1
00:00:17,700 --> 00:00:19,220
Live Text Access.

2
00:00:19,650 --> 00:00:22,620
Training
for real time intralingual subtitlers.

3
00:00:24,830 --> 00:00:27,140
Unit 5. Respeaking.

4
00:00:27,770 --> 00:00:30,970
Element 1. Psycho-cognitive skills.

5
00:00:33,540 --> 00:00:37,430
Remember full sentences
while lagging behind.

6
00:00:37,930 --> 00:00:41,560
Created by SSML and Velotype.

7
00:00:44,900 --> 00:00:47,740
On completion of this training sequence,

8
00:00:47,840 --> 00:00:52,490
you will be able to repeat
or reformulate the source text

9
00:00:52,560 --> 00:00:53,840
while listening to it,

10
00:00:53,810 --> 00:00:57,070
remember full sentences
while lagging behind,

11
00:00:57,530 --> 00:01:00,700
activate exit strategies
while respeaking

12
00:01:00,770 --> 00:01:03,310
if you realize
the speaker is challenging.

13
00:01:06,350 --> 00:01:09,050
This is the agenda of this presentation.

14
00:01:09,810 --> 00:01:12,780
I will first recall
the skills of the respeaker.

15
00:01:13,570 --> 00:01:16,280
Then,
I will concentrate on the memory system,

16
00:01:16,610 --> 00:01:20,830
which is at the basis of respeaking
and simultaneous interpreting,

17
00:01:21,230 --> 00:01:24,960
a mental activity
very close to real time subtitling.

18
00:01:25,880 --> 00:01:30,570
I will finally focus on the inferencing
and memorizing strategies,

19
00:01:30,960 --> 00:01:34,430
that you will be requested to practice
along the course

20
00:01:34,820 --> 00:01:38,520
to develop, reinforce
and finally maintain this skill

21
00:01:38,590 --> 00:01:44,060
in two different contexts:
verbatim and sensatim subtitling.

22
00:01:46,840 --> 00:01:49,350
Section 1 – The challenging speaker.

23
00:01:51,390 --> 00:01:52,510
As you have seen,

24
00:01:52,580 --> 00:01:56,280
a respeaker has to do
many things at the same time.

25
00:01:56,970 --> 00:02:02,080
In Element 1, LO1, we have dealt
with Psycho-cognitive skills,

26
00:02:02,450 --> 00:02:05,910
meaning all those skills related
to being able

27
00:02:05,980 --> 00:02:08,160
to listen and speak at the same time,

28
00:02:08,650 --> 00:02:12,740
which is the first set of skills
a respeaker has to possess.

29
00:02:13,340 --> 00:02:14,340
In particular,

30
00:02:14,490 --> 00:02:18,120
for a respeaker to be able
and become a professional in the field,

31
00:02:18,650 --> 00:02:22,480
he or she has also to train
a good short-term memory

32
00:02:22,840 --> 00:02:26,870
to be able and process
as many concepts as possible,

33
00:02:27,260 --> 00:02:30,500
and to apply
exit strategies or solutions

34
00:02:30,560 --> 00:02:34,680
to any problem one may encounter
in the subtitling process.

35
00:02:35,810 --> 00:02:38,710
While you can only develop
these competences

36
00:02:38,780 --> 00:02:42,340
by the time
and thanks to ad hoc exercises,

37
00:02:42,770 --> 00:02:47,260
you may be interested in knowing
how memory and short-term memory work.

38
00:02:48,020 --> 00:02:50,860
Before, it is important to recall.

39
00:02:51,350 --> 00:02:55,080
That is what we are going to see
in the next section.

40
00:02:57,660 --> 00:02:59,670
In this Element, Element 2,

41
00:03:00,130 --> 00:03:03,040
we will deal with memory
and the memory system.

42
00:03:04,750 --> 00:03:06,800
Section 2 – Memory.

