18:23,840 --> 00:18:26,240
I mean, do you think I would be
able to identify someone?

350
00:18:26,240 --> 00:18:27,760
I really don't know.

351
00:18:27,760 --> 00:18:30,240
Well, there might be a problem,
you see,

352
00:18:30,240 --> 00:18:35,120
because I have a thing called
face blindness or prosopagnosia,

353
00:18:35,120 --> 00:18:38,320
that I don't recognise faces.
This is true.

354
00:18:38,320 --> 00:18:42,960
So do you, from your considerable
experience, accept that the ability

355
00:18:42,960 --> 00:18:47,640
of people to identify must vary,
as in fact, we've already seen,

356
00:18:47,640 --> 00:18:50,280
given that we have three people
in one way or another identified?

357
00:18:50,280 --> 00:18:53,200
Yes, and that's why we have to
choose the most robust

358
00:18:53,200 --> 00:18:57,200
identification procedure, so that
it will boost everyone's ability

359
00:18:57,200 --> 00:18:59,200
to make accurate identifications.

360
00:18:59,200 --> 00:19:01,440
Thank you very much. Thank you.

361
00:19:01,440 --> 00:19:03,640
ANNOUNCEMENT: Professor Flaherty,
you may step down.

362
00:19:03,640 --> 00:19:06,200
APPLAUSE

363
00:19:07,520 --> 00:19:12,480
Both Charlie and Heather's research
is revealing how to best help people

364
00:19:12,480 --> 00:19:15,520
successfully recall facial memories.

365
00:19:15,520 --> 00:19:18,960
But we all find it easier
to recognise people that we know,

366
00:19:18,960 --> 00:19:20,560
members of our family, for example,

367
00:19:20,560 --> 00:19:24,160
but difficult if the individuals
are strangers.

368
00:19:24,160 --> 00:19:27,680
But there are others who seem to be
sort of superhuman,

369
00:19:27,680 --> 00:19:32,240
and have superpowers, and they're
called super recognisers.

370
00:19:32,240 --> 00:19:35,640
So I want you to welcome a retired
police officer who established

371
00:19:35,640 --> 00:19:38,840
and ran the world's first
super recogniser unit

372
00:19:38,840 --> 00:19:40,360
at New Scotland Yard.

373
00:19:40,360 --> 00:19:42,560
Detective Chief Inspector
Mike Neville.

374
00:19:42,560 --> 00:19:45,480
APPLAUSE

375
00:19:50,320 --> 00:19:51,920
A super recogniser.

376
00:19:51,920 --> 00:19:55,520
Tell me what that means, and tell me
about the unit that you set up.

377
00:19:55,520 --> 00:19:57,720
Right, so a super recogniser
is a person

378
00:19:57,720 --> 00:19:59,760
with a fantastic natural memory.

379
00:19:59,760 --> 00:20:01,600
You're just born with
this wonderful memory.

380
00:20:01,600 --> 00:20:03,240
You can remember lots of faces.

381
00:20:03,240 --> 00:20:06,040
And when I worked at Scotland Yard,
I had responsibility for taking

382
00:20:06,040 --> 00:20:07,880
all the CCTV images in,

383
00:20:07,880 --> 00:20:10,160
and as I circulated them
and said to officers,

384
00:20:10,160 --> 00:20:11,640
"Who are these criminals?",

385
00:20:11,640 --> 00:20:13,880
there were certain officers
who made lots and lots

386
00:20:13,880 --> 00:20:15,200
and lots of identification,

387
00:20:15,200 --> 00:20:18,280
and we identified these people
who've got this superb skill.

388
00:20:18,280 --> 00:20:21,200
Can you give us one of your success
stories, if you wouldn't mind?

389
00:20:21,200 --> 00:20:23,200
Everybody likes a forensic story.

390
00:20:23,200 --> 00:20:25,880
Well, the greatest success,
I think, is the Novichok case,

391
00:20:25,880 --> 00:20:29,200
which you'll remember some Russian
assassins came to Salisbury,

392
00:20:29,200 --> 00:20:32,640
and they needed to find out who
they were, where they'd come from.

393
00:20:32,640 --> 00:20:36,320
So two super recognisers were
tasked to watch thousands of hours

394
00:20:36,320 --> 00:20:39,600
of footage of people entering
this country from Russia.

395
00:20:39,600 --> 00:20:42,480
Then they were tasked to watch
a week's footage of Salisbury,

396
00:20:42,480 --> 00:20:44,840
and they were asked,
is anybody in set A in set B?

397
00:20:44,840 --> 00:20:46,280
And that's how they found them.

398
00:20:46,280 --> 00:20:49,320
So it's good how the human mind
is better than a computer.

399
00:20:49,320 --> 00:20:53,000
And what percentage of the
population have got this skill?

400
00:20:53,000 --> 00:20:56,000
We think 1-2% have got some of
the skills.

401
00:20:56,000 --> 00:20:57,720
They could be here right now.

402
00:20:57,720 --> 00:21:00,520
So we're going to test
where our super recognisers

403
00:21:00,520 --> 00:21:02,520
might be in the audience
this evening.

404
00:21:02,520 --> 00:21:04,160
Thank you very, very much.
Thank you.

405
00:21:04,160 --> 00:21:07,200
APPLAUSE

406
00:21:09,240 --> 00:21:13,520
Now, we've now got Steve and Rob
here from WiFi Wars,

407
00:21:13,520 --> 00:21:16,000
and you're going to run us
through a game

408
00:21:16,000 --> 00:21:20,720
to see if we can spot any
super recognisers in our audience.

409
00:21:20,720 --> 00:21:22,760
That's exactly right.
So over to you. Thank you.

410
00:21:22,760 --> 00:21:26,200
All right. So, yes, we are now
going to test all of you

411
00:21:26,200 --> 00:21:28,880
and see how you get on
at recognising some suspects.

412
00:21:28,880 --> 00:21:30,880
So I need you all to have
your phones at the ready.

413
00:21:30,880 --> 00:21:32,080
You've already connected.

414
00:21:32,080 --> 00:21:35,200
Now I'm going to show you a face
from three different angles.

415
00:21:35,200 --> 00:21:39,480
So take a look at the big screen
and commit this to memory.

416
00:21:39,480 --> 00:21:40,640
There's the first person.

417
00:21:42,440 --> 00:21:43,880
From three angles.

418
00:21:43,880 --> 00:21:47,520
And now you have ten seconds to
tell me which of these three faces

419
00:21:47,520 --> 00:21:49,640
was that person, starting now?

420
00:21:49,640 --> 00:21:51,720
So A, B or C on your phone.

421
00:21:51,720 --> 00:21:54,360
Choose A, B or C, which one
you think it was.

422
00:21:54,360 --> 00:21:56,320
You've got a few more seconds
to do that.

423
00:21:56,320 --> 00:21:58,880
There's lots of nodding. You seem
confident, but let's find out.

424
00:21:58,880 --> 00:22:00,240
Time is up.

425
00:22:00,240 --> 00:22:02,440
I can now tell you -
you're all green at the moment,

426
00:22:02,440 --> 00:22:04,920
let's see how that goes.

427
00:22:04,920 --> 00:22:07,440
First thing to tell you is
the correct answer was B.

428
00:22:09,160 --> 00:22:12,000
OK, 77%. So we can see a few reds
around, but not too many.

429
00:22:12,000 --> 00:22:14,840
You can put your phones down for
now. Commiserations to the red ones.

430
00:22:14,840 --> 00:22:17,280
We're going to go again
with the second face, please, Rob.

431
00:22:17,280 --> 00:22:20,960
So a different face now. Three
angles again for those who remain.

432
00:22:20,960 --> 00:22:24,680
There's that face from all
three angles. Ten seconds.

433
00:22:24,680 --> 00:22:28,240
Was that person A, B or C?

434
00:22:28,240 --> 00:22:31,160
But for those who remain,
hopefully you made your choice

435
00:22:31,160 --> 00:22:32,880
because time is up.

436
00:22:32,880 --> 00:22:34,840
The correct - yeah, well done.

437
00:22:34,840 --> 00:22:38,120
The correct answer this time was A.

438
00:22:38,120 --> 00:22:42,440
And after that, 66% of you remain.
So only two-thirds of you now.

439
00:22:42,440 --> 00:22:45,680
People who are still in,
memorise these six faces.

440
00:22:47,320 --> 00:22:50,360
But now the line-up is going to be
more difficult to choose from.

441
00:22:50,360 --> 00:22:51,800
No conferring!

442
00:22:51,800 --> 00:22:54,360
Which of these three
was in that line-up of six?

443
00:22:55,640 --> 00:22:58,760
Horribly cloudy night - A, B or C?

444
00:22:58,760 --> 00:23:01,560
Only one of them. They all look
the same to me from down here.

