Content uploaded by Sijia Chen
Author content
All content in this area was uploaded by Sijia Chen on Sep 17, 2020
Content may be subject to copyright.
The Journal of Specialised Translation Issue 34 – July 2020
100
The impact of directionality on the process and product in
consecutive interpreting between Chinese and English: Evidence
from pen recording and eye tracking
Sijia Chen, Southwest University, Chongqing
ABSTRACT
Directionality is a recurring topic in interpreting studies. Focusing on consecutive
interpreting (CI), this study aims to contribute further empirical data to the topic by
analysing the impact of directionality on the process and product of CI, with special
attention paid to the cognitive aspect. Pen recording and eye tracking were used to record
CI with notes amongst professional interpreters with Chinese as L1 and English as L2. The
direction of interpreting was found to affect the cognitive processing and product in both
phases of CI. Phase I of L2 to L1 interpreting seemed to burden the interpreters with a
higher level of cognitive load than L1 to L2 interpreting, leading to a strategic choice of
using more language notes (full words rather than abbreviations) and fewer symbols.
Phase II of L2 to L1 interpreting appeared to be less cognitively demanding than the other
direction and the interpreters produced a more fluent target speech, but the target speech
was less accurate.
KEYWORDS
Directionality, consecutive interpreting, process, product, cognitive load.
1. Introduction
Directionality is a recurring topic in translation and interpreting studies (Gile
2005). It refers to the question as to whether there is any difference
between working from an interpreter’s first language (L1) into a second
language (L2) or the reverse1. On the one hand, interpreters may have
personal preferences to work into a certain language. For example, they
may prefer to produce interpretations into their L1, which is stronger, given
the greater cognitive effort required for language production (Nicodemus
and Emmorey 2013). On the other hand, directionality is a potentially
important factor that impacts the quality of interpretation (Godijns and
Hinderdael 2005; Kalina 2005). Some have even argued that interpreters
can only create high-quality target speech when working into their L1,
especially in simultaneous interpreting (SI) (Seleskovitch 1999).
However, most studies focusing on the issue of directionality in interpreting
have only investigated SI and signed language interpreting, while
consecutive interpreting (CI) has received inadequate attention. Apart from
that, there is not enough empirical evidence overall and the jury is still out
there on the topic (Gile 2005). This study aims to contribute further
empirical data by analysing both the process and product of CI and in both
L2 to L1 and L1 to L2 directions, with special attention paid to the cognitive
aspect. The issue of directionality is not only meaningful for translation and
interpreting research. This study helps to shed light on the ways the brain
The Journal of Specialised Translation Issue 34 – July 2020
101
engages cognition to language comprehension and language production,
while the differences found between the two directions of CI could be taken
as evidence for the asymmetry between L1 and L2 processing.
2. Directionality in interpreting
One of the regularly recurring themes in interpreting research is that of
directionality and the associated issue of whether preference should be
given to interpreting into L1 or L2. This section will provide a brief overview
of the relevant literature, with a main focus on the empirical findings, rather
than the theoretical views such as those by Seleskovitch and Lederer (1989)
and Denissenko (1989), which have already been reviewed in many related
studies (see Gile (2005) for example).
Of the previously studied themes on interpreting directionality, two emerge
as the most discussed: (1) the influence of directionality on interpreting,
especially the quality of interpreting; (2) the factors interacting with
directionality, particularly those that might cancel out its influence.
2.1. The influence of directionality on interpreting
Studies in this group mainly focus on the influence of directionality from
three aspects: interpreting quality (e.g. Tommola and Helevä 1998; Van
Dijk et al. 2011; Fu 2012, 2013; Nicodemus and Emmorey 2015; Lin et al.
2018), strategy use (e.g. Al-Salman and Al-Khanji 2002; Bartłomiejczyk
2006; Chang and Schallert 2007; Gumul 2017; Wu and Liao 2018), and
cognitive load (e.g. Rinne et al. 2000).
The impact on interpreting quality is the most extensively researched aspect,
with the majority of the attention paid to information completeness, fluency
of delivery, and target language quality.
With novice SI interpreters as participants, a study by Tommola and Helevä
(1998) found a slight advantage of the L1 to L2 direction in information
completeness as measured by propositional accuracy. Experimenting with
professional SI interpreters, Rinne et al. (2000) reported higher information
completeness in the L1 to L2 direction. A similar result was found in signed
language interpreting: Van Dijk et al. (2011) found that their professional
interpreters produced more accurate products when working from L1
(spoken Dutch) into L2 (Sign Language of the Netherlands).
Standing in contrast, Chang and Schallert (2007) detected higher
information completeness rendered in L2 to L1 SI with Chinese L1
interpreters (but no difference was found with English L1 interpreters).
Nicodemus and Emmorey (2015) also observed that in signed language
interpreting, novice interpreters were more accurate when interpreting into
their L1, whereas expert interpreters were equally accurate in the two
directions.
