CHAPTER FOUR RESEARCH FINDINGS
After the Sight Translation (henceforth ST) task, the subjects’ recordings are
analyzed. The research findings are discussed in the following three dimensions:
first, fidelity, which covers miss- interpretation, omissions, and unnatural Chinese;
second, ST chunking, which covers chunk- moving and chunk-connection; third,
delivery, which covers backtracks, fillers, pauses, and time.
The general outcome is as follows: Group A, which consists of GITI students,
does the best in the first dimension, fidelity. It has the fewest units of
miss- interpretation, omissions, and units of unnatural Chinese. Group B, which
consists of graduate students from English department, ranks the second and Group C,
which consists of government employees with the same level of English proficiency,
is the third in the three subdivisions.
In the second dimension, ST chunking, Group A has the fewest chunks moving
too far, and the highest number of successful connections. Group B has the most
chunks moving too far and the second highest number of successful connections.
Group C has the second fewest chunks moving too far, and the fewest successful
connections. Again, Group A performs the best of the three groups. As successful
chunk-connection can create more sensible rendition and thus should be more
important than chunk- moving, Group B appears to perform better in terms of ST
chunking than Group C.
Finally, in delivery, Group A has the fewest backtracks, fillers, pauses, and
spent the least time. Group B has the most backtracks and fillers, the second most
pauses, and spent more time than Group A. Group C has the second most backtracks
and fillers, the most pauses and spent the longest time. Still, Group A outperforms
the other two Groups in this dimension, while the results for Group B and Group C
are mixed; Group B does better in pauses and time control while Group C does better
in backtracks and fillers.
4.1 Comparisons of Fidelity
Fidelity being the most important element in ST, is observed first. According to
Barik (121), in SI, the interpreter’s version may depart from the original version in
three different ways: omissions, additions, and errors. Through the pilot study of
this case study, the same phenomena are also observed. Based on the notion and the
definition of fidelity described in the previous chapter, in this section, the researcher
intends to observe the following major phenomena that are often related to fidelity.
They are: miss- interpretation (errors), omissions, plus unnatural Chinese (or so-called
“translationese”), which might result from a shallow or semi understanding of the text.
Additions, due to their scarcity in this case study, are not discussed. Although there
could be various reasons for miss- interpretation, omissions, and unnatural Chinese as
well, such as insufficient comprehension, poor coordination, or carelessness, etc., they
are all categorized here in the fidelity dimension for the convenience of the
observation.
4.1.1 The Segmentation of meaning units
There are 12 sentences in the text, numbered from 1 to 12. Each sentence is
divided into several meaning units. There could be different ways of segmenting
meaning units, depending on the purpose of segmentation. Therefore, the meaning
units here are segmented according to their ST purposes instead of linguistic ones.
That is, they are segmented into small chunks for ST based also on the target language
instead of merely morphemes, words, or phrases.
For example, the first sentence of the speech text can be divided into 9 units--(1)
This dialogue (2) is designed (3) to help us (4) cooperate (5) where our
interests coincide, (6) and (7) to narrow (8) differences (9) wherever they
exist. Linguistically, the segmentation is not very consistent because there are
meaning units of words, phrases, and even clauses. However, if the possible
renditions in Chinese are also considered, the segmentation would make more sense.
For example, (4) cooperate (word) and (5) where our interests coincide (clause) can
both be considered one single meaning unit, or a “chunk,” in the target language. (9)
wherever they exist (clause) is usually interpreted or omitted as a unit in Chinese as
well. The details of the segmentation can be found in Appendix C.
4.1.2 Comparisons of Miss-interpretation among the Groups
For the three groups, the miss- interpretations usually appear when there are
unfamiliar words or complicated sentence structures to them. For example,
“coincide” translated into 「衝突」in Chinese (which means “collide”), or sentence 8
“I could not fairly represent them if I did not emphasize America’s belief that
organized and peaceful political expression is not a crime or a threat;” translated into
「我能夠代表人民，是因為我知道而且我也非常強調美國人民的理念是，不是一
種罪惡或者是一種威脅的方式，那也不是一種有組織的那種，也不是一種罪過，
也不是一種威脅方式」in Chinese (which can be roughly translated back into “The
reason why I can represent people is that I know and I put a lot of emphasis on the
American people’s ideal that is not a crime or a way of threat, not the organized type,
not a crime, and not a way of threat, either.” and after checking the original sentence,
we can see that the meaning has been somewhat distorted) The total numbers of the
miss- interpreted units made by each group are shown below in the bar graph. For
details of the miss- interpreted units made by the subjects of each group, please refer
to Appendix D.