445
00:23:01,560 --> 00:23:04,040
If you're still in,
you're better than I am.

446
00:23:04,040 --> 00:23:06,880
And I hope you've done it quick
because the time is up.

447
00:23:06,880 --> 00:23:08,440
And the final one was...

448
00:23:09,560 --> 00:23:11,720
B, ha-ha!

449
00:23:11,720 --> 00:23:13,400
How many are we left with?

450
00:23:13,400 --> 00:23:15,440
6%! Wonderful.

451
00:23:15,440 --> 00:23:18,600
Right, those 6% of you,
where are you, where are my 6%?

452
00:23:18,600 --> 00:23:19,880
Yeah, good, good, good.

453
00:23:19,880 --> 00:23:21,880
And what we'll do now
is we're going to go back

454
00:23:21,880 --> 00:23:24,920
through all of those rounds. Rob,
you're going to select the person

455
00:23:24,920 --> 00:23:27,440
who was fastest at getting it right
through all of them,

456
00:23:27,440 --> 00:23:29,800
That is your ultimate
super recogniser in the room.

457
00:23:29,800 --> 00:23:32,280
And a round of applause
for that person, who is...

458
00:23:33,360 --> 00:23:36,000
Artemis! So where are you, Artemis?
All the way at the back there.

459
00:23:36,000 --> 00:23:37,480
Well done to you.

460
00:23:37,480 --> 00:23:38,680
APPLAUSE

461
00:23:38,680 --> 00:23:40,720
Is this something
you're surprised by?

462
00:23:40,720 --> 00:23:43,200
Do you think you have
those superpowers?

463
00:23:43,200 --> 00:23:45,880
No. Well, maybe we've just taught
you something about yourself

464
00:23:45,880 --> 00:23:48,760
that you didn't know, maybe there's
a future career there for you,

465
00:23:48,760 --> 00:23:50,320
do you think? Maybe. Maybe?

466
00:23:50,320 --> 00:23:53,720
So when we pulled individual four
out in the line-up,

467
00:23:53,720 --> 00:23:55,840
was it four that you had chosen?

468
00:23:55,840 --> 00:24:00,040
Yeah. I wonder if the room feels
more confident in the selection

469
00:24:00,040 --> 00:24:04,480
of number four that our very own
Royal Institution super recogniser

470
00:24:04,480 --> 00:24:07,680
also pulled out number four?
Well done, congratulations.

471
00:24:12,880 --> 00:24:16,920
Now you will know that the police
use CCTV imagery like that sourced

472
00:24:16,920 --> 00:24:20,160
by the super recogniser team
all the time,

473
00:24:20,160 --> 00:24:25,600
and an average Londoner
is captured 70 - seven zero -

474
00:24:25,600 --> 00:24:28,560
times a day on a CCTV camera.

475
00:24:28,560 --> 00:24:33,840
And this dense network of cameras
means that we can use CCTV to track

476
00:24:33,840 --> 00:24:36,200
most movements of city dwellers.

477
00:24:36,200 --> 00:24:37,840
So have a look at this.

478
00:24:37,840 --> 00:24:40,200
This is a photograph, oh,
a train station!

479
00:24:40,200 --> 00:24:44,120
I wonder who that might be walking
through the train station?

480
00:24:44,120 --> 00:24:47,440
Oh, there you are again,
heading along the walkway.

481
00:24:47,440 --> 00:24:49,640
Gosh, I look really tired, don't I?

482
00:24:49,640 --> 00:24:51,840
That's what happens
when you do rehearsals

483
00:24:51,840 --> 00:24:54,480
for the Royal Institution.
Standing on the train platform.

484
00:24:54,480 --> 00:24:57,480
And I had no idea
this was being done.

485
00:24:57,480 --> 00:24:58,760
Absolutely none.

486
00:24:58,760 --> 00:25:02,400
All I did was to say what times
I was at different places

487
00:25:02,400 --> 00:25:05,720
within that journey
to the rehearsals.

488
00:25:05,720 --> 00:25:09,640
So that CCTV footage
is incredibly important.

489
00:25:09,640 --> 00:25:11,800
And to analyse that kind of footage,

490
00:25:11,800 --> 00:25:15,440
I want you to welcome a forensic
image analyst and an academic

491
00:25:15,440 --> 00:25:17,920
who's provided expert witness
testimony

492
00:25:17,920 --> 00:25:20,440
in courtrooms across the world.

493
00:25:20,440 --> 00:25:22,360
Please welcome Ray Evans.

494
00:25:22,360 --> 00:25:24,840
APPLAUSE

495
00:25:29,080 --> 00:25:34,400
Ray, as a CCTV image analyst,
will you talk us through

496
00:25:34,400 --> 00:25:39,360
the heist video and tell us
what you see through your eyes?

497
00:25:39,360 --> 00:25:42,560
Yes, indeed, so we can see
it's not great footage.

498
00:25:42,560 --> 00:25:47,120
There's a lot of noise in there,
the detail is lacking.

499
00:25:47,120 --> 00:25:49,280
I think it's important for me
to say at this point

500
00:25:49,280 --> 00:25:52,640
that unlike the experts
that you've already seen,

501
00:25:52,640 --> 00:25:55,200
they specialise in recognition.

502
00:25:55,200 --> 00:25:57,960
Well, what we deal with
is identification.

503
00:25:57,960 --> 00:26:01,680
All the work we do is geared
to looking at the differences

504
00:26:01,680 --> 00:26:04,480
between sets of images.

505
00:26:04,480 --> 00:26:07,320
You could almost say the task
is like a glorified

506
00:26:07,320 --> 00:26:08,800
spot the difference.

507
00:26:08,800 --> 00:26:12,600
Now we have a suspect chosen
by everybody in the room here.

508
00:26:12,600 --> 00:26:15,200
And it's a man who is identified
from the line-up

509
00:26:15,200 --> 00:26:16,800
that we saw earlier.

510
00:26:16,800 --> 00:26:19,640
And you have some images
of that individual

511
00:26:19,640 --> 00:26:21,800
and you're going to take us
through a comparison.

512
00:26:21,800 --> 00:26:24,040
Yes, I'll take it through
very quickly.

513
00:26:24,040 --> 00:26:27,600
So we see the person come in,
slow him down a little bit there.

514
00:26:27,600 --> 00:26:31,120
So this is the suspect that we took
out of the line, this is Ben.

515
00:26:31,120 --> 00:26:34,880
OK, can we see any differences
between those two people?

516
00:26:34,880 --> 00:26:38,480
The perpetrator
is sucking his lips in.

517
00:26:38,480 --> 00:26:40,480
So let's move forward
a little bit further.

518
00:26:40,480 --> 00:26:43,760
So you can have differences, but
as long as you have an explanation

519
00:26:43,760 --> 00:26:46,400
for that differences?
Precisely, precisely.

520
00:26:46,400 --> 00:26:48,160
So we'll move that through there.

521
00:26:48,160 --> 00:26:51,680
So on this, I've highlighted
the most glaring aspect,

522
00:26:51,680 --> 00:26:52,880
which is the ear.

523
00:26:52,880 --> 00:26:55,640
So we can have a look at that ear
and we can say, well,

524
00:26:55,640 --> 00:26:58,720
what are the differences
between these two sets of ears?

525
00:26:58,720 --> 00:27:04,280
Well, apart from the poor quality,
I don't see too many things.

526
00:27:04,280 --> 00:27:07,280
We can see the outline of the helix
there, which is the outside

527
00:27:07,280 --> 00:27:09,000
of the ear.

528
00:27:09,000 --> 00:27:10,520
It's quite rolled.

529
00:27:10,520 --> 00:27:13,040
We can see the lower parts
of the lobe.

530
00:27:13,040 --> 00:27:14,880
It's semi attached there.

531
00:27:14,880 --> 00:27:17,320
So they would go
into the similarity box.

532
00:27:17,320 --> 00:27:22,360
But remember, similarities does not
mean that it is the same person.

533
00:27:22,360 --> 00:27:24,560
Even though we've got CCTV footage,

534
00:27:24,560 --> 00:27:28,240
we've got a suspect within
the first hour or so,

535
00:27:28,240 --> 00:27:30,760
it's the quality of the image
that's the problem, isn't it?

536
00:27:30,760 --> 00:27:32,280
That's right, that's right.

537
00:27:32,280 --> 00:27:33,720
Thank you, Ray.

538
00:27:33,720 --> 00:27:35,440
Call Ray Evans.

539
00:27:38,000 --> 00:27:41,880
Now, then, sir, despite some
pretty heavy prompting

540
00:27:41,880 --> 00:27:46,120
from Professor Black,
you can't be in any way confident

541
00:27:46,120 --> 00:27:48,920
that you are looking at
the perpetrator, can you?

542
00:27:48,920 --> 00:27:51,000
Well, I have a level of confidence.