43
00:03:11,060 --> 00:03:12,060
As we know,

44
00:03:12,260 --> 00:03:14,870
in any multitasking process
which involves

45
00:03:14,940 --> 00:03:17,010
doing two things at the same time,

46
00:03:17,640 --> 00:03:20,880
two main types of strategies
are involved in respeaking:

47
00:03:21,570 --> 00:03:22,860
The first is understanding,

48
00:03:22,920 --> 00:03:25,400
aimed at comprehending the source text.

49
00:03:25,830 --> 00:03:27,050
The second is respeaking,

50
00:03:27,380 --> 00:03:30,840
aimed
at making the most out of understanding

51
00:03:30,910 --> 00:03:32,430
under stressful conditions

52
00:03:32,490 --> 00:03:36,020
and with a reduced processing capacity
than usual.

53
00:03:37,540 --> 00:03:39,290
In this, memory is crucial.

54
00:03:40,180 --> 00:03:43,580
But what is memory
and how does its system work?

55
00:03:44,840 --> 00:03:47,340
Memory is commonly described as a system

56
00:03:47,410 --> 00:03:49,980
for simultaneous storage and processing.

57
00:03:50,540 --> 00:03:54,500
Several studies have attempted
to investigate this aspect

58
00:03:54,640 --> 00:03:57,410
belonging to interpreter’s
and respeaker’s ability.

59
00:03:59,220 --> 00:04:02,920
While several different models of memory
have been proposed,

60
00:04:03,410 --> 00:04:06,450
the stage model of memory is often used

61
00:04:06,550 --> 00:04:10,050
to explain the basic structure
and function of memory.

62
00:04:11,090 --> 00:04:12,090
Initially,

63
00:04:12,160 --> 00:04:18,160
proposed in 1968 by Richard Atkinson
and Richard Shiffrin,

64
00:04:18,730 --> 00:04:22,980
this theory outlines
three separate stages of memory:

65
00:04:23,450 --> 00:04:27,770
sensory memory,
short-term memory and long-term memory.

66
00:04:28,890 --> 00:04:32,290
Sensory memory is the earliest stage
of memory.

67
00:04:32,850 --> 00:04:34,140
During this stage,

68
00:04:34,340 --> 00:04:37,800
sensory information
from the environment is stored

69
00:04:37,900 --> 00:04:40,140
for a very brief period of time,

70
00:04:40,610 --> 00:04:45,160
generally for no longer
than a half-second for visual information

71
00:04:45,520 --> 00:04:48,920
and 3 or 4 seconds
for auditory information.

72
00:04:49,710 --> 00:04:53,510
We attend to only certain aspects
of this sensory memory,

73
00:04:54,170 --> 00:04:58,330
allowing some of this information
to pass into the next stage:

74
00:04:58,720 --> 00:04:59,940
short-term memory.

75
00:05:01,560 --> 00:05:07,010
Short-term memory stores information
for approximately 20-30 seconds.

76
00:05:07,600 --> 00:05:09,980
It is also known as Working Memory,

77
00:05:10,340 --> 00:05:13,210
which refers
to the processes that are used

78
00:05:13,280 --> 00:05:17,600
to temporarily store,
organize and manipulate information.

79
00:05:18,320 --> 00:05:23,240
In the next slide, we will focus
on STM as a crucial system

80
00:05:23,670 --> 00:05:26,810
for the simultaneous interpreter
and for the respeaker.

81
00:05:28,030 --> 00:05:32,420
Long-term memory refers
to the continuing storage of information.

82
00:05:33,040 --> 00:05:37,130
This information is largely outside
of our awareness

83
00:05:37,500 --> 00:05:41,660
but can be called into working memory
to be used when needed.