0 50 100 150 200 250
Group A Group B Group C
Group A Group B Group C
Figure 4.1 The comparison of miss-interpreted units among the groups
From the figure above, it is obvious that the numbers of miss- interpreted units
are similar between Group A and Group B, which possess similarly high language
proficiency, while Group C, the relatively lowest in terms of language proficiency
among the three, has the most miss-interpretation units.
That is, people with high language proficiency and professional training have the
lowest units of miss-interpretation; those with similarly high language proficiency but
without any professional training have the second lowest, and those with lower
language proficiency and introductory ST training have more miss-interpreted units.
Note that even though Group B hasn’t received any ST training, its
miss- interpretation units are still quite few, only a bit more than Group A. However,
Group C, which has received some ST training, still has far more miss- interpretation
units than the other two groups, which proves that its low language proficiency has
more influence than its ST training over the ST task in this subdivision.
Therefore, it could be inferred that language proficiency comes before
professional training, or the former should be the groundwork of the latter for higher
fidelity and thus a better ST performance. Also, it could mean that Group C hasn’t
got sufficient language proficiency for this speech text, which originally served as a
teaching material for T & I students. However, results of the three groups are
consistent with expectations, which proves the text to be appropriate and valid in
testing the subjects.
4.1.3 Comparisons of Omissions among the Groups
In this case study, omissions most often occur when complicated sentence
structure or condensed information appears because the effort required for interpreting
it could exceed the effort available according to Gile’s effort model (Basic Concepts
170-72, 183). For example, when translating sentence 8, which is the most complex
in the text because it includes 2 embedded sentences or 13 chunks (meaning units),
the subjects encounter more trouble reformulating it and words like “organized,”
“crime” or “threat” are usually missing. And, in sentence 3, where many things are
mentioned, words like “proliferation,” “stability,” and “the resolution” were omitted
or substituted by many subjects. (Please refer to Appendix E)
Please note that the phenomenon called “substitution” is also calculated here
because it’s different from miss- interpretation and is usually a kind of “dilution” of
the original, and sometimes it would cause slight changes to the original as well.
Examples are “stability on the Korean Peninsula” substituted by 「韓國的問題」
(which means “the Korean issues”), “profoundly distressed” substituted by 「非常失
望」 (which means “very disappointed”), etc. The numbers of omissions, including
substitutions, made by each group are shown below in the bar graph. For more
details of omissions made by the subjects of each group, please refer to Appendix E.
Figure 4.2 The comparison of omissions among the groups
Interestingly enough, the result of this part is the reverse of the previous one.
That is, Group A has the most omissions, Group B the second most, and Group C the
fewest of the three. Since it is not possible to tell the exact reason behind each
omission and substitution when there are no interviews with the subjects after the case
study, all possibilities should be considered. That is, they could be carelessness,
insufficient efforts available, or ST skills such as avoidance of unfamiliar words and
choice of a terser rendition. The results of the three groups indicate that Group A
0 20 40 60 80 100 120
Group A
Group B
Group C
Group A Group B Group C
departs from the original text the most, whatever the reasons maybe, Group B the
second most, and Group C sticks to the original text the most.
4.1.4 Comparisons of Unnatural Chinese (Translationese) among Groups
Unnatural Chinese, or translationese (unnatural rendition of the target language),
happens when the interpreter doesn’t fully understand certain words or is trapped by
some complex sentence structure. For instance, “as your leaders do with us”
translated into 「就像你們的領導對我們做的一樣」(word for word translation), “I
could not fairly represent them” translated into 「我不能十分的代表他們」 (strange
collocation),etc. For the detailed calculation of the translationese made by the
subjects, please refer to Appendix F.
0 20 40 60 80 100 120 140
Group A
Group B
Group C
Group A Group B Group C
Figure 4.3 The comparison of translationese units among the groups
This pattern resembles that of miss-interpretation--group A has the least number
of translationese units, group B has more, and group C has the biggest number.