543
00:27:51,000 --> 00:27:56,160
Should the members of the jury
understand that in the work you do,

544
00:27:56,160 --> 00:28:01,600
you can either exclude or include?
That's correct, yes.

545
00:28:01,600 --> 00:28:05,520
And to exclude somebody is perhaps
easier because there may be

546
00:28:05,520 --> 00:28:08,560
quite a distinctive facial
difference, that sort of thing?

547
00:28:08,560 --> 00:28:13,080
To exclude somebody, it is easier
because you've got something

548
00:28:13,080 --> 00:28:16,240
that separates those two faces.

549
00:28:17,160 --> 00:28:21,360
To include somebody, that is harder?

550
00:28:21,360 --> 00:28:25,200
Yes, it is harder, but people
can share similar features.

551
00:28:25,200 --> 00:28:29,240
And if those features are similar
enough, it might be difficult

552
00:28:29,240 --> 00:28:32,320
to distinguish between one person
and another.

553
00:28:32,320 --> 00:28:35,120
So what the members of the jury
have here

554
00:28:35,120 --> 00:28:38,960
is that there are some differences?

555
00:28:38,960 --> 00:28:40,160
Correct.

556
00:28:40,160 --> 00:28:43,960
There are also some areas
where there appears

557
00:28:43,960 --> 00:28:46,280
to be a clear difference?

558
00:28:47,640 --> 00:28:50,120
Relatively clear, I wouldn't say
they're absolutely clear.

559
00:28:50,120 --> 00:28:52,720
And in other areas,
there are some similarities?

560
00:28:52,720 --> 00:28:54,360
Yes, yes, that's correct.

561
00:28:54,360 --> 00:28:57,360
Just in a term or a phrase...

562
00:28:57,360 --> 00:29:00,560
..describe your level of confidence
in this case.

563
00:29:00,560 --> 00:29:04,840
I would probably use the term
limited support.

564
00:29:04,840 --> 00:29:07,440
Limited... Maybe as much as
moderate support,

565
00:29:07,440 --> 00:29:09,480
but certainly limited support.

566
00:29:09,480 --> 00:29:11,560
Thank you very much.
Thank you.

567
00:29:11,560 --> 00:29:14,280
Ray Evans, you may step down.

568
00:29:14,280 --> 00:29:17,680
Thank you very much, Ray,
very much appreciated, thank you.

569
00:29:17,680 --> 00:29:20,160
APPLAUSE

570
00:29:22,400 --> 00:29:24,040
This is the bit you've been
waiting for -

571
00:29:24,040 --> 00:29:26,480
I need a volunteer, please.

572
00:29:26,480 --> 00:29:28,360
Would you come forward, please?

573
00:29:28,360 --> 00:29:30,080
There we go, thank you.

574
00:29:35,600 --> 00:29:36,840
And your name?

575
00:29:36,840 --> 00:29:38,280
Carly. Carly.

576
00:29:38,280 --> 00:29:41,600
So Carly, I'm going to ask you
to head off here with Isla.

577
00:29:41,600 --> 00:29:45,160
We're going to pop
a little something on you.

578
00:29:45,160 --> 00:29:47,960
Nothing horrible, I do promise you.

579
00:29:47,960 --> 00:29:51,680
And what we're going to do
is for identification purposes,

580
00:29:51,680 --> 00:29:54,320
we're going to look at gait.

581
00:29:54,320 --> 00:29:57,360
Now, I don't mean that thing
at the end of a fence that you open

582
00:29:57,360 --> 00:30:01,640
and close, because when we saw
our suspect running away,

583
00:30:01,640 --> 00:30:03,240
that is their gait.

584
00:30:03,240 --> 00:30:07,320
The gait is the way that people
carry their body or move their legs

585
00:30:07,320 --> 00:30:09,720
and their arms as they walk.

586
00:30:09,720 --> 00:30:13,680
Now it's up to the court to decide
what kind of evidence

587
00:30:13,680 --> 00:30:15,640
is admissible into the courtroom,

588
00:30:15,640 --> 00:30:18,880
and not all evidence
gets into the courtroom.

589
00:30:18,880 --> 00:30:21,960
And by and large,
judges are generally

590
00:30:21,960 --> 00:30:24,320
not scientifically trained.

591
00:30:24,320 --> 00:30:27,040
So we do have to make sure
that the science

592
00:30:27,040 --> 00:30:29,120
that goes into them is robust.

593
00:30:29,120 --> 00:30:32,920
And it's important that the expert
witnesses can inform the courtroom

594
00:30:32,920 --> 00:30:35,760
whether an evidence type
is strong or weak.

595
00:30:35,760 --> 00:30:37,600
Just as we heard from Ray Evans,

596
00:30:37,600 --> 00:30:40,280
that strength of conviction
is important.

597
00:30:40,280 --> 00:30:43,720
Now, you come and join us
over here.

598
00:30:43,720 --> 00:30:46,360
Just step further back a little bit
there and we're going to...

599
00:30:46,360 --> 00:30:48,640
That's perfect, all I'm going
to ask you to do...

600
00:30:48,640 --> 00:30:50,600
Hand by your side,
if you wouldn't mind?

601
00:30:50,600 --> 00:30:53,200
You can see that we've got
some lights on your head

602
00:30:53,200 --> 00:30:54,600
and down the side here.

603
00:30:54,600 --> 00:30:56,880
We're going to turn the lights off
in here.

604
00:30:56,880 --> 00:31:00,640
And the camera, if you watch,
is going to track how you walk.

605
00:31:00,640 --> 00:31:05,000
So I'm going to ask you to walk
just over to that cushion there,

606
00:31:05,000 --> 00:31:06,760
very slowly.

607
00:31:06,760 --> 00:31:08,840
Are you ready?
Lights are going down.

608
00:31:08,840 --> 00:31:11,880
Off you go, and when you get there,
I want you to stop.

609
00:31:11,880 --> 00:31:14,880
So you walk across, there you go,
you can see the leg movement.

610
00:31:14,880 --> 00:31:17,440
You can see the arm movement.

611
00:31:17,440 --> 00:31:19,280
So that's gait.

612
00:31:19,280 --> 00:31:22,200
Now, if you stop
and if you walk backwards,

613
00:31:22,200 --> 00:31:24,080
I'll stop you, don't worry.

614
00:31:24,080 --> 00:31:25,560
Look what happens with backward.

615
00:31:25,560 --> 00:31:29,480
The leg moves, bit of arm movement
as well.

616
00:31:29,480 --> 00:31:31,760
Can I ask you to do
something silly now?

617
00:31:31,760 --> 00:31:36,480
Would you, on your left leg, stand
and hop over to the other side?

618
00:31:36,480 --> 00:31:38,960
Oh, look, it's a completely
different pattern, isn't it?

619
00:31:38,960 --> 00:31:41,760
Carly, thank you very much indeed,
well done, well done.

620
00:31:41,760 --> 00:31:43,240
Thank you.

621
00:31:46,320 --> 00:31:50,320
Now, as you might expect, there's
been a tremendous amount of research

622
00:31:50,320 --> 00:31:52,480
has gone into
how our skeletons move.

623
00:31:52,480 --> 00:31:56,000
And it's really important in terms
of high-performance sport,

624
00:31:56,000 --> 00:31:58,960
but also in medicine,
so for orthopaedic intervention.

625
00:31:58,960 --> 00:32:02,280
So we know a lot about gait,
but when you don't know the person,

626
00:32:02,280 --> 00:32:06,400
identifying gait of that individual
is really quite difficult.

627
00:32:06,400 --> 00:32:09,800
And sometimes there are some judges
who are quite uncomfortable

628
00:32:09,800 --> 00:32:13,080
about allowing gait analysis
into the courtroom,

629
00:32:13,080 --> 00:32:16,440
and all that tells us is
that there's more research to do.

630
00:32:16,440 --> 00:32:20,840
Mr Findlay, could I ask you
whether gait analysis

631
00:32:20,840 --> 00:32:23,520
is something that has appeared
in any of the cases

632
00:32:23,520 --> 00:32:26,120
where you've been involved
and whether it's an evidence type

633
00:32:26,120 --> 00:32:27,680
you'd be comfortable with?

634
00:32:27,680 --> 00:32:31,280
I've never had to deal with
gait analysis in any trial

635
00:32:31,280 --> 00:32:32,960
I've ever been involved in.

636
00:32:32,960 --> 00:32:36,840
We have expert evidence
in a large number of areas,

637
00:32:36,840 --> 00:32:39,240
but it is important
to have two things.

638
00:32:39,240 --> 00:32:43,560
Firstly, there must be
an identifiable area of science.

639
00:32:43,560 --> 00:32:46,440
You know, you can't just take
a hobby and call yourself an expert

640
00:32:46,440 --> 00:32:48,200
and then you become
an expert witness.