84
00:05:44,920 --> 00:05:48,920
Working memory is key
in interpreting and respeaking,

85
00:05:49,280 --> 00:05:53,440
due to the operation of several factors
including:

86
00:05:53,600 --> 00:05:57,660
the time interval between
the moment the original speech is heard

87
00:05:58,120 --> 00:06:01,650
and the moment its processing
for comprehension is finished;

88
00:06:02,640 --> 00:06:04,290
the time interval between

89
00:06:04,360 --> 00:06:08,780
the moment the message to be respoken
or interpreted is determined

90
00:06:09,080 --> 00:06:11,260
and the completion of its formulation;

91
00:06:11,850 --> 00:06:16,040
tactical moves which are used,
for instance,

92
00:06:16,340 --> 00:06:20,430
if a speech segment is unclear
to the interpreter or respeaker

93
00:06:20,730 --> 00:06:23,990
because of bad sound quality,
a strong accent,

94
00:06:24,260 --> 00:06:28,280
unclear logic
or errors in the source text itself.

95
00:06:30,400 --> 00:06:31,420
We will now see

96
00:06:31,490 --> 00:06:34,090
how the working memory system
is structured.

97
00:06:34,160 --> 00:06:37,560
I will refer to the model put forward

98
00:06:37,660 --> 00:06:40,920
by Baddeley & Hitch in 1974.

99
00:06:42,140 --> 00:06:43,230
Initially,

100
00:06:43,330 --> 00:06:46,070
Baddeley and Hitch conceptualized
a system

101
00:06:46,140 --> 00:06:49,540
involving three main components of WM.

102
00:06:50,490 --> 00:06:55,110
This model included an attentional system
known as the central executive,

103
00:06:55,610 --> 00:07:00,030
which is responsible for controlling
and coordinating the flow of information

104
00:07:00,130 --> 00:07:02,510
from the two subordinate systems.

105
00:07:02,900 --> 00:07:05,150
These systems are
the visuo-spatial scratch pad

106
00:07:05,380 --> 00:07:08,120
which processes
visual and spatial information,

107
00:07:08,580 --> 00:07:12,930
and the phonological loop
which manages verbal information.

108
00:07:13,820 --> 00:07:18,410
In turn, the phonological loop consists
of a phonological store

109
00:07:18,740 --> 00:07:22,400
where information is held
for about 2 seconds

110
00:07:22,730 --> 00:07:25,080
before it begins to decay.

111
00:07:25,970 --> 00:07:29,330
The second component is
a rehearsal process

112
00:07:29,400 --> 00:07:35,180
that refreshes information
in the store via sub-vocal articulation.

113
00:07:36,590 --> 00:07:41,910
The central executive is considered
as the most important aspect of WM.

114
00:07:42,530 --> 00:07:48,240
It is a regulatory mechanism
that controls information flow in WM,

115
00:07:48,640 --> 00:07:52,430
information retrieval
from other memory systems

116
00:07:52,600 --> 00:07:56,460
and information storage
and processing in WM.

117
00:07:57,290 --> 00:08:02,570
The model allows for both storage
and manipulation of information.

118
00:08:04,280 --> 00:08:08,440
This memory model was revised
by Baddeley in 2000

119
00:08:08,870 --> 00:08:11,050
to include an ‘episodic buffer’,

120
00:08:11,380 --> 00:08:14,780
a multidimensional storage
and processing component

121
00:08:15,140 --> 00:08:19,530
assumed to provide
a temporary store of limited capacity

122
00:08:19,860 --> 00:08:22,860
that is capable
of integrating information

123
00:08:23,160 --> 00:08:28,340
from the subsidiary systems
with that of LTM.

124
00:08:29,530 --> 00:08:34,870
It is assumed to be important
for the chunking of information in STM.

125
00:08:35,570 --> 00:08:39,620
Buddley proposes
that the buffer is a mechanism

126
00:08:39,720 --> 00:08:44,540
for recalling the gist of a message
rather than verbatim information.

127
00:08:45,630 --> 00:08:48,900
Now, we have said
that memory is a system

128
00:08:48,960 --> 00:08:52,660
for simultaneous storage
and processing of the discourse.