The Journal of Specialised Translation Issue 34 – July 2020
102
Directionality was found to be an important factor that influences the
fluency of delivery in interpreting. In Lin et al. (2018), the SI performance
of novice interpreters was more fluent in the L2 to L1 direction. The same
trend was found in signed language interpreting. In a group of professional
interpreters, Wang and Napier (2015) found that native signers performed
significantly better in terms of delivery in the English (L2) to Auslan (L1)
direction. Similarly, Nicodemus and Emmorey (2015) found that their
novice interpreters’ speed of interpretation was scored significantly higher
(meaning the delivery was well-paced, neither too fast nor too slow) when
working into English (L1) than into American Sign Language (L2). At the
same time, however, they found that professional interpreters were slightly
(but not significantly) better when interpreting from L1 into L2 in terms of
flow.
Only a couple of studies tackled the issue of directionality in CI, and the
focus was invariantly on fluency. Mead (2005) used both novice and
professional interpreters as participants and found that they were both
more fluent in L2 to L1 interpreting, with novices making a high proportion
of long filled pauses in L2 production, showing marked disfluencies. Fu
(2013) detected a significant directional impact on novice interpreters’
disfluency features. He found more ungrammatical pauses in the L1 to L2
direction, but more repairs in the L2 to L1 direction.
In terms of the influence of directionality on target language quality,
Nicodemus and Emmorey (2015) reported that novice signed language
interpreters produced more natural prosody when working into L1, but there
was no difference between the two directions in expert interpreters. Wang
and Napier (2015) found that the native signers were significantly superior
to the non-native signers in terms of the target text features in English-to-
Auslan interpreting.
Varied strategic approaches may be taken by interpreters to cope with the
different demands of the two interpreting directions. For example,
Bartłomiejczyk (2006) found that, in SI from L2 to L1, novice interpreters
used more inferencing, parallel reformulation, and transcodage, whereas in
SI from L1 to L2, they used more syntactic transformation, approximation,
and paraphrase. In a study looking specifically at explicitation in
interpreting, Gumul (2017) found that novice interpreters used explicitation
more frequently in L1 to L2 SI, especially by means of adding connectives,
reiteration, meaning specification, and disambiguating metaphors.
According to the retrospective report, this was mainly due to the
interpreting constraints felt by the participants in L2 production, and they
were adopting repair or preventive strategies.
The influence of directionality on cognitive load in interpreting has been
investigated in several studies which started from an interdisciplinary
perspective. Gran and Fabbro (1988) reported significantly increased
The Journal of Specialised Translation Issue 34 – July 2020
103
disruption in the tapping rate for L1 to L2 SI as compared to the other
direction, due to the less automatic and more cognitively taxing language
production control in L2. Kurz (1995) detected EEG differences between the
two SI directions, and found an increase of interhemispheric coherence in
the beta band in the L1 to L2 direction, indicating a higher level of cognitive
load. Hyönä et al. (1995) used pupil dilation as a measure of cognitive load
in SI and found that the pupil dilated more when the subjects performed
lexical translation from L1 to L2 than when they translated from L2 to L1.
Rinne et al. (2000) observed that the brain activation patterns of
interpreters were modulated by direction, and recorded more extensive
activation during SI from L1 to L2.
2.2. Factors that interact with directionality in interpreting
The influence of directionality on interpreting, albeit clearly shown by
empirical evidence in various studies, is modulated and sometimes even
“offset” (Gile 2005: 9) by other relevant factors. These factors that interact
with directionality in interpreting can be roughly divided into two types:
interpreter-related and task-related factors. On the one hand, the
interpreter characteristics of motivation (Pavlović 2007), training and
working experience (Fernández 2005; Chmiel 2016), language proficiency
(de Bot 2000), and working memory (Lin et al. 2018) have a bearing on
how individual interpreters might deal with interpreting into different
directions. On the other hand, the task characteristics, such as interpreting
mode (Nicodemus and Emmorey 2013), language pair (Padilla 2005),
features of the speech (Dose 2017), features of the speaker and the
audience (Chang and Schallert 2007), and market (Fernández 2005; Lim
2005), need to be carefully controlled or systematically manipulated so as
not to obscure the actual influence of directionality.
More importantly, the interpreter and task characteristics often work
together to jointly interact with directionality. Survey studies (Lim 2005;
Nicodemus and Emmorey 2013) confirmed the stronger motivation to work
into L1, supporting the traditional bias of spoken language interpreters for
the L2 to L1 direction (e.g. Seleskovitch and Lederer 1989: 135; Déjean le
Féal 1998, 2005). But this preference was found to be modulated by
language pair, interpreting mode, and training and working experience.
Another survey conducted by Pavlović (2007) among translators and
interpreters in Croatia (mostly working with L1 Croatian and L2 English)
showed that one-third of the respondents preferred the L1 to L2 direction,
one-third preferred the other direction, and the rest reported having no
preference regarding directionality. Furthermore, questionnaire results
from Al-Salman and Al-Khanji (2002) showed that the majority of
respondents were more comfortable when interpreting from Arabic into
English (L1 to L2) than vice versa.