641
00:32:48,200 --> 00:32:51,640
So we need to understand
what the science is.

642
00:32:51,640 --> 00:32:55,520
And then the person who is
the expert must have devoted

643
00:32:55,520 --> 00:32:57,760
a good deal of time to study,
to research,

644
00:32:57,760 --> 00:32:59,760
to the practical application.

645
00:32:59,760 --> 00:33:02,600
If there isn't a lot of information
supporting gait

646
00:33:02,600 --> 00:33:04,960
going into the courtroom,
there's a tremendous amount

647
00:33:04,960 --> 00:33:07,480
of evidence in terms of DNA.

648
00:33:07,480 --> 00:33:11,920
And we've been able to retrieve some
DNA samples from the glass cabinet.

649
00:33:11,920 --> 00:33:14,600
You know that when you're looking
at the television programmes,

650
00:33:14,600 --> 00:33:17,400
the minute they say
the DNA evidence, everyone goes,

651
00:33:17,400 --> 00:33:20,520
"The case is over", because surely
if there's DNA evidence,

652
00:33:20,520 --> 00:33:24,240
it must be absolutely spot on,
we've got the right individual.

653
00:33:24,240 --> 00:33:27,560
But even DNA evidence
isn't free from doubt.

654
00:33:27,560 --> 00:33:30,640
So to present the DNA evidence
and how it's used in the court,

655
00:33:30,640 --> 00:33:33,400
please welcome the Professor
of Genetics at the University

656
00:33:33,400 --> 00:33:35,800
of Leicester, Professor Turi King.

657
00:33:35,800 --> 00:33:38,280
APPLAUSE

658
00:33:39,360 --> 00:33:41,400
Lovely to see you.

659
00:33:43,040 --> 00:33:46,720
So, Turi, we're talking about DNA
and DNA always being the thing

660
00:33:46,720 --> 00:33:48,920
that we think solves everything.
That's right.

661
00:33:48,920 --> 00:33:52,960
But we want to talk about transfer?
We do. And persistence maybe,

662
00:33:52,960 --> 00:33:56,080
but transfer certainly? Yeah.
So what do we mean by transfer?

663
00:33:56,080 --> 00:33:59,120
Can we show what we mean
by DNA transfer?

664
00:33:59,120 --> 00:34:03,720
Yes, and for that,
we need six volunteers.

665
00:34:03,720 --> 00:34:06,600
You do two...

666
00:34:06,600 --> 00:34:09,200
You come down.

667
00:34:09,200 --> 00:34:12,440
You come down,
and you in the pink jumper.

668
00:34:12,440 --> 00:34:14,520
Come down. And over this side...

669
00:34:14,520 --> 00:34:17,760
Let's take you and...you.

670
00:34:17,760 --> 00:34:19,440
Come on down.

671
00:34:20,840 --> 00:34:22,960
Come and stand here.

672
00:34:22,960 --> 00:34:25,640
Now, I'm going to send you two out
where you're going to get

673
00:34:25,640 --> 00:34:28,640
looked after by our team.

674
00:34:28,640 --> 00:34:32,400
Right, now then, the rest of you,
what I need you to do

675
00:34:32,400 --> 00:34:35,240
is I need you to wear some, sadly,
especially from your outfit,

676
00:34:35,240 --> 00:34:37,960
non-Christmassy jumpers,
I'm afraid

677
00:34:37,960 --> 00:34:40,080
So you need to put those on.

678
00:34:40,080 --> 00:34:43,520
Can I welcome back
our two lovely volunteers

679
00:34:43,520 --> 00:34:45,480
who are going to come in?

680
00:34:51,680 --> 00:34:53,840
You look amazing!

681
00:34:54,880 --> 00:34:57,720
And you are
covered in DNA balls.

682
00:34:57,720 --> 00:35:00,360
Now, what I want you to do
is I want the six of you

683
00:35:00,360 --> 00:35:04,640
to form a tight circle here,
and I don't want the two of you

684
00:35:04,640 --> 00:35:07,320
to stand next to each other,
so if you can form a circle,

685
00:35:07,320 --> 00:35:10,280
form a circle facing in
and, oop, that's right.

686
00:35:10,280 --> 00:35:12,560
Perfect, now I'm going to
go over here.

687
00:35:12,560 --> 00:35:15,560
So what I want you to do is, I want
you to take your left hand,

688
00:35:15,560 --> 00:35:18,360
put it into the middle, and grab
the hand of somebody...

689
00:35:18,360 --> 00:35:20,240
Not right next to you.

690
00:35:20,240 --> 00:35:23,040
OK, now you're going to take your
right hand and you're going

691
00:35:23,040 --> 00:35:25,840
to do the same thing, try not to
grab the hand of somebody

692
00:35:25,840 --> 00:35:28,560
who you've already got the hand of,
if that makes sense?

693
00:35:28,560 --> 00:35:31,280
Great, OK, now,
what I want you to do

694
00:35:31,280 --> 00:35:34,880
is I want you
to untangle yourselves.

695
00:35:34,880 --> 00:35:39,480
You can pivot, you can step
over arms if you want.

696
00:35:39,480 --> 00:35:44,080
So, three, two, one... go!

697
00:35:45,920 --> 00:35:49,960
Excellent, keep holding hands,
but try and untangle.

698
00:35:51,280 --> 00:35:53,160
Oh, it's tricky, isn't it?

699
00:35:54,680 --> 00:35:56,280
Oh, one in the middle.

700
00:35:58,320 --> 00:36:01,360
Now, because you can turn around
and go backwards...

701
00:36:01,360 --> 00:36:04,000
OK, do we, I don't know,
do we give up?

702
00:36:04,000 --> 00:36:06,320
I think maybe what we do
is we get you in a line,

703
00:36:06,320 --> 00:36:08,080
so let's get you in a line.

704
00:36:08,080 --> 00:36:13,680
So when we started, we had the two
of you with the lovely hats on,

705
00:36:13,680 --> 00:36:18,000
had got DNA bobbles on you, but now,
do you want to turn around?

706
00:36:19,440 --> 00:36:21,680
Oh, my goodness,
check that out.

707
00:36:21,680 --> 00:36:27,800
OK, so what this is showing, though,
is that there is DNA transfer.

708
00:36:27,800 --> 00:36:33,640
So DNA has transferred from
these two onto everybody else.

709
00:36:33,640 --> 00:36:37,080
So you've got blue going here,
that's primary transfer.

710
00:36:37,080 --> 00:36:41,280
But then if I got this from this
person, that's secondary transfer.

711
00:36:41,280 --> 00:36:43,400
Now that is happening
all of the time.

712
00:36:43,400 --> 00:36:45,880
All you need to do is shake
somebody's hand

713
00:36:45,880 --> 00:36:47,960
and then go and touch a door handle,

714
00:36:47,960 --> 00:36:51,400
and you've put somebody else's
DNA onto the door handle.

715
00:36:51,400 --> 00:36:53,160
Thank you very much, everybody.

716
00:36:53,160 --> 00:36:56,120
That was fantastic.
If you guys want to go out.

717
00:36:56,120 --> 00:36:58,200
APPLAUSE

718
00:37:02,320 --> 00:37:04,120
Call Professor Turi King.

719
00:37:06,640 --> 00:37:10,280
Tell me, Professor,
if you will, please.

720
00:37:10,280 --> 00:37:15,320
DNA comes from various
body fluids,

721
00:37:15,320 --> 00:37:17,200
skin cells and so on.

722
00:37:17,200 --> 00:37:18,800
Yes.

723
00:37:18,800 --> 00:37:24,240
And when it is analysed, you can
connect that to an individual

724
00:37:24,240 --> 00:37:27,520
to odds which are pretty
well astronomical?

725
00:37:27,520 --> 00:37:31,280
Yes. You never say absolutely,
but as near absolutely as we can?

726
00:37:31,280 --> 00:37:32,520
Yes.

727
00:37:32,520 --> 00:37:37,560
What can the presence of DNA
on an object tell you, firstly,

728
00:37:37,560 --> 00:37:41,120
about how that DNA
came to be on that object?

729
00:37:41,120 --> 00:37:43,960
We cannot answer that question.

730
00:37:43,960 --> 00:37:47,320
Or how long has it been
on that object?

731
00:37:47,320 --> 00:37:49,600
We can't answer that
question either.

732
00:37:49,600 --> 00:37:53,520
If, for example, we've seen
a video of a case,

733
00:37:53,520 --> 00:37:58,160
if I was to take my watch off
and place it in the case

734
00:37:58,160 --> 00:38:00,720
and come back in two years' time,

735
00:38:00,720 --> 00:38:03,600
would it still have my DNA
on the watch?

736
00:38:03,600 --> 00:38:06,480
The DNA will have degraded,
but it's possible.