129
00:08:53,390 --> 00:08:58,070
To help the respeaker or interpreter
in his or her understanding

130
00:08:58,400 --> 00:09:00,840
and accurately convey the message,

131
00:09:01,270 --> 00:09:06,120
the WM is crucial in carrying out
all the processes

132
00:09:06,450 --> 00:09:09,090
during the respeaking
or interpreting task,

133
00:09:09,650 --> 00:09:14,930
as the respeaker or interpreter retains
what he or she has just heard,

134
00:09:15,230 --> 00:09:18,040
in order to convey
the message to the audience.

135
00:09:19,020 --> 00:09:20,020
About this,

136
00:09:20,250 --> 00:09:24,570
the following slide explains
how spoken language is understood,

137
00:09:25,100 --> 00:09:28,860
based on the strategic model
of discourse comprehension.

138
00:09:32,990 --> 00:09:35,860
The strategic model
of discourse comprehension,

139
00:09:35,920 --> 00:09:39,820
developed
by van Dijk and Kintsch in 1983,

140
00:09:40,180 --> 00:09:43,050
explains
how spoken language is understood.

141
00:09:43,640 --> 00:09:44,930
By adapting it,

142
00:09:45,160 --> 00:09:46,780
it is possible to claim

143
00:09:46,850 --> 00:09:49,880
that a respeaker understands
the source text

144
00:09:50,280 --> 00:09:56,090
by adopting the following strategies:
form strategies, content strategies,

145
00:09:56,550 --> 00:10:01,960
textual strategies, procedural strategies
and general knowledge strategies.

146
00:10:02,690 --> 00:10:05,890
Form strategies are related
to understanding

147
00:10:05,950 --> 00:10:08,630
the words and grammar
of the source text

148
00:10:08,690 --> 00:10:10,870
from sounds and grammar elements.

149
00:10:11,760 --> 00:10:16,640
Content strategies capitalise
on form strategies

150
00:10:16,710 --> 00:10:18,530
to understand the meaning of a text.

151
00:10:20,140 --> 00:10:23,540
Textual strategies contextualise
this content.

152
00:10:24,070 --> 00:10:26,120
Procedural strategies relate

153
00:10:26,180 --> 00:10:29,120
the ongoing comprehension
of the source text

154
00:10:29,380 --> 00:10:32,980
to other texts
similar to the one which is heard.

155
00:10:33,400 --> 00:10:34,400
Finally,

156
00:10:34,460 --> 00:10:37,860
general knowledge strategies allow
to more easily understand

157
00:10:38,030 --> 00:10:42,050
how the source text related
to the rest of the world.

158
00:10:46,080 --> 00:10:49,410
After having strategically understood
the ST,

159
00:10:49,810 --> 00:10:54,030
the respeaker adopts
production strategies to come to the TT.

160
00:10:54,960 --> 00:10:56,180
By adapting the model

161
00:10:56,250 --> 00:11:01,200
proposed by Kohn e Kalina in 1996
for simultaneous interpreting,

162
00:11:01,720 --> 00:11:04,530
it has been said that
in respeaking contexts

163
00:11:04,730 --> 00:11:07,040
professionals follow some strategies,

164
00:11:07,430 --> 00:11:09,970
as highlighted in Element 1, LO1.

165
00:11:11,100 --> 00:11:16,010
Within the production strategies
an interpreter or respeaker has to master

166
00:11:16,380 --> 00:11:19,710
the following are strongly supported
by the memory.

167
00:11:20,530 --> 00:11:21,660
There is inferencing,

168
00:11:21,750 --> 00:11:25,520
whereby the respeaker anticipates
linguistic elements.

169
00:11:26,080 --> 00:11:30,600
This means that the interpreter needs
to attentively follow the text

170
00:11:30,660 --> 00:11:33,900
if he or she wants to anticipate
some element

171
00:11:34,160 --> 00:11:37,790
in order to close the sentence
in the shortest time possible.