The Journal of Specialised Translation Issue 34 – July 2020
104
In signed language interpreting, the anecdotal preference of L1 (spoken
language) to L2 (signed language) interpreting was confirmed by
Nicodemus and Emmorey (2013). Apart from discovering the different
directionality preferences among spoken and signed language interpreters,
the study also revealed that spoken language interpreters received an
education with a slight bias for more training in L2 to L1 interpreting, while
the signed language interpreters received almost twice as much instruction
in the L1 to L2 direction.
The above-mentioned literature highlights the relation between
directionality and interpreting. Nevertheless, the majority of the studies
cast their attention on SI and signed language interpreting, leaving CI quite
under-researched in comparison. Since the interpreting mode could
potentially affect whether or not an interpreter would work in one direction
or the other (for example, in Lim (2005), those who were willing to work
into L2 in CI preferred not to work in that direction in SI), it would be
beneficial to devote more research effort to the consecutive mode. In
addition, research into the impact of directionality on CI will be meaningful
for interpreter education programmes to enhance curriculum designs.
Furthermore, interpreting is a cognitively demanding language task which
can serve as a useful tool to investigate the cognitive processes involved in
the bilingual brain, and CI taps into different mental processes from the
other interpreting modes. This study, as a result, tries to contribute some
empirical evidence on the topic of directionality in CI.
2.3. Tapping into the process and product of CI
According to Gile’s Effort Models (2009: 175–176), phase I of CI is the
comprehension phase, consisting of the sub-processes of listening and
analysis, note-taking, short-term memory operations, and coordination;
phase II is the speech production phase, consisting of the sub-processes of
remembering, note-reading, and production.
The product of phase I is the written notes produced by the interpreters,
whereas the product of phase II is the translated speech in the target
language. Extensive research has been carried out on the product of both
phase I (i.e. the notes, see Chen (2016) for a review) and phase II (i.e. the
target speech, see Collados Aís and García Becerra (2015) for an overview).
The process of CI, however, seems under-researched in comparison. Earlier
process research on CI relied on the method of video recording (e.g. Andres
2002), which involved manually checking the timing of note-writing, was
labour-intensive and not very accurate. Digital pen recording is a newer and
viable research method to tap into phase I of CI, which has been applied in
more recent studies (Orlando 2010, 2014; Chen 2017, 2020). As to phase
II, a research method with substantial potential for investigating the
cognitive processes is eye tracking, a method that has long been applied in
fields such as reading and cognitive sciences, but only became a popular
The Journal of Specialised Translation Issue 34 – July 2020
105
research method in Translation Studies during roughly the past decade
(Hvelplund 2017).
This study used digital pen recording and eye tracking to collect data from
a group of professional interpreters while they were performing CI in both
the L2 to L1 and the L1 to L2 direction, with an aim to further our
understanding of directionality in CI. The study attempts to address the
following questions:
1. What is the impact of directionality on the process and product of CI
phase I, the comprehension phase?
2. What is the impact of directionality on the process and product of CI
phase II, the speech production phase?
More specifically, the interest lies in exploring the impact of directionality
on the process of CI phase I (as indicated by the temporal features of note-
writing captured via pen recording), the product of phase I (the written
notes), the process of phase II (as indicated by the eye movements
captured via eye tracking), and the product of phase II (as indicated by the
information completeness and fluency of delivery of the target speech).
3. Method
3.1. Participants, interpreting tasks, and procedures
The data collection period lasted from August 2015 to July 2016. A sample
of 26 professional interpreters with Chinese as L1 and English as L2 were
recruited to perform CI with notes. They all had Professional Interpreter
accreditation by NAATI, the Australian National Accreditation Authority for
Translators and Interpreter. The sample mean age was 36.4 years (SE =
2.36). The participants had an average of 7.4 years of professional
experience (SE = 1.43) and had provided 167 CI services on average in the
past 12 months (SE = 58.1). They lived and worked in Australia, a country
where their L2 was spoken.
Two source speeches (one in Chinese and one in English) were created and
strictly controlled for variances to make the two CI tasks as similar as
possible (see Table 1), differing only in directionality. Here is a summary of
the procedures (please refer to (Chen 2017) for more details): (1) two
English scripts on similar topics were created and an analysis by CPIDR 5
(Brown et al. 2008) showed that they had similar word count, proposition
count, and idea density2; (2) one of the scripts was translated by the author
(who is a professional translator/interpreter) to make a Chinese script; (3)
two speeches were recorded by native speakers of Chinese and English in
sound-proof studios using the relevant scripts; (4) the recorded speeches
were edited using Audacity and controlled for variables such as pauses,
duration, and speed.
The Journal of Specialised Translation Issue 34 – July 2020
106
Table 1. A summary of the two tasks.
The fact that the C-E script was a translated text rather than originally
drafted in Chinese potentially limits the comparability of E-C and C-E texts.
In order to mitigate this effect, the translated text was refined stylistically
and grammatically by two Chinese speaking editors working at a local
Chinese radio station. The editors were highly professional for the job due
to the nature of their work (editing scripts for radio broadcasting). It was
also worth mentioning that the texts, although talking about business
registration and property purchase, were of a general rather than
(semi-)legal nature and no legal terminology was involved (please refer to
(Chen 2018) for a copy of the texts).