737
00:38:06,480 --> 00:38:10,200
So we don't know when, we don't know
how, we don't know for how long,

738
00:38:10,200 --> 00:38:13,680
and we know that it may be
by direct contact,

739
00:38:13,680 --> 00:38:16,120
and we've seen that
in your demonstration,

740
00:38:16,120 --> 00:38:18,560
but also by secondary transfer?

741
00:38:18,560 --> 00:38:21,120
Yes. Or tertiary transfer?

742
00:38:21,120 --> 00:38:23,600
Yes. And I can't remember the one
for four, but it could be

743
00:38:23,600 --> 00:38:25,080
four down the road? It is.

744
00:38:25,080 --> 00:38:27,880
I mean, the amount of DNA
that will be transferred

745
00:38:27,880 --> 00:38:29,880
will decrease each time.

746
00:38:29,880 --> 00:38:33,320
And it depends on a number
of different factors.

747
00:38:33,320 --> 00:38:36,000
How do you measure the amount
of DNA?

748
00:38:36,000 --> 00:38:40,200
A standard unit of measurement
for DNA

749
00:38:40,200 --> 00:38:41,640
is known as a nanogram.

750
00:38:41,640 --> 00:38:42,960
What's a nanogram?

751
00:38:42,960 --> 00:38:44,680
What's in a nanogram?

752
00:38:44,680 --> 00:38:47,200
It's a billionth of a gram.

753
00:38:47,200 --> 00:38:49,680
So a tiny amount?

754
00:38:49,680 --> 00:38:54,480
Even the tiniest amount of DNA
could involve somebody

755
00:38:54,480 --> 00:38:56,320
potentially in a crime.

756
00:38:56,320 --> 00:38:57,600
Yes.

757
00:38:57,600 --> 00:39:01,520
The tiniest amount of DNA
could involve this young man...

758
00:39:02,640 --> 00:39:05,800
..the light of innocence
shining from his eyes.

759
00:39:05,800 --> 00:39:07,720
LAUGHTER

760
00:39:11,240 --> 00:39:16,600
Might it be important to know
why somebody's DNA

761
00:39:16,600 --> 00:39:18,840
might be in a particular place,
Professor?

762
00:39:18,840 --> 00:39:22,600
Indeed. We simply present
the evidence... Oh.

763
00:39:22,600 --> 00:39:25,480
..but we cannot say how or when.

764
00:39:25,480 --> 00:39:28,480
I suppose if somebody worked
in a place like this,

765
00:39:28,480 --> 00:39:30,640
that might be very important.

766
00:39:30,640 --> 00:39:31,920
Definitely.

767
00:39:31,920 --> 00:39:33,320
Tell them where you work.

768
00:39:33,320 --> 00:39:35,960
Royal Institution
of Great Britain. Here!

769
00:39:35,960 --> 00:39:38,080
Your DNA's all over the place.

770
00:39:38,080 --> 00:39:40,520
Good lad. Thank you, Professor.

771
00:39:40,520 --> 00:39:43,160
APPLAUSE

772
00:39:43,160 --> 00:39:45,600
Professor King, you may step down.

773
00:39:48,880 --> 00:39:52,120
An extra piece of evidence
has come in.

774
00:39:52,120 --> 00:39:54,720
What we hadn't realised
was that there was a camera

775
00:39:54,720 --> 00:39:59,440
in the cabinet above
where the ring was stolen.

776
00:39:59,440 --> 00:40:02,240
And you can see the camera
in place there.

777
00:40:02,240 --> 00:40:05,960
You can see the hand come in
and steal the ring.

778
00:40:05,960 --> 00:40:09,560
So we've seen identification
from faces -

779
00:40:09,560 --> 00:40:12,800
can we tell anything from hands?

780
00:40:12,800 --> 00:40:14,880
Now we're quite used to
the front of our hands

781
00:40:14,880 --> 00:40:16,880
with our fingerprints
being identifiable.

782
00:40:16,880 --> 00:40:19,240
But this is the back of a hand.

783
00:40:19,240 --> 00:40:24,000
So we have a research team based
both at the University of Lancaster

784
00:40:24,000 --> 00:40:27,880
and at the University of Dundee,
who are looking at identification

785
00:40:27,880 --> 00:40:29,800
from the back of the hand.

786
00:40:29,800 --> 00:40:33,080
So could you please welcome
Ricky Boswell-Challand, please?

787
00:40:33,080 --> 00:40:34,760
APPLAUSE

788
00:40:38,640 --> 00:40:43,080
So we had an image up here
that showed in the robbery

789
00:40:43,080 --> 00:40:44,680
just the back of a hand.

790
00:40:44,680 --> 00:40:48,760
What is it that you can see and tell
from the back of a hand?

791
00:40:48,760 --> 00:40:51,720
OK, so the back of the hand's
really interesting because there

792
00:40:51,720 --> 00:40:56,120
are a number of features that show
variation between individuals.

793
00:40:56,120 --> 00:40:59,200
So I'd like to invite you all
to take a look at the back

794
00:40:59,200 --> 00:41:01,000
of your own hand.

795
00:41:01,000 --> 00:41:03,200
There's always that phrase,
isn't it, you know it like

796
00:41:03,200 --> 00:41:04,400
the back of your own hand?

797
00:41:04,400 --> 00:41:06,080
Yeah, I forget mine all the time.

798
00:41:06,080 --> 00:41:09,080
So if you take a look at the back
of your hand, you'll see,

799
00:41:09,080 --> 00:41:11,560
and I'll look at mine, you'll see
a number of features.

800
00:41:11,560 --> 00:41:13,680
We've all got creases
in our knuckles.

801
00:41:13,680 --> 00:41:18,040
We've got various skin
pigmentation, freckles...

802
00:41:18,040 --> 00:41:20,480
Birthmarks? Birthmarks, scars.

803
00:41:20,480 --> 00:41:23,640
So we've got a number of different
features like that.

804
00:41:23,640 --> 00:41:26,840
We see, we find more
if we go under the skin.

805
00:41:26,840 --> 00:41:29,800
As we go under the skin,
we've got a model to show here.

806
00:41:29,800 --> 00:41:31,600
Well, that's a big hand.

807
00:41:34,520 --> 00:41:38,240
OK, so here we've got the internal
structure of the hand.

808
00:41:38,240 --> 00:41:39,760
Obviously, we've got the skeleton,

809
00:41:39,760 --> 00:41:41,960
you can see the bones
of the fingers.

810
00:41:41,960 --> 00:41:44,640
We've got tendons, we've got
ligaments, we've got the musculature

811
00:41:44,640 --> 00:41:46,760
of the hand here, shown in red.

812
00:41:46,760 --> 00:41:49,840
They give the overall form
and definition of the hand.

813
00:41:49,840 --> 00:41:52,680
And then we've got something
really important.

814
00:41:52,680 --> 00:41:57,080
We've got the veins,
and veins show a great deal

815
00:41:57,080 --> 00:42:00,040
of variability between individuals.

816
00:42:00,040 --> 00:42:02,120
So if you were to look at the back
of your right hand

817
00:42:02,120 --> 00:42:04,480
and your left hand, would you
expect them to be different?

818
00:42:04,480 --> 00:42:07,120
I would expect... In the same
individual? Very, very different.

819
00:42:07,120 --> 00:42:09,560
We'd expect to see different
patterns between

820
00:42:09,560 --> 00:42:11,400
identical twins, even.

821
00:42:11,400 --> 00:42:15,520
So is there anything about this vein
pattern that's genetic?

822
00:42:15,520 --> 00:42:19,160
Because you've said that identical
twins, it would be different.

823
00:42:19,160 --> 00:42:22,240
But if you were looking at a family,
you know, mum, dad, brothers,

824
00:42:22,240 --> 00:42:25,600
sisters, is there anything in there?
We'd still expect differences.

825
00:42:25,600 --> 00:42:27,240
To the best of our knowledge,

826
00:42:27,240 --> 00:42:30,760
we do not think there is a genetic
influence on the vein patterns.

827
00:42:30,760 --> 00:42:33,840
And there's a database
associated with the images

828
00:42:33,840 --> 00:42:37,200
that I know you're looking at.
Have you been able to yet find

829
00:42:37,200 --> 00:42:41,760
the same pattern of superficial
veins in any two individuals?

830
00:42:41,760 --> 00:42:43,560
Never in two hands
that we have studied.

831
00:42:43,560 --> 00:42:45,160
And what's your database size?

832
00:42:45,160 --> 00:42:48,840
We're currently nearing
5,000 individuals,

833
00:42:48,840 --> 00:42:51,320
two hands for most
of those individuals.

834
00:42:51,320 --> 00:42:54,680
So within that 10,000,
we've seen no duplication.

835
00:42:54,680 --> 00:42:56,280
Brilliant, thank you, Ricky.