Meanwhile, the gaps among groups are bigger than those of miss- interpretation
(especially between Group A and Group B) and omissions, which manifests that
language proficiency and professional training both can make greater difference on
the occurrence of translationese than miss- interpretation and omissions. The bigger
gap between Group A and Group B indicates a stronger influence of professional ST
training on this part. In ST training (and other interpretation training as well),
understandable delivery and natural language use are stressed. Therefore, it is not a
surprise that group A subjects know better how to avoid translationese while Group B
subjects seem to be more submitted to literal translation.
4.1.5 Implications for ST Training
Group C has the highest numbers of miss- interpretation and translationese units,
which manifests that language proficiency should be fundamental to fidelity. At the
same time, its smallest number of omissions could imply either Group C is more loyal
to the original text or is not flexible enough to make changes. For Group A and
Group B, when they possess similar language proficiency, it seems that the impact of
professional training is to reduce miss-interpretation and translationese but to increase
omissions.
Therefore, the implications for training are: first, the screening of the trainees is
crucial. The language proficiency of the trainees should be sufficient so that they
can perform well in terms of fidelity. Second, although omissions could be the
coping strategies in ST, they should be used with great discretion or they might dilute,
exaggerate, or even distort the meaning of the original. The priority in ST should be
fidelity.
4.2 Comparisons of ST Chunking
In SI, the interpreter has to follow the speech very closely to stay in the
“simultaneous mode,” and that’s when chunking, breaking source language sentences
into separate chunks and using appropriate connectives to organize the chunks into the
target language, comes into play. (Chang 張維為 41) Since ST serves as an
important assisting skill in SI with text, it is crucial that chunking is used in ST.
In this case study, chunking is observed next to fidelity; however, it is not easy to
“quantify” chunking in a standardized way because there are too many possibilities
for different types of chunking, just like one sentence could have many different
renditions. Therefore, the researcher defines that all possible arrangements of
chunking would be acceptable as long as the shifts are within the range of half a
sentence, which contains about 7 to 9 words. The range is determined based on a
statement about “the magic number seven” that “short-term memory can hold only
seven plus or minus two items.” (Miller 81-97) The words that the subjects can deal
with at one time should be around the same number. For a sentence shorter than 10
words, chunking will not be calculated. For a sentence longer than 20 words, the
relative clause or the part after a semicolon, can be considered as an appropriate
chunk, and the sentence can thus be divided into 3 or 4 chunks. Another rationale
behind this is the concept that for a sentence of medium length, the interpreter should
wait for no more than half a sentence before interpreting (Chang 張維為 57)
4.2.1 The Skills of Chunking
The purpose of chunking in SI is to narrow the time difference between the
speech and the interpretation, and to reduce memory load. Although the two needs
don’t exist in ST, chunking skills are still needed in ST for it is so closely related to SI,
as a warm-up or when doing SI with text. Therefore, whether chunking skills are
applied and how they are applied in ST can help us tell the difference of ST training.
The difficulties of chunking in English to Chinese ST consist in the very
different structures of the two languages. As mentioned previously, English has
basically right-branching sentence structures (the “tree diagram” develops from left to
right) while Chinese has left-branching structures (usually VPs and NPs). (Setton, A
Pragmatic Theory 143) Therefore, if one wants to convert English sentences into
Chinese, usually the best practice is to jump out of the original structures and focus on
finding equivalences in meaning. However, if one has to also follow the original
sentence structures and orders (as is the case in ST), s/he has to segment the original
sentence into proper chunks, make necessary movements, and then use connectives to
link the chunks in the target language for an appropriate interpretation. Therefore,
the second dimension, ST chinking, is subdivided into chunk-moving and
chunk-connection.
4.2.2 Comparisons of Chunk-moving among Groups
For sentences under 30 words and without complex patterns (e.g. double
negative, embedded clauses, etc.), the subjects basically follow the original structures.
The sentences that could have chunk- moving problems are number 8, 10, 11. The
chunks of these sentences that are moved more than half a sentence away in the target
language are calculated below:
Table 4.1 Chunks of the complex sentences moved too far by the 3 groups Chunk
Group
In sentence 8 In sentence 10 In sentence 11 Total
A 2 6 1 9
B 6 9 1 16
C 3 2 0 5
Source: compiled by this study
From the table above, we can see that Group B moves more chunks than Group
A and Group C. Group B is the only group that has never received any ST training.