The participants were tested individually in a language laboratory. They
performed the two CI tasks in randomised order so that about half of them
performed L2 to L1 interpreting first, and other L1 to L2 interpreting first.
They were allowed to take a break for as long as they needed between the
tasks. During the experiment, the participants’ note-taking process was
recorded by a digital tablet equipped with a digital pen, powered by the Eye
and Pen software (http://eyeandpen.net/en/). The entire interpreting
process was also recorded by a head-mounted eye tracker (the SMI ETG).
A retrospection session (audio-recorded) cued by the written notes followed
the tasks, and the participants were asked to recall whatever they could
remember about the interpreting process, and report the content, form, and
language of each note unit.
3.2. Data and analysis
Due to the combined data loss of pen recording and eye tracking, only 18
participants had both their pen and eye data successfully collected.
During the experiment, the Eye and Pen software recorded not only all the
written note units, but also the timestamp of every event, including the
start/end time for writing each note unit and the start time for playing each
source speech audio, allowing for the calculation of the ear-pen span (EPS).
The EPS was calculated as the time span between the moment a speech
unit was heard and the moment it was written down as notes, and used as
an indicator of the cognitive processing in phase I (see (Chen 2017) for
Direction Topic Word
count
Proposition
count Idea
density Duration
E-C
How to register a
business in Australia
631 321 0.51 4'59"
C-E
How to purchase a
property in Australia
630* 324* 0.51* 4'47"
* The calculations were based on the original English script to enable
comparison between the two texts. The Chinese character count was 944.
The Journal of Specialised Translation Issue 34 – July 2020
107
more detailed explanations). The form and language of the notes were also
analysed, following the rules specified in Dam (2004a, 2004b).
The eye tracker recorded the eye movements and a video of the entire
interpreting process. The eye measure of average fixation duration was
used as an indicator of the cognitive processing in phase II. The quality of
interpreting was measured in terms of information completeness and
fluency of delivery. To measure information completeness, the target
speeches were transcribed, propositionalised, and scored against the source
text propositions; a group of three raters assessed the information
completeness and the final score was the average of the three (see (Chen
2017) for more specifications). To measure fluency of delivery, Audacity
was used to prepare the audios for the automatic rating (e.g. the source
speech was removed from the recording). Praat (Boersma and Weenink
2018) was used to automatically analyse the target speech audios, using
Praat Script Syllable Nuclei v2 (Quené et al. 2010) with default settings.
Paired samples t-tests were used to compare the two directions of
interpreting. All statistical analyses were performed by running IBM SPSS
Statistics 24. Two-tailed p values less than 0.05 were considered to be
statistically significant. Cohen's d (the difference between the means
divided by the pooled standard deviation) was used to indicate the effect
sizes, which were classified as small (d = 0.2), medium (d = 0.5), and large
(d = 0.8).
4. Results
4.1. The impact of directionality on CI phase I
The time lag between listening and note-writing, namely the EPS (reported
in milliseconds), was significantly different between the two directions
(Figure 1). The EPS in the L2 to L1 task (M = 2262, SD = 635) was
significantly shorter than that in the L1 to L2 task (M = 2632, SD = 681);
t(21) = -3.39, p = .003, d = 0.72. This means that the interpreters followed
the source speech more closely when listening to an L2 speech compared
to an L1 speech.
The Journal of Specialised Translation Issue 34 – July 2020
108
Figure 1. The EPS in the two directions of interpreting.
There was no significant difference in the quantity of notes taken by the
participants between the L2 to L1 direction (M = 182.14, SD = 27.71) and
the L1 to L2 direction (M = 179.32, SD = 27.56), t(21) = 0.90, p = .38.
The participants in this study preferred language to symbol in interpreting
overall, but there were differences between the two directions (Table 2).
The percentage of language notes in the L2 to L1 direction was significantly
higher than that in the L1 to L2 direction, whereas the percentage of symbol
notes in the L2 to L1 direction was significantly lower than that in the L1 to
L2 direction. This means that the participants used more language notes
and fewer symbol notes when listening to an L2 speech compared to an L1
speech.
L2 to L1
L1 to L2
M
SD
M
SD
t
df
d
Language (%)
75.2%
7.3%
68.6%
8.1%
8.49***
21
1.81
Symbol (%)
19.8%
7.6%
22.5%
8.5%
-3.59**
21
0.77
** p < .01, *** p < .001
Table 2. Differences between the two directions in terms of the
language/symbol choice.
The participants did not have a clear preference for abbreviation or full word
in either direction. There was a significant directional difference in terms of
their use of full words, but not abbreviations (Table 3). The percentage of
full word notes in the L2 to L1 direction was significantly higher than that in
the L1 to L2 direction. That is to say, the participants used a higher
proportion of full words in note-taking when listening to L2 as opposed to
L1.
The Journal of Specialised Translation Issue 34 – July 2020
109
L2 to L1
L1 to L2
M
SD
M
SD
t
df
d
Abbreviation (%)
39.6%
10.2%
37.6%
7.3%
Full word (%)
35.5%
7.6%
31.0%
8.5%
3.76**
21
0.80
** p < .01
Table 3. Differences between the two directions in terms of the
abbreviation/full word choice.