836
00:42:56,280 --> 00:42:59,520
And I have to say it, but you have
to give him a big hand.

837
00:42:59,520 --> 00:43:01,880
GROANING AND APPLAUSE

838
00:43:04,160 --> 00:43:08,200
One of the reasons that DNA
has been so successful

839
00:43:08,200 --> 00:43:11,720
in identification is because of
the very large databases

840
00:43:11,720 --> 00:43:13,440
that we can interrogate.

841
00:43:13,440 --> 00:43:17,560
And to be able to interrogate
large data sets of images,

842
00:43:17,560 --> 00:43:21,360
we can't do it individually,
we need to involve a computer.

843
00:43:21,360 --> 00:43:26,360
So we start to use artificial
intelligence to look at every image

844
00:43:26,360 --> 00:43:30,160
and say, can we extract
the anatomical information

845
00:43:30,160 --> 00:43:32,160
out of those images?

846
00:43:32,160 --> 00:43:36,000
Now, several teams have already
presented this work

847
00:43:36,000 --> 00:43:39,880
in the courtroom, so it has been
deemed admissible evidence.

848
00:43:39,880 --> 00:43:41,880
So that's, I think,
important to know.

849
00:43:41,880 --> 00:43:43,480
So what I'd like to do now is

850
00:43:43,480 --> 00:43:48,760
I'd like to welcome in a computer
expert, and he will talk us through

851
00:43:48,760 --> 00:43:51,840
what is happening in the next
stage of research

852
00:43:51,840 --> 00:43:54,040
in this hand identification.

853
00:43:54,040 --> 00:43:57,480
So please welcome, from Lancaster
University, Dr Brian Williams.

854
00:43:57,480 --> 00:43:59,360
APPLAUSE

855
00:44:02,120 --> 00:44:05,320
Thank you, Brian.
Thank you, Brian.

856
00:44:05,320 --> 00:44:08,960
So we've just been looking,
we've got an image of the hand

857
00:44:08,960 --> 00:44:11,560
of the person involved in the crime.

858
00:44:11,560 --> 00:44:14,440
And of course, we have a hand
of our suspect.

859
00:44:14,440 --> 00:44:17,400
What would you do in terms
of a comparison?

860
00:44:17,400 --> 00:44:19,760
What sort of information
are we looking for

861
00:44:19,760 --> 00:44:22,480
and how are we using computers
to help with that?

862
00:44:22,480 --> 00:44:26,440
So in terms of the hand images,
what we're looking for is reliable

863
00:44:26,440 --> 00:44:30,000
anatomical information that we can
find in both the suspect

864
00:44:30,000 --> 00:44:31,840
and the offender images.

865
00:44:31,840 --> 00:44:35,320
So what we do is we take
thousands of examples of images

866
00:44:35,320 --> 00:44:38,480
that have already been marked up
by experts, such as Sue,

867
00:44:38,480 --> 00:44:41,560
in order to show the vein patterns,
the knuckle creases,

868
00:44:41,560 --> 00:44:43,920
some detail in the fingernails.

869
00:44:43,920 --> 00:44:47,360
And then we feed these into computer
algorithms that we develop

870
00:44:47,360 --> 00:44:50,440
so that later on, when we come
with case images like this,

871
00:44:50,440 --> 00:44:53,960
we can use the trained models
to extract this data again.

872
00:44:53,960 --> 00:44:57,880
So you have images of hands
and you've trained the computer,

873
00:44:57,880 --> 00:45:01,280
I just need to get this right,
you've trained a computer to say,

874
00:45:01,280 --> 00:45:03,840
"Find the veins and show me
the pattern,"

875
00:45:03,840 --> 00:45:07,040
and then you're able
to compare the two patterns

876
00:45:07,040 --> 00:45:09,000
of suspect and offender?

877
00:45:09,000 --> 00:45:13,920
But how confident might you be that
they could be the same individual?

878
00:45:13,920 --> 00:45:16,040
Well, that depends on
the level of information

879
00:45:16,040 --> 00:45:17,520
that we've got available.

880
00:45:17,520 --> 00:45:20,160
So one of the important things
when we're developing a system

881
00:45:20,160 --> 00:45:23,680
for something like this is it should
be a system that can evaluate itself

882
00:45:23,680 --> 00:45:26,800
so it can evaluate how good
is the quality in this image

883
00:45:26,800 --> 00:45:29,680
and how important,
how relevant is the information

884
00:45:29,680 --> 00:45:31,560
that we can extract?

885
00:45:31,560 --> 00:45:35,600
So in this case, we had excellent
quality in the finger knuckles

886
00:45:35,600 --> 00:45:37,840
and in the dorsal side of the hand.

887
00:45:37,840 --> 00:45:40,280
So are you quite content
that you just might have

888
00:45:40,280 --> 00:45:42,000
the same individual?

889
00:45:42,000 --> 00:45:46,560
I am confident in the extraction of
the data, and in the evidence image,

890
00:45:46,560 --> 00:45:49,840
we have very clear images
of the knuckle creases.

891
00:45:49,840 --> 00:45:51,640
So I am pretty confident.

892
00:45:51,640 --> 00:45:54,560
And there we can see that again,
so here is the image.

893
00:45:54,560 --> 00:45:58,760
So what we see here is we see
some very clear images, especially

894
00:45:58,760 --> 00:46:02,320
of the index finger, the major
knuckle on the index finger

895
00:46:02,320 --> 00:46:05,520
and the major knuckle of
the middle finger.

896
00:46:05,520 --> 00:46:08,680
And we can see both of these clearly
in this hand as well.

897
00:46:08,680 --> 00:46:12,560
We can also see the dorsal side
of the hand where we can extract

898
00:46:12,560 --> 00:46:15,000
visible vein patterns from that.

899
00:46:15,000 --> 00:46:16,400
Thank you, Brian.

900
00:46:16,400 --> 00:46:18,120
Call Dr Brian Williams.

901
00:46:21,000 --> 00:46:24,680
I didn't quite understand this,
no doubt members of the jury do,

902
00:46:24,680 --> 00:46:27,520
but who carried out this exercise?

903
00:46:27,520 --> 00:46:31,640
You, or a box
of electronic gadgetry?

904
00:46:31,640 --> 00:46:33,120
It was both of us.

905
00:46:33,120 --> 00:46:35,320
Well, given the box of
electronic gadgetry

906
00:46:35,320 --> 00:46:37,560
couldn't put
the information into itself,

907
00:46:37,560 --> 00:46:40,360
who put the information
into the box in the first place?

908
00:46:40,360 --> 00:46:43,560
The evidence images
were entered by me.

909
00:46:43,560 --> 00:46:48,040
And the background software
that you use is entered by?

910
00:46:48,040 --> 00:46:51,920
The background software is developed
by my team of computer scientists

911
00:46:51,920 --> 00:46:54,200
and mathematicians
and anthropologists.

912
00:46:54,200 --> 00:46:56,040
And to be fair, you are,
of course, a doctor.

913
00:46:56,040 --> 00:46:57,720
I am, yes. Of medicine?

914
00:46:57,720 --> 00:47:01,000
No... Not of medicine?
I'm a doctor of mathematics.

915
00:47:01,000 --> 00:47:03,440
Ah. Who is actually
carrying this out -

916
00:47:03,440 --> 00:47:05,200
you or the machine?

917
00:47:05,200 --> 00:47:07,280
It's a combination.

918
00:47:07,280 --> 00:47:11,280
But it's a combination of a machine
that can only run on the basis

919
00:47:11,280 --> 00:47:15,720
of the information that's put
in by somebody who is not

920
00:47:15,720 --> 00:47:17,640
a doctor of medicine.

921
00:47:17,640 --> 00:47:20,800
I'm afraid, Doctor, I'm going
to have to suggest that

922
00:47:20,800 --> 00:47:23,480
you are not an expert in this area.

923
00:47:23,480 --> 00:47:26,760
And for your information,
I will be inviting this jury

924
00:47:26,760 --> 00:47:29,360
to reject your evidence totally.

925
00:47:29,360 --> 00:47:32,080
Dr Williams, you may step down.

926
00:47:32,080 --> 00:47:33,840
Thank you, Brian.

927
00:47:33,840 --> 00:47:35,800
APPLAUSE

928
00:47:38,560 --> 00:47:40,600
Calling Professor Sue Black.

929
00:47:40,600 --> 00:47:42,840
Oh, my, OK.

930
00:47:42,840 --> 00:47:44,680
Tell me...

931
00:47:44,680 --> 00:47:49,640
So in this case you are looking
to make, I assume, and the jury

932
00:47:49,640 --> 00:47:53,440
would hope, a fair comparison
between two parts

933
00:47:53,440 --> 00:47:55,120
of somebody's body.