172
00:11:39,240 --> 00:11:40,300
There is memorising,

173
00:11:40,560 --> 00:11:44,060
whereby the respeaker postpones
linguistic material.

174
00:11:44,620 --> 00:11:48,750
This means to retrieve full sentences
while laggind behind

175
00:11:49,010 --> 00:11:50,750
because of several reasons.

176
00:11:51,780 --> 00:11:55,480
Then,
as we have expalined in Element 1 LO1,

177
00:11:55,880 --> 00:12:00,200
among the production strategies
are also editing strategies,

178
00:12:00,330 --> 00:12:04,190
either before going on air
or while being on air

179
00:12:04,520 --> 00:12:09,510
either avoiding possible mistakes
through synonymy or paraphrase

180
00:12:09,600 --> 00:12:12,870
or by correcting mistakes
before thay are aired.

181
00:12:13,550 --> 00:12:14,550
Finally,

182
00:12:14,650 --> 00:12:19,870
exit strategies or 3 Gs are
those strategies

183
00:12:19,930 --> 00:12:25,050
which are adopted in extreme conditions
to compensate for information loss.

184
00:12:25,710 --> 00:12:29,040
This can be done
through generalisation strategies,

185
00:12:29,400 --> 00:12:33,170
whereby the respeaker uses
more general synonyms

186
00:12:33,500 --> 00:12:37,160
or repeats previously mentioned content.

187
00:12:38,220 --> 00:12:41,940
Another exit strategy is
the gordian knot strategy,

188
00:12:42,410 --> 00:12:44,850
also called cut and knit strategy,

189
00:12:45,310 --> 00:12:48,780
whereby the respeaker omits
bits of a sentence

190
00:12:49,110 --> 00:12:50,850
but manages to produce

191
00:12:50,950 --> 00:12:54,020
a grammatically correct
and coherent sentence.

192
00:12:55,210 --> 00:12:58,410
The last exit strategy,
called Garwood strategy,

193
00:12:58,740 --> 00:13:01,940
is named after
my simultaneous interpreting teacher,

194
00:13:02,010 --> 00:13:04,550
who used to repeat this all the time.

195
00:13:07,350 --> 00:13:09,930
Section 3 – Memory Strategies.

196
00:13:14,350 --> 00:13:18,440
In order to deal with the complex process
of respeaking as a whole,

197
00:13:18,910 --> 00:13:22,570
a professional has to implement
strategic moves realtime,

198
00:13:23,110 --> 00:13:26,940
and memory is paramount
while working as a supportive system.

199
00:13:28,030 --> 00:13:31,660
While you can only develop
this competence by the time

200
00:13:31,730 --> 00:13:34,700
and thanks to ad hoc exercises,

201
00:13:35,130 --> 00:13:38,760
you may be interested in understanding
the rationale behind it,

202
00:13:39,150 --> 00:13:42,750
through some examples of strategies,
as a way to develop it.

203
00:13:43,380 --> 00:13:48,290
In particular, we will focus
on omissions, compression or condensation

204
00:13:48,430 --> 00:13:50,290
and expansion or addition.

205
00:13:54,860 --> 00:13:58,100
An aspect which can lead
the respeaker to omit

206
00:13:58,160 --> 00:14:01,430
a section or a sentence
of the source text

207
00:14:01,820 --> 00:14:04,700
depends on how fast the speaker is.

208
00:14:05,520 --> 00:14:10,400
Here is an example of omission
taken from the press conference

209
00:14:10,470 --> 00:14:15,060
on the appointment of the city of London
for the Olympic Games.

210
00:14:16,070 --> 00:14:18,240
The pace of the speech is very fast.