Looking at the two sets of data together (Table 2 and Table 3), we can see
that the participants not just used more language notes when interpreting
from L2 to L1, but the increased language notes were mostly full words
instead of abbreviations.
The participants in this study preferred L2 (English) to L1 (Chinese) when
choosing the language for note-taking in both directions, but there were
differences between the two directions (Table 4). The percentage of L1
notes in the L2 to L1 direction was significantly lower than the other
direction, whereas the percentage of L2 notes in the L2 to L1 direction was
significantly higher than the other direction. What this shows is that the
participants’ preference for L2 in note-taking was modulated by the
direction of interpreting: the preference was enhanced when
comprehension was in L2, and weakened when it was in L1.
L2 to L1
L1 to L2
M
SD
M
SD
t
df
d
L1 (%)
16.4%
13.4%
26.7%
16.4%
-2.60*
21
0.55
L2 (%)
58.8%
15.2%
41.9%
16.5%
4.16***
21
0.89
* p < .05, *** p < .001
Table 4. Differences between the two directions in terms of the choice of
language.
4.2. The impact of directionality on CI phase II
In phase II, the participants’ average fixation duration in the L2 to L1
direction was shorter than that in the L1 to L2 direction, and the data shows
a clear tendency to significance (Table 5). This seems to indicate that when
the participants were producing a target speech in L1, the cognitive load
tended to be lower.
The Journal of Specialised Translation Issue 34 – July 2020
110
L2 to L1
L1 to L2
M
SD
M
SD
t
df
d
Average fixation
duration
271.48 70.99 281.77 76.34 -2.03* 17 0.48
* p = .058
Table 5. Difference between the two directions in the average fixation duration
in CI phase II.
There were significant differences between the two directions in terms of
the information completeness of the target speech (Table 6). The
information completeness score in the L2 to L1 direction was significantly
lower than that in the L1 to L2 direction. That is to say, the quality of
interpreting performance in terms of information completeness was better
in L1 to L2 interpreting compared with L2 to L1 interpreting.
L2 to L1
L1 to L2
M
SD
M
SD
t
df
d
Information
completeness
73.27 9.54 77.24 8.73 -3.72** 21 0.79
** p < .01
Table 6. Difference between the two directions in terms of information
completeness.
Praat was used to analyse the target speech on such parameters as the
number of silent pauses, duration, phonation time, and speech rate
(number of syllables/duration). The duration and phonation time of the
target speeches in the two directions were not significantly different (Table
7). Differences were found in the fluency of delivery between the two
directions as measured by the number of silent pauses and speech rate
(Table 7). The number of silent pauses in L2 to L1 direction was significantly
lower than that in the L1 to L2 direction, while the speech rate of the L2 to
L1 direction was significantly higher than that in the L1 to L2 direction.
Taken together, the data show that when producing an L1 target speech,
the participants had fewer pauses and a higher speech rate, pointing to a
better fluency of delivery in the L2 to L1 direction.
L2 to L1
L1 to L2
M
SD
M
SD
t
df
d
No. of silent
pauses
121.00 41.39 160.73 44.67 -4.68*** 21 1.00
Duration (s) 382.88 82.92 392.89 86.93 -.56 21
Phonation
time (s)
283.54 46.74 265.62 60.24 1.79 21
Speech rate 3.31 0.53 2.95 0.49 4.79*** 21 1.00
*** p < .001
Table 7. Differences between the two directions in terms of fluency of delivery.
The Journal of Specialised Translation Issue 34 – July 2020
111
I also calculated the proportion of pause duration in the target speech and
found that silent pauses counted for 24.5% of the entire duration of speech
in the L2 to L1 direction, and this was significantly lower than the 31.8% in
the L1 to L2 direction (t(21) = -4.85, p < .001, d = 1.03). This is converging
evidence for better fluency in L1 target speech production.
5. Discussion
The impact of directionality was found on both the process and product in
the two phases of CI. In the listening and note-taking phase, when the
participants listened to an L2 speech, they acted more hastily and there was
a shorter time lag between listening and note-taking. An earlier study (Chen
2020) showed that a very short EPS was detected when interpreters came
across numbers, way shorter than the average EPS in the entire task.
Considering that numbers have long been regarded as a drain on the
cognitive resources (Mead 2015), a shorter EPS might be the interpreters’
strategy to cope with the high cognitive load. Similarly, the shorter EPS in
L2 to L1 interpreting found in this study might indicate a strategic choice
resulting from a higher level of cognitive load associated with listening and
comprehension in L2 compared with L1. This is both intuitively plausible and
empirically supported by listening studies (e.g. Borghini and Hazan 2018).