It is thus probable that the subjects don’t pay attention to chunking skills but
translated the sentences based solely on the structure of the target language, which is
very different from the source language. Something worth noting is that while
Group C follows the original structures most strictly, its sentence connections are
mostly problematic, which will be calculated in the next part. This is perhaps
because they were taught of the chunking principle when they did not possess enough
language flexibility to make their renditions right.
4.2.3 Comparisons of Chunk-connection among the Groups
This subsection concerns the connections between chunks (clauses, participles,
etc.) The connections could be made unsuccessfully, unnaturally, or simply missing.
For instance, sentence 2 “It has made possible a pair of historic summits in
Washington and Beijing.” was translated to 「這也會造成一個歷史上的會談，在華
盛頓和北京。」. The inappropriate connection (or no connection at all) makes the
translation sound incomplete. The first half of Sentence 3 “It has allowed us to work
together in areas of vital significance” was translated into 「也讓我們能夠一起工
作 ， 就 某 些 領 域 來 說 。 」 . Here, chunking was used unsuccessfully. The
unsuccessful connections indicate the existence of interferences mentioned by Gile
(Basic Concepts 184) in Chapter 2. These kinds of failure are calculated below:
Table 4.2 The unsuccessful connections made by the 3 groups
Group Connection
A B C
In sentence 1 0 0 1
In sentence 2 1 1 4
In sentence 3 1 0 7
In sentence 4 0 0 1
In sentence 5 1 4 9
In sentence 6 0 2 4
In sentence 8 3 7 8
In sentence 9 2 3 4
In sentence 10 1 2 8
In sentence 11 0 5 6
In sentence 12 1 1 3
Total 10 25 55
Source: compiled by this study
As expected, group A has the fewest unsuccessful connections, Group B the
second fewest, and Group C the most (about twice of Group B’s) among the three.
This implies that even the subjects have some knowledge of chunking (like Group C
subjects), they still can’t manage the skills without sufficient language proficiency.
In other words, language proficiency comes before ST training in terms of successful
subjects possess both good language proficiency and ST training.
4.2.4 Implications for ST Training
Compared with Group B, Group A moves fewer chunks and make fewer
unsuccessful connections. Meanwhile, Group C moves the fewest chunks among the
three, but creates more unsuccessful connections than the other two.
This indicates that, without sufficient language proficiency, even if students are
introduced of the chunking principles and skills (like Group C subjects), what they
can actually do is just follow the original sentence structures but not create natural
renditions. On the other hand, people with sufficient language proficiency but
without professional training (like Group B subjects) don’t quite follow the rules of
chunking and are more easily caught in the complex sentences than those who possess
both sufficient language proficiency and ST training (in this case, Group A subjects).
Therefore, it might be inferred that chunk-moving is something that can be
taught and a notion that can be acquired in a short period of time, even by people with
intermediate language proficiency (such as Group C subjects); however, successful
sentence connection, which is more important in terms of communication, takes better
language proficiency and longer training to manage. Once again, language
proficiency should thus be a prerequisite in this dimension: ST chunking.
4.3 Comparisons of Delivery
Delivery is essential in all types of interpretation, because interpretation is
basically a kind of “performance,” which involves language proficiency, knowledge
of the two languages, peripheral knowledge, interpretation skills, etc (Yang 165). In
this section, backtracks, fillers, pauses, and average time spent that concerns Delivery
are observed.
4.3.1 Backtracks, fillers, pauses, unnatural renditions, and time
Backtracks mean unnecessary repetitions (ranging from a word to a sentence)
by the subjects, especially when they’re having trouble organizing sentences. Here,
backtracks are calculated regardless of their linguistic differences. That is, a
backtrack could be a word, a phrase, or even a whole sentence.
Fillers mean words or phrases like 「嗯」、「那」、「然後」, etc., which the subjects
use either to buy time or to ease tension. Also, fillers are calculated despite of their
different types, as long as their appearance is not necessary in the interpretation.
Pauses mean unnecessary and unnatural stops in the subject’s delivery. In 2.2.3,
Déjean Le Féal claims that there are two types of pauses: syntactic pauses and
hesitation pauses, and Messina argues that there may be non-syntactic pauses not
necessarily caused by hesitation in reading. However, in this case study, the most
possible reason for pauses should be hesitation for a proper rendition. According to
the average delivery of the subjects in this case study, a syntactic pause usually lasts
for one second or so, while a hesitation pause can last from 2 to 10 seconds.