211
00:14:18,800 --> 00:14:23,720
When one of the members,
Dame Tanni Grey Thompson takes the floor,

212
00:14:24,250 --> 00:14:28,040
he starts speaking much too fast,
by saying these words:

213
00:14:28,670 --> 00:14:33,190
“I feel very proud to be
a British athlete and a Paralympian.

214
00:14:33,850 --> 00:14:38,110
There is not another country in the world
with so much attention

215
00:14:38,180 --> 00:14:40,950
to Paralympian athletes as the UK,

216
00:14:41,010 --> 00:14:44,940
both in terms of financial support
and of media coverage".

217
00:14:46,190 --> 00:14:50,420
In this case,
his way of speaking is too fast.

218
00:14:50,750 --> 00:14:53,520
This is confusing for the respeaker

219
00:14:53,620 --> 00:14:56,690
who cannot retain
the whole sentence on his memory,

220
00:14:57,350 --> 00:15:01,900
thus deciding to omit
some portions of the source text.

221
00:15:03,820 --> 00:15:04,670
He writes:

222
00:15:06,820 --> 00:15:11,770
"I feel very proud to be
a British athlete and a Paralympian.

223
00:15:12,430 --> 00:15:17,280
Both in terms of financial support
and of media coverage".

224
00:15:19,790 --> 00:15:20,790
However,

225
00:15:22,590 --> 00:15:26,550
the reader of the subtitles has
the impressions

226
00:15:26,780 --> 00:15:28,760
that something is being omitted,

227
00:15:29,090 --> 00:15:33,550
but the logics behind
the whole sentence is still maintained.

228
00:15:34,280 --> 00:15:36,190
In this case,

229
00:15:36,290 --> 00:15:40,410
this strategy really helps the respeaker.

230
00:15:41,930 --> 00:15:44,270
Important to say is also the fact

231
00:15:44,340 --> 00:15:47,840
that the question
of omission intimately concerns

232
00:15:47,900 --> 00:15:50,780
the question of quality,
as well as context.

233
00:15:51,470 --> 00:15:55,260
If an omission is considered
unquestionably valid,

234
00:15:55,530 --> 00:15:59,880
then this is surely because
“high quality” is not the same thing

235
00:15:59,950 --> 00:16:02,990
as rendering everything
in the source text.

236
00:16:03,710 --> 00:16:07,080
In fact, false starts, hesitations,

237
00:16:07,140 --> 00:16:11,000
and unnecessary repetitions
are routinely omitted,

238
00:16:11,500 --> 00:16:16,280
basically since such improvements
in the quality of discourse are seen

239
00:16:16,350 --> 00:16:19,020
as part
of the respeaker’s service function.

240
00:16:21,430 --> 00:16:25,690
The strategy of omission can be divided
into two different kind.

241
00:16:26,150 --> 00:16:31,130
Semantic omissions occur when losing
some words which are important

242
00:16:31,230 --> 00:16:33,610
for the whole understanding
of the discourse,

243
00:16:34,100 --> 00:16:38,990
and that the audience cannot get
from the rest of the context.

244
00:16:39,680 --> 00:16:45,590
The non-semantic omissions occur
when the respeaker automatically deletes

245
00:16:45,720 --> 00:16:47,440
some parts of the spoken text,

246
00:16:47,600 --> 00:16:52,450
for example: "you see"
or "it’s, it’s it’s very interesting…",

247
00:16:52,650 --> 00:16:55,190
where "it's" is repeated many times,

248
00:16:55,750 --> 00:16:57,400
and redundant words.

249
00:16:58,130 --> 00:17:00,110
These cases mainly happen

250
00:17:00,170 --> 00:17:03,280
when the respeaker has to cut
something in the speech,

251
00:17:04,990 --> 00:17:08,790
not to overcharge
his or her working-memory.

252
00:17:09,940 --> 00:17:10,970
In this cases,

253
00:17:11,030 --> 00:17:15,390
omissions go to the benefit
both of the memory load of the respeaker

254
00:17:15,620 --> 00:17:19,250
and to the benefit
of the cognitive load of the audience.