The product of CI phase I, that is the notes, showed differences in both the
form and language between the two directions. The participants not only
used a higher proportion of language notes and a lower proportion of symbol
notes in L2 to L1 interpreting than vice versa, but their increased proportion
of language notes were full words, rather than abbreviations. That is to say,
when faced with the more cognitive demanding L2 listening and
comprehension, the interpreters opted for fewer symbols and more full
words in note-taking. This is in line with previous studies which have stated
that if symbols and abbreviations are not fully mastered by the interpreter,
“retrieving them from one’s memory when they are needed for writing may
take up too much time and processing capacity” (Gile 2009: 179).
In terms of the language of note-taking, this study found that the
participants preferred L2, which was English, in both directions of
interpreting. This might have to do with the interpreters’ working
experience and L2 competence (the participants are professional
interpreters who live and work in a country where their L2 is spoken), as
well as the language combination (compared to Chinese, the alphabetic
language of English might be preferred because phonetic spelling and
misspelling can be used for note-taking).
However, directionality had an impact on this preference for L2. The
preference was enhanced when interpreters were listening to an L2 speech
and weakened when they were dealing with an L1 speech. A possible
explanation is that comprehension in L2 and taking notes in L2 is intra-
The Journal of Specialised Translation Issue 34 – July 2020
112
language processing, arguably less cognitively demanding than the inter-
language processing of comprehension in L1 and taking notes in L2.
In phase II of CI, this study found that the average fixation duration during
note-reading was shorter in the L2 to L1 direction, potentially indicating a
lower level of cognitive load. This is in line with the findings in a number of
studies that production in L1 is less cognitively taxing than in L2 (Hyönä et
al. 1995; Kurz 1995; Gran and Fabbro 1988; de Bot 2000; Rinne et al.
2000).
The quality of interpreting was also found to be affected by directionality.
This study measured the target speech quality on two criteria: information
completeness and fluency of delivery.
It was found that the information completeness was higher in the L1 to L2
direction, corroborating findings in some studies (Tommola and Helevä
1998; Rinne et al. 2000; Van Dijk et al. 2011) but contradicting others
(Chang and Schallert 2007; Nicodemus and Emmorey 2015). For example,
Chang and Schallert (2007) observed higher information completeness in
SI from L2 to L1. The current study differs from their study in the
interpreting mode studied, and also the participants in this study lived and
worked in an English-speaking country and had high L2 proficiency. These
could potentially explain the contradiction in the findings. As to Nicodemus
and Emmorey (2015), they studied signed language interpreting, and it was
likely that the mode of interpreting came into play, causing the difference
in the findings.
This study also examined the fluency of delivery between the two directions
of interpreting and found that the fluency of L2 to L1 interpreting was better
than that of L1 to L2 interpreting, as evidenced by the number of silent
pauses, speech rate, and proportion of pause duration. This finding echoes
a number of studies which found L1 production to be more fluent, be it in
SI (Lin et al. 2018), CI (Mead 2005), or signed language interpreting
(Nicodemus and Emmorey 2015; Wang and Napier 2015).
6. Conclusion
This study examined a group of Chinese(L1)/English(L2) professional
interpreters while they performed CI in both directions. To the best of my
knowledge, this study is the first to examine the varied influences of
directionality on both the process and the product of the two phases in CI.
In phase I, the comprehension phase, the participants followed the source
speech more closely and used higher proportions of language and full word
notes and a lower proportion of symbol notes when listening to L2 compared
with L1. The evidence seems to reveal that, faced with a higher level of
cognitive load in L2 comprehension, interpreters would adapt their strategy
by using more language notes (full words rather than abbreviations) and
The Journal of Specialised Translation Issue 34 – July 2020
113
less symbols. In phase II, producing a target speech in L1 was associated
with a lower average fixation duration, better fluency, but a lower level of
information completeness. The data seems to indicate that while the
interpreters were experiencing a lower level of cognitive load during L1
production, generating a more fluent target speech, the target speech was
less accurate than the other direction. Findings of this study could be
informative for developing compatible training strategies to tackle the
specific challenges in the two interpreting directions.
Examining the findings in this study together with those in previous studies,
it becomes evident that a series of factors might come into play on
directionality in interpreting. This study managed to identify some clear
differences between the two directions by confining the scope of research.
It only looked at one interpreting mode (CI), with only one type of
participant (professional interpreters) and one language combination
(Chinese L1 and English L2). It would be very interesting to see how
generalisable these findings are when some of the parameters are
systematically manipulated.
What this study would also like to conclude is that CI provides a unique
avenue for investigating the links between different languages in the human
brain. What CI presents is a challenging bilingual processing task naturally
divided into two distinct phases, one featuring comprehension and the other,
production. With further effort devoted into the topic, the differences in
cognitive processing found between the two directions of CI in the two
phases could provide potential evidence for the asymmetry between L1 and
L2 processing, and between language comprehension and production.
Acknowledgments
This work has been supported by the Fundamental Research Funds for the
Central Universities (grant number SWU1809688) and the China National
Committee for Translation & Interpreting Education (grant number
MTIJZW201903).
References
• Al-Salman, Saleh and Raja’i Al-Khanji (2002). "The native language factor in
simultaneous interpretation in an Arabic/English context." Meta 47(4), 607–626.
• Andres, Dörte (2002). Konsekutivdolmetschen und notation. Frankfurt: Peter Lang.