934
00:47:55,120 --> 00:48:00,080
Yes. Well, how many areas did you
compare between the suspect's hand

935
00:48:00,080 --> 00:48:04,600
and the other hand that was taken
from the video of the robbery?

936
00:48:04,600 --> 00:48:09,600
So I compared initially
knuckle crease patterns.

937
00:48:09,600 --> 00:48:15,640
And I also looked for vein patterns,
but I was not able to identify veins

938
00:48:15,640 --> 00:48:18,080
because veins are not always visible
in all individuals.

939
00:48:18,080 --> 00:48:23,840
So there was one basis, knuckle
crease patterns across all fingers.

940
00:48:23,840 --> 00:48:29,520
One characteristic?
Across all fingers.

941
00:48:29,520 --> 00:48:31,920
I think we have that, don't you?

942
00:48:31,920 --> 00:48:36,480
And of course, there has to be
a reasonable representation.

943
00:48:36,480 --> 00:48:40,200
The two hands, they were clearly
in the same position?

944
00:48:40,200 --> 00:48:42,880
No, they were not.

945
00:48:42,880 --> 00:48:45,360
They were quite distinctive
in their positioning?

946
00:48:45,360 --> 00:48:48,240
They were. So that would have to be
taken into account, wouldn't it?

947
00:48:48,240 --> 00:48:51,200
Absolutely. So we have
the poor quality of imaging. Yep.

948
00:48:51,200 --> 00:48:55,040
We have the fact that you did
a very, very limited comparison...

949
00:48:55,040 --> 00:48:57,720
I did what was possible
on the comparison.

950
00:48:57,720 --> 00:49:00,080
..only on creases in the fingers...
Across all fingers.

951
00:49:00,080 --> 00:49:02,200
..and on hands that were not
in the same position.

952
00:49:02,200 --> 00:49:03,400
Have I got all of that?

953
00:49:03,400 --> 00:49:05,320
Some of it, yes. Yes.

954
00:49:05,320 --> 00:49:07,120
LAUGHTER

955
00:49:07,120 --> 00:49:08,920
It's been a pleasure, as always.

956
00:49:08,920 --> 00:49:10,920
As always, thank you.
Thank you very much.

957
00:49:10,920 --> 00:49:12,520
Professor Black, you may step down.

958
00:49:14,360 --> 00:49:17,200
I had no idea where that was going.

959
00:49:17,200 --> 00:49:19,520
And that's really important.

960
00:49:19,520 --> 00:49:23,440
The courtroom is an alien place
for a scientist.

961
00:49:23,440 --> 00:49:26,160
We don't know what somebody
is going to ask us.

962
00:49:26,160 --> 00:49:29,680
We may have the burning thing
that we want to say,

963
00:49:29,680 --> 00:49:33,280
but if the advocates don't ask us
the question we want,

964
00:49:33,280 --> 00:49:35,280
we're not able to say it.

965
00:49:35,280 --> 00:49:39,680
The courtroom is really where
science and evidence can be tested.

966
00:49:39,680 --> 00:49:41,760
Now, what the Crown will say to you,

967
00:49:41,760 --> 00:49:45,360
I'm putting the Crown's words
into my mouth, as it were.

968
00:49:45,360 --> 00:49:49,720
The Crown will tell you that
a crime was committed this evening,

969
00:49:49,720 --> 00:49:53,560
that the crime was seen
on CCTV footage.

970
00:49:53,560 --> 00:49:58,240
That some eyewitnesses
identified a suspect.

971
00:49:58,240 --> 00:50:02,840
That that suspect has since,
their hand has been compared

972
00:50:02,840 --> 00:50:06,920
and the gait perhaps has been
looked at, the DNA analysis

973
00:50:06,920 --> 00:50:10,400
has been done, and we've
looked at similarities

974
00:50:10,400 --> 00:50:12,560
and differences across the faces.

975
00:50:12,560 --> 00:50:16,840
The Crown's case would be that
there is sufficient evidence

976
00:50:16,840 --> 00:50:21,160
to find this man guilty of stealing
the diamond ring

977
00:50:21,160 --> 00:50:25,880
from the Royal Institution, a crime
that you witnessed this evening.

978
00:50:25,880 --> 00:50:29,520
But that would be the premise
of the Crown's case.

979
00:50:30,560 --> 00:50:33,760
Mr Findlay, the defence case.

980
00:50:35,200 --> 00:50:40,600
It has been my privilege
to represent this young man.

981
00:50:42,320 --> 00:50:45,840
Decent lad, decent home.

982
00:50:45,840 --> 00:50:50,360
Loving son, spends his time
in this building,

983
00:50:50,360 --> 00:50:55,360
earning a modest living to look
after his old, grey-haired mother.

984
00:50:58,600 --> 00:51:05,280
And yet the prosecution say he is
guilty of the crime of robbery.

985
00:51:05,280 --> 00:51:07,200
On what basis?

986
00:51:07,200 --> 00:51:08,720
The evidence.

987
00:51:08,720 --> 00:51:11,080
And that's what you have
to consider.

988
00:51:11,080 --> 00:51:15,640
We start off with a little exercise

989
00:51:15,640 --> 00:51:17,480
where we do an EVOfit

990
00:51:17,480 --> 00:51:20,440
and we discover that, yes,

991
00:51:20,440 --> 00:51:23,040
there is a close vote between

992
00:51:23,040 --> 00:51:26,760
two persons who may be the suspect.

993
00:51:26,760 --> 00:51:29,520
But from the very start,
there is a close vote

994
00:51:29,520 --> 00:51:30,880
between two.

995
00:51:30,880 --> 00:51:32,600
How many others?

996
00:51:32,600 --> 00:51:35,040
How many others could we bring
into this court

997
00:51:35,040 --> 00:51:37,840
who would fit into one or other
categories?

998
00:51:37,840 --> 00:51:41,240
Then we have new technology,
new science that is being developed.

999
00:51:41,240 --> 00:51:43,080
And what does that show?

1000
00:51:43,080 --> 00:51:47,360
That produces a third potential
suspect for this robbery.

1001
00:51:47,360 --> 00:51:48,640
So we now have three.

1002
00:51:48,640 --> 00:51:50,960
How many more do the Crown want?

1003
00:51:50,960 --> 00:51:55,000
And how many more do the Crown
want you to ignore?

1004
00:51:55,000 --> 00:51:56,680
This young man,

1005
00:51:56,680 --> 00:52:00,600
to feed his family,
works in this very building,

1006
00:52:00,600 --> 00:52:02,280
and in this very building

1007
00:52:02,280 --> 00:52:05,840
his DNA will be in every single
corner of the building.

1008
00:52:05,840 --> 00:52:09,360
The fact that it is connected
with the robbery is irrelevant.

1009
00:52:09,360 --> 00:52:11,160
It is neither here nor there.

1010
00:52:11,160 --> 00:52:14,120
And you now know, you now know,

1011
00:52:14,120 --> 00:52:16,440
when you get home tonight,

1012
00:52:16,440 --> 00:52:20,240
you will be contaminated
with the DNA

1013
00:52:20,240 --> 00:52:24,720
of people you didn't even know
existed before you came here.

1014
00:52:24,720 --> 00:52:27,880
And Ray Evans,
who in a balanced and fair way

1015
00:52:27,880 --> 00:52:30,640
despite being pushed and prodded
by the prosecution,

1016
00:52:30,640 --> 00:52:32,120
time and time again,

1017
00:52:32,120 --> 00:52:35,040
relentlessly, remorselessly
driving him in their direction,

1018
00:52:35,040 --> 00:52:37,920
would not budge, would not budge.

1019
00:52:37,920 --> 00:52:41,320
And said really he couldn't say
very much

1020
00:52:41,320 --> 00:52:43,520
other than it was
a limited support.

1021
00:52:43,520 --> 00:52:45,200
Where is this Crown case?

1022
00:52:45,200 --> 00:52:47,520
There is no Crown case

1023
00:52:47,520 --> 00:52:50,200
until of course
the prosecution advocate

1024
00:52:50,200 --> 00:52:51,720
joins in herself

1025
00:52:51,720 --> 00:52:53,440
to try and provide evidence

1026
00:52:53,440 --> 00:52:55,160
and talks about a hand

1027
00:52:55,160 --> 00:52:57,880
where there are creases
on one hand.

1028
00:52:58,920 --> 00:53:01,160
And that is it.

1029
00:53:01,160 --> 00:53:04,120
A hand that's not in the same
position as a comparison hand,

1030
00:53:04,120 --> 00:53:06,040
an image which is so poor

1031
00:53:06,040 --> 00:53:09,240
that you have seen the consequences
in this court

1032
00:53:09,240 --> 00:53:11,440
of trying to identify somebody

1033
00:53:11,440 --> 00:53:13,720
from such poor quality produce.