255
00:17:20,300 --> 00:17:21,720
Here are some examples.

256
00:17:23,210 --> 00:17:24,210
In example 1,

257
00:17:24,590 --> 00:17:28,920
the source text says:
"Thank you, James, for…" etcetera.

258
00:17:29,610 --> 00:17:33,570
The target text says:
"Thank you for…" etcetera.

259
00:17:34,360 --> 00:17:37,170
In example 2, source text says:

260
00:17:37,300 --> 00:17:40,370
"We determine,
as a council, three priorities".

261
00:17:41,190 --> 00:17:44,990
The target text says:
"We determined three priorities".

262
00:17:45,290 --> 00:17:49,250
In this case,
omission helps the respeaker

263
00:17:51,260 --> 00:17:54,200
in his or her cognitive load

264
00:17:54,690 --> 00:17:58,160
without losing the meaning
of the whole sentence.

265
00:17:59,740 --> 00:18:01,490
Compression is a strategy

266
00:18:01,550 --> 00:18:03,730
that overlaps with omission,
in a certain way.

267
00:18:04,690 --> 00:18:08,250
In fact, compression and omission have
something in common.

268
00:18:09,440 --> 00:18:10,630
Compression happens

269
00:18:10,700 --> 00:18:14,060
when the original meaning is rendered
by the respeaker

270
00:18:14,130 --> 00:18:16,410
in a more general and concise way,

271
00:18:16,970 --> 00:18:19,670
deleting what is repetitive or redundant.

272
00:18:20,420 --> 00:18:25,300
Omission is used when the respeaker omits
incomprehensible input,

273
00:18:25,400 --> 00:18:31,080
unnecessarily repetitive, redundant,
un-important, or unacceptable utterances.

274
00:18:32,660 --> 00:18:36,260
Therefore, the application
of one of them may entail

275
00:18:36,320 --> 00:18:37,710
the employment of the other.

276
00:18:39,580 --> 00:18:40,580
In particular,

277
00:18:40,740 --> 00:18:44,770
condensation strategy facilitates
working under time pressure,

278
00:18:45,130 --> 00:18:47,380
while conveying complete information.

279
00:18:48,900 --> 00:18:52,950
It consists in formulating
concise and synthetic utterances

280
00:18:53,250 --> 00:18:55,170
and deleting superfluous words.

281
00:18:55,990 --> 00:18:59,880
In the example that follows,
the source text reads:

282
00:19:00,580 --> 00:19:04,640
“The setup program results
in an update of the registry”.

283
00:19:05,260 --> 00:19:06,780
The target text says:

284
00:19:07,410 --> 00:19:10,540
“The setup program updates
the registry”.

285
00:19:11,340 --> 00:19:15,860
In the target text a strong verb is used
by the respeaker.

286
00:19:16,380 --> 00:19:17,510
The verb is “to update”,

287
00:19:17,900 --> 00:19:21,990
instead of its nominalized variant
in the source text,

288
00:19:22,160 --> 00:19:24,670
that is the “update of the registry”.

289
00:19:25,330 --> 00:19:27,510
This way of facilitating the sentence

290
00:19:27,800 --> 00:19:31,100
helps the respeaker retain
the right information

291
00:19:31,430 --> 00:19:34,860
by relieving the cognitive load
on the working memory.

292
00:19:37,010 --> 00:19:39,680
Expansion represents a strategy

293
00:19:39,750 --> 00:19:44,430
in which the respeaker has to trust
his or her memory system

294
00:19:44,700 --> 00:19:46,980
in terms of retention of information.

295
00:19:47,600 --> 00:19:52,680
The respeaker adds information
or expands the source discourse,

296
00:19:53,010 --> 00:19:56,080
so as to better convey
or clarify the message

297
00:19:56,500 --> 00:19:59,760
and avoid unclear information
in the target discourse.