• Bartłomiejczyk, Magdalena (2006). "Strategies of simultaneous interpreting and
directionality." Interpreting 8(2), 149–174.
• Boersma, Paul and David Weenink (2018). "Praat: doing phonetics by computer."
(Version 6.0.37). http://www.praat.org/ (consulted 25.05.2019).
The Journal of Specialised Translation Issue 34 – July 2020
114
• Borghini, Giulia and Valerie Hazan (2018). "Listening effort during sentence
processing is increased for non-native listeners: A pupillometry study." Frontiers in
Neuroscience 12, 1–13.
• Brown, Cati et al. (2008). "Automatic measurement of propositional idea density from
part-of-speech tagging." Behavior Research Methods 40(2), 540–545.
• Chang, Chia-chien and Diane L. Schallert (2007). "The impact of directionality on
Chinese/English simultaneous interpreting." Interpreting 9(2), 137–176.
• Chen, Sijia (2016). "Note-taking in consecutive interpreting: A review with special
focus on Chinese-English literature." The Journal of Specialised Translation 26, 151–
171.
• Chen, Sijia (2017). "Note-taking in consecutive interpreting: New data from pen
recording." Translation and Interpreting 9(1), 4–23.
• Chen, Sijia (2018). Exploring the Process of Note-taking and Consecutive Interpreting:
A Pen-eye-voice Approach towards Cognitive Load. PhD thesis. Macquarie University.
• Chen, Sijia (2020). "The process of note-taking in consecutive interpreting: A digital
pen recording approach." Interpreting 22(1), 117–139.
• Chmiel, Agnieszka (2016). "Directionality and context effects in word translation tasks
performed by conference interpreters." Poznań Studies in Contemporary Linguistics
52(2), 269–295.
• Collados Aís, Ángela and Olalla García Becerra (2015). "Quality." Holly Mikkelson
and Renée Jourdenais (eds) (2015). The Routledge Handbook of Interpreting. New York:
Routledge, 368–383.
• Dam, Helle V. (2004a). "Interpreters' notes: On the choice of form and language."
Gyde Hansen, Kirsten Malmkjær and Daniel Gile (eds) (2004). Claims, Changes and
Challenges in Translation Studies. Amsterdam: John Benjamins, 251–261.
• Dam, Helle V. (2004b). "Interpreters' notes: On the choice of language." Interpreting
6(1), 3–17.
• de Bot, Kees (2000). "Simultaneous interpreting as language production." Birgitta
Englund Dimitrova and Kenneth Hyltenstam (eds) (2000). Language Processing and
Simultaneous Interpreting: Interdisciplinary Perspectives. Amsterdam/Philadelphia:
John Benjamins, 65–88.
• Déjean le Féal, Karla (1998). "Non nova, sed nove." The Interpreters' Newsletter 8,
41–49.
• Déjean le Féal, Karla (2005). "Can and should interpretation into a second language
be taught." Rita Godijns and Michaël Hinderdael (eds) (2005). Directionality in
Interpreting: The ‘Retour’ or the Native? Ghent: Communication and Cognition, 167–
194.
• Denissenko, Jurij (1989). "Communicative and interpretative linguistics." Laura Gran
and John Dodds (eds) (1989). The Theoretical and Practical Aspects of Teaching
Conference Interpretation. Udine: Campanotto, 155–158.
• Dose, Stefanie (2017). "Assessing directionality in context." Stellenbosch Papers in
Linguistics 47, 67–87.
The Journal of Specialised Translation Issue 34 – July 2020
115
• Fernández, Emilia Iglesias (2005). "Bidirectionality in interpreter training in Spanish
universities: An empirical study." Rita Godijns and Michaël Hinderdael (eds) (2005).
Directionality in Interpreting: The ‘Retour’ or the Native? Ghent: Communication and
Cognition, 101–125.
• Fu, Rongbo (2012). "英汉双向交替传译中译语停顿的对比研究."
外语教学与研究
44(3), 437–
447.
• Fu, Rongbo (2013). "口译方向性对译语非流利产出的影响."
现代外语
(
季刊
) 36(2), 198–205.
• Gile, Daniel (2005). "Directionality in conference interpreting: A cognitive view." Rita
Godijns and Michaël Hinderdael (eds) (2005). Directionality in Interpreting: The ‘Retour’
or the Native? Ghent: Communication and Cognition, 9–26.
• Gile, Daniel (2009). Basic Concepts and Models for Interpreter and Translator Training
(revised edition). Amsterdam/Philadelphia: John Benjamins.
• Godijns, Rita and Michaël Hinderdael (2005). "Introduction." Rita Godijns and
Michaël Hinderdael (eds) (2005). Directionality in Interpreting: The ‘Retour’ or the
Native? Ghent: Communication and Cognition, 3–8.
• Gran, Laura and Franco Fabbro (1988). "The role of neuroscience in the teaching of
interpretation." The Interpreters’ Newsletter 1, 23–41.
• Gumul, Ewa (2017). "Explicitation and directionality in simultaneous interpreting."