1034
00:53:15,600 --> 00:53:17,960
Members of the jury,

1035
00:53:17,960 --> 00:53:20,120
if there was to be a crime
committed,

1036
00:53:20,120 --> 00:53:22,480
it would be here tonight

1037
00:53:22,480 --> 00:53:25,200
in convicting this young man

1038
00:53:25,200 --> 00:53:27,640
of the charge of robbery.

1039
00:53:27,640 --> 00:53:30,480
Without hesitation, I ask you...

1040
00:53:31,520 --> 00:53:33,480
..I ask you...

1041
00:53:33,480 --> 00:53:35,600
to reject the Crown case,

1042
00:53:35,600 --> 00:53:37,760
and at this time of year,

1043
00:53:37,760 --> 00:53:41,480
to allow him to walk out...

1044
00:53:43,000 --> 00:53:44,400
..into Christmas...

1045
00:53:44,400 --> 00:53:45,880
GIGGLING

1046
00:53:45,880 --> 00:53:48,600
..to return home
to his mother...

1047
00:53:48,600 --> 00:53:50,080
LAUGHTER

1048
00:53:51,320 --> 00:53:54,280
..and share a modest,

1049
00:53:54,280 --> 00:53:57,120
very small chicken sandwich

1050
00:53:57,120 --> 00:53:59,680
for Christmas dinner,

1051
00:53:59,680 --> 00:54:02,600
because it's the best he can do.

1052
00:54:02,600 --> 00:54:05,040
I invite you...

1053
00:54:05,040 --> 00:54:06,480
..to acquit.

1054
00:54:07,720 --> 00:54:10,240
Home and dry, boy, home and dry.
Thank you very much.

1055
00:54:16,240 --> 00:54:20,280
Never knowingly overacted,
Mr Findlay. Never.

1056
00:54:20,280 --> 00:54:22,880
OK, the moment has arrived

1057
00:54:22,880 --> 00:54:25,040
where you now have to make
a decision.

1058
00:54:25,040 --> 00:54:27,160
And in all seriousness,

1059
00:54:27,160 --> 00:54:29,200
I know we've had a bit of fun
with this,

1060
00:54:29,200 --> 00:54:32,520
but in all seriousness,
you've had some evidence.

1061
00:54:32,520 --> 00:54:34,280
You have to make a decision.

1062
00:54:34,280 --> 00:54:36,520
I don't think there is
a more difficult decision

1063
00:54:36,520 --> 00:54:38,360
than anybody could make

1064
00:54:38,360 --> 00:54:40,080
when they are a member of the jury.

1065
00:54:40,080 --> 00:54:44,040
Is there sufficient evidence
to find Ben guilty?

1066
00:54:44,040 --> 00:54:46,720
Mr Findlay does not need to
prove innocence.

1067
00:54:46,720 --> 00:54:50,280
He just needs reasonable doubt.

1068
00:54:50,280 --> 00:54:52,800
I'm going to ask you, first of all,

1069
00:54:52,800 --> 00:54:56,120
whether you find Ben guilty of...

1070
00:54:56,120 --> 00:54:57,720
GIGGLING

1071
00:54:57,720 --> 00:55:00,520
..of the theft of a diamond ring

1072
00:55:00,520 --> 00:55:02,840
from the Royal Institution.

1073
00:55:02,840 --> 00:55:06,600
Is Ben guilty of that crime?

1074
00:55:06,600 --> 00:55:08,120
Please raise your hand.

1075
00:55:10,800 --> 00:55:13,040
Please raise your hand
if you believe

1076
00:55:13,040 --> 00:55:15,320
that there is reasonable doubt.

1077
00:55:18,040 --> 00:55:20,440
Isn't that interesting?

1078
00:55:20,440 --> 00:55:22,360
Do you know,
there's one of the problems

1079
00:55:22,360 --> 00:55:24,120
we have in a courtroom,

1080
00:55:24,120 --> 00:55:27,720
which is if you have an even number
of jurors,

1081
00:55:27,720 --> 00:55:30,840
then you can have a split decision.

1082
00:55:30,840 --> 00:55:35,120
How many jurors do we have
in a Scottish court, Mr Findlay?

1083
00:55:35,120 --> 00:55:37,920
15. 15.
Why do we do that in Scotland?

1084
00:55:37,920 --> 00:55:41,720
Because we can have a balance
of probability

1085
00:55:41,720 --> 00:55:46,400
that says one side is going to have
a higher number than the other.

1086
00:55:46,400 --> 00:55:47,960
You're not in Scotland,

1087
00:55:47,960 --> 00:55:51,200
but if you were,
there's a third verdict

1088
00:55:51,200 --> 00:55:54,480
and that third verdict
is not proven.

1089
00:55:54,480 --> 00:55:56,720
So we might be in London,

1090
00:55:56,720 --> 00:55:59,840
we might be
in the Royal Institution,

1091
00:55:59,840 --> 00:56:02,320
but I would suggest that perhaps

1092
00:56:02,320 --> 00:56:05,720
Scotland's not proven verdict

1093
00:56:05,720 --> 00:56:08,520
is what you've come forward with
this evening,

1094
00:56:08,520 --> 00:56:11,520
because based on the evidence
that you've heard,

1095
00:56:11,520 --> 00:56:15,640
I'm not happy
that he could be convicted,

1096
00:56:15,640 --> 00:56:18,320
given... I'm not finished yet.

1097
00:56:18,320 --> 00:56:19,840
LAUGHTER

1098
00:56:19,840 --> 00:56:23,360
..given the level of doubt
that is in the room.

1099
00:56:23,360 --> 00:56:27,480
And I think we'd have to be
much more confident than that.

1100
00:56:27,480 --> 00:56:29,760
I'm going to let him go

1101
00:56:29,760 --> 00:56:33,560
because I think that would be
the right thing to do.

1102
00:56:33,560 --> 00:56:35,360
Did you do it? No.

1103
00:56:35,360 --> 00:56:37,480
No, of course you didn't.

1104
00:56:37,480 --> 00:56:39,200
Of course he didn't do it.

1105
00:56:39,200 --> 00:56:43,160
Ladies and gentlemen, you have made
an incredibly difficult decision.

1106
00:56:43,160 --> 00:56:47,440
You have taken what is some science
that you know you see on television

1107
00:56:47,440 --> 00:56:50,800
and television says to you
it's never wrong.

1108
00:56:50,800 --> 00:56:53,640
Well, sometimes we have to
ask questions,

1109
00:56:53,640 --> 00:56:55,760
the really difficult questions.

1110
00:56:55,760 --> 00:56:58,040
And some of my best research

1111
00:56:58,040 --> 00:57:01,600
has come out of
the really difficult questions

1112
00:57:01,600 --> 00:57:06,000
that a defence advocate will ask,

1113
00:57:06,000 --> 00:57:08,280
because if we don't know the answer
to it,

1114
00:57:08,280 --> 00:57:10,520
then we need to go find out.

1115
00:57:10,520 --> 00:57:13,720
So science never stands still,
it moves forward.

1116
00:57:13,720 --> 00:57:17,640
It is at the core of
the most important element

1117
00:57:17,640 --> 00:57:19,800
of our civilised society.

1118
00:57:19,800 --> 00:57:21,640
It is about justice,

1119
00:57:21,640 --> 00:57:24,200
it is about finding truth.

1120
00:57:24,200 --> 00:57:27,320
And you have done
an exceptional job on that.

1121
00:57:27,320 --> 00:57:31,080
Mr Findlay, I thank you very much
indeed for this evening.

1122
00:57:31,080 --> 00:57:35,480
And may we never, ever end up
in court again.

1123
00:57:35,480 --> 00:57:37,200
Thank you very much indeed.

1124
00:57:37,200 --> 00:57:38,320
Thank you.

1125
00:57:38,320 --> 00:57:39,720
APPLAUSE

1126
00:57:54,440 --> 00:57:58,240
It's...it's a Royal Institution
Christmas lecture.

1127
00:57:58,240 --> 00:58:01,280
And of course, right at the core
of everything we had

1128
00:58:01,280 --> 00:58:04,480
was a diamond ring.

1129
00:58:04,480 --> 00:58:08,920
Diamonds, of course,
being a girl's best friend.

1130
00:58:08,920 --> 00:58:11,240
And we've retrieved
the diamond ring.

1131
00:58:11,240 --> 00:58:13,600
Don't ask us how, but we have.

1132
00:58:13,600 --> 00:58:17,600
But it would not be a Royal
Institution Christmas lecture

1133
00:58:17,600 --> 00:58:19,960
if it didn't go out with a bang.

1134
00:58:19,960 --> 00:58:21,040
Three.

1135
00:58:21,040 --> 00:58:22,200
ALL: Two.

1136
00:58:22,200 --> 00:58:23,240
One.

1137
00:58:23,240 --> 00:58:25,240
AUDIENCE: Ooh!