298
00:20:00,790 --> 00:20:03,760
As we have already examined
in the previous example,

299
00:20:04,020 --> 00:20:08,150
respeakers tend to stick
to the source text as much as possible,

300
00:20:08,510 --> 00:20:12,630
and to omit something
only in cases of false starts,

301
00:20:12,930 --> 00:20:18,570
hesitations, unnecessary repetitions
or prolix and redundant ritual wording.

302
00:20:19,430 --> 00:20:22,170
Also, a respeaker can omit something

303
00:20:22,240 --> 00:20:24,910
when in trouble
with the fast pace of the speaker.

304
00:20:25,370 --> 00:20:27,720
In this framework, one could think

305
00:20:27,810 --> 00:20:32,070
that there is no much room for expansions
in the respeaking process,

306
00:20:32,730 --> 00:20:35,070
intended as addition of information

307
00:20:35,140 --> 00:20:38,280
or clarification
of particularly difficult concepts.

308
00:20:38,940 --> 00:20:42,270
However,
expansions sometimes are necessary

309
00:20:42,570 --> 00:20:48,310
in order to retrieve previous information
or to clarify not very clear concepts.

310
00:20:49,100 --> 00:20:53,320
Here is an example
of expansion at a sentence level.

311
00:20:54,740 --> 00:20:56,190
The source text says:

312
00:20:56,950 --> 00:21:00,910
“Gunmen have settled
a ferocious battle with police”.

313
00:21:01,670 --> 00:21:03,450
The target text says:

314
00:21:04,180 --> 00:21:09,230
“Reports say that people have started
to fight with the police”.

315
00:21:10,320 --> 00:21:14,610
In this case, not only
the respeaker introduces a hedging

316
00:21:14,810 --> 00:21:17,220
which is not present
in the original text,

317
00:21:17,640 --> 00:21:21,340
but he also simplifies
the lexis of the source text,

318
00:21:21,740 --> 00:21:25,660
without weakening the language
at the level of the sentence.

319
00:21:27,930 --> 00:21:28,930
Summary.

320
00:21:30,760 --> 00:21:36,670
In this LO2, we have introduced
the main competence of the respeaker,

321
00:21:37,100 --> 00:21:39,410
the psycho-cognitive skill of memory.

322
00:21:40,040 --> 00:21:43,500
In particular, we have dealt
with two important aspects,

323
00:21:43,860 --> 00:21:48,020
a theoretical one that is
the memory divided into sensory memory,

324
00:21:48,250 --> 00:21:52,050
long-term memory
and short-term memory or working memory,

325
00:21:52,540 --> 00:21:54,420
and a more practical one,

326
00:21:54,770 --> 00:21:57,680
on the inferencing
and memorising strategies

327
00:21:57,840 --> 00:21:59,260
used by the respeaker.

328
00:21:59,820 --> 00:22:02,070
In the case of inferencing strategies,

329
00:22:02,130 --> 00:22:06,360
the respeaker recovers
lost or incomprehensible information

330
00:22:06,450 --> 00:22:10,910
on the basis of the speech context
and his or her general knowledge,

331
00:22:11,470 --> 00:22:15,500
or he anticipates lexico-grammar
or conceptual elements.

332
00:22:16,160 --> 00:22:18,200
In the case of memorising strategies,

333
00:22:18,470 --> 00:22:23,280
the respeaker postpones
lexico-grammar or conceptual elements.

334
00:22:23,910 --> 00:22:27,940
One way to go for that is
trying to semi-automatise

335
00:22:28,000 --> 00:22:29,090
these actions.

336
00:22:29,520 --> 00:22:34,140
To do so,
memorising is the exercise to start with

337
00:22:34,540 --> 00:22:35,960
together with inferencing.

338
00:22:36,390 --> 00:22:40,510
In the homework session,
we will see how to do this in practice.