Linguistica Silesiana 38, 311–329.
• Hvelplund, Kristian T. (2017). "Eye tracking in translation process research." John W.
Schwieter and Aline Ferreira (eds) (2017). The Handbook of Translation and Cognition.
Hoboken: John Wiley & Sons, 248–264.
• Hyönä, Jukka, Tommola, Jorma and Anna-Mari Alaja (1995). "Pupil dilation as a
measure of processing load in simultaneous interpretation and other language tasks."
The Quarterly Journal of Experimental Psychology 48(3), 598-612.
• Kalina, Sylvia (2005). "Quality in the interpreting process: What can be measured and
how?". Rita Godijns and Michaël Hinderdael (eds) (2005). Directionality in Interpreting:
The ‘Retour’ or the Native? Ghent: Communication and Cognition, 27–46.
• Kurz, Ingrid (1995). "Interdisciplinary research—Difficulties and benefits." Target 7(1),
165–179.
• Lim, Hyang-Ok (2005). "Working into the B language: The condoned taboo?" Meta
50(4), 1–18.
• Lin, Yumeng et al. (2018). "Predicting fluency with language proficiency, working
memory, and directionality in simultaneous interpreting." Frontiers in Psychology 9, 1–
13.
• Mead, Peter (2005). "Directionality and fluency: An experimental study of pausing in
consecutive interpretation into English and Italian." Communication and Cognition.
Monographies 38(1–2), 127–146.
• Mead, Peter (2015). "Numbers." Franz Pöchhacker (ed.) (2015). Routledge
Encyclopedia of Interpreting Studies. Abingdon: Routledge, 286–287.
The Journal of Specialised Translation Issue 34 – July 2020
116
• Nicodemus, Brenda and Karen Emmorey (2013). "Direction asymmetries in spoken
and signed language interpreting." Bilingualism: Language and Cognition 16(3), 624–
636.
• Nicodemus, Brenda and Karen Emmorey (2015). "Directionality in ASL-English
interpreting: Accuracy and articulation quality in L1 and L2." Interpreting 17(2), 145–
166.
• Orlando, Marc (2010). "Digital pen technology and consecutive interpreting: Another
dimension in note-taking training and assessment." The Interpreters’ Newsletter 15,
71–86.
• Orlando, Marc (2014). "A study on the amenability of digital pen technology in a hybrid
mode of interpreting: Consec-simul with notes." Translation and Interpreting 6(2), 39–
54.
• Padilla, Presentación (2005). "Cognitive implications of the English-Spanish direction
for the quality and the training of simultaneous interpreting." Rita Godijns and Michaël
Hinderdael (eds) (2005). Directionality in Interpreting: The ‘Retour’ or the Native?
Ghent: Communication and Cognition, 47–62.
• Pavlović, Nataša (2007). "Directionality in translation and interpreting practice:
Report on a questionnaire survey in Croatia." Forum 5(2), 79–99.
• Quené, Hugo et al. (2010). "Praat Script Syllable Nuclei." (Version 2).
https://sites.google.com/site/speechrate/Home/praat-script-syllable-nuclei-v2
(consulted 25.05.2019).
• Rinne, Juha O. et al. (2000). "The translating brain: Cerebral activation patterns
during simultaneous interpreting." Neuroscience Letters 294(2), 85–88.
• Seleskovitch, Danica (1999). "The teaching of conference interpretation in the course
of the last 50 years." Interpreting 4(1), 55–66.
• Seleskovitch, Danica and Marianne Lederer (1989). Pédagogie raisonnée de
l’interprétation. Paris: Didier.
• Tommola, Jorma and Marketta Helevä (1998). "Language direction and source text
complexity: Effects on trainee performance in simultaneous interpreting." Lynne Bowker
et al. (eds) (1998). Unity in Diversity? Current Trends in Translation Studies. Manchester:
St. Jerome, 177–186.
• Van Dijk, Rick et al. (2011). "Directionality effects in simultaneous language
interpreting: The case of sign language interpreters in the Netherlands." American
Annals of the Deaf 156(1), 47–55.
• Wang, Jihong and Jemina Napier (2015). "Directionality in signed language
interpreting." Meta 60(3), 518–541.
• Wu, Yinyin and Posen Liao (2018). "Re-conceptualising interpreting strategies for
teaching interpretation into a B language." The Interpreter and Translator Trainer 12(2),
188–206.
The Journal of Specialised Translation Issue 34 – July 2020
117
Biography
Sijia Chen is an Assistant Professor at Southwest University (China) and
an Honorary Research Fellow at Macquarie University. She is interested in
cross-disciplinary research on cognitive processes in translation and
interpreting, using a combination of methods such as pen recording, eye
tracking, psychometrics, and think-aloud protocols.
Email: sijia.rachel.chen@gmail.com
Notes
1 This study uses the L1 and L2 labels following Chmiel (2016) since the nomenclature is
sufficiently neutral for this interdisciplinary research.
2 A proposition consists of the main verb and all its arguments, whereas the idea density
is calculated as the number of propositions divided by the number of words (Brown et al.
2008).