In this case study, any pause 3 seconds and above is calculated. This is a more
relaxed standard based on the requirement in the evaluation form in Yang’s book,
pp.163, where any pause over 2 seconds is calculated. The use of 3 instead of 2
seconds is because the form is all-purpose; that is, it can be used to evaluate SI, CI, or
ST, and in ST, time pressure is not as serious. Besides, in this study, 2 seconds is a
gray area, especially for slow speakers. Therefore, the 3 second-and-above standard
is applied.
Finally, average time of the subject’s delivery is also calculated. Because
pauses are not further divided into different time ranges, the subjects that pause the
same number of times could have very different delivery time, which has a lot to do
with fluency, or the delivery flow.
4.3.2 Comparisons of Backtracks among the Groups
A very important principle of fluency is to avoid backtracking, which not only
fragments the delivery but also is a waste of time. In this subsection, the numbers of
backtracks generated by the three groups are shown in the figure below. For a more
detailed calculation, please refer to Appendix G.
0 50 100 150 200 250 300
Group A Group B Group C
Figure 4.4 Numbers of backtracks made by the three groups
It can be observed that Group B backtracks the most. Group C ranks the
second, and Group A the third. This happens to be in inverse proportion to their
training time (Group A the longest time, Group C the second, and Group B the
third — no time at all). Therefore, it can be inferred that the more ST training one
receives, the fewer backtracks s/he makes.
4.3.3 Comparisons of Fillers among the Groups
Fillers, which instructors of interpretation are also always preaching students to
avoid, make one’s delivery sound sloppy and shaky. In this subsection, the numbers
of fillers generated by the three groups are shown in the figure below. For a more
detailed calculation, please refer to Appendix H.
0 50 100 150 200 250 300 350 400
Group A Group B Group C
Figure 4.5 Numbers of Fillers used by the three groups
Once again, the rank is the same as that of backtracks. Group B uses the most
fillers, Group C the second, and Group A the third. However, the gap between
Group C and Group A is very small. Accordingly, their usage of fillers is also in
inverse proportion to their training time.
4.3.4 Comparisons of Pauses among Groups
Although both ST and SI should be aware of time pressure, in pure ST, time
pressure can be taken out of the equation. Therefore, whe n processing the speech
text or encountering trouble, subjects tend to allow themselves to stop and think — the
difference is “how long.” To those with ST training, the time is supposed to be
shorter for their awareness. In this subsection, the numbers of pauses made by the
three groups are calculated respectively. As for their lengths, because there are too
many combinations and possibilities, the calculation of average time spent in the next
subsection will be a generalization of this respect. That is, in this subsection, only
the times of pauses are calculated, not the lengths. The following is the figure of the
numbers of pauses made by the three groups. For a detailed calculation, please go to
Appendix I.
0 50 100 150 200 250
Group A Group B Group C
Figure 4.6 Numbers of pauses made by the three groups
Group C makes the most pauses. Group B the second, and Group A the third.
It was expected that subjects with ST training would have fewer pauses. However,
the results show that even with ST training, the subjects could still make more
pauses than those without training (Group C and Group B subjects). This
indicates that language proficiency also plays a role in this subdivision.
First of all, it might be safe to infer that subjects with higher language
proficiency will run into fewer obstacles when processing the speech text, and thus
less need to pause. In terms of language proficiency, Group A and Group B are
both very high, and thus should have similar numbers of pauses. However, it
turns out that Group A makes less pauses, and manifests the effect of ST training.
However, the fact that Group C has more pauses than Group B indicates the
importance of language proficiency over ST training in the respect of delivery.
4.3.5 Comparisons of Time Spe nt among the Groups
ST being an important skill in SI with text, should include time control in its
training, because the interpreter can’t afford lagging behind the speaker too much to
stay in the simultaneous mode. Therefore, the average time spent in delivery had
better not exceed the speech when interpreted into mother tongue. (Yang 163) In
this subsection, the average time spent on delivery by the three groups is shown in the
figure below. For a more detailed calculation, please refer to Appendix J.
Table 4.3 Average time spent by the three groups respectively
Group A B C
Average Time 3’38” 4’03” 5’00”
Source: compiled by this study
Group A uses the least time, Group B a little longer, and Group C the longest of the
three. This can also be referred back to the pause calculation, because they are in
direct proportion. (The more often they pause, the more time they spend.)
4.3.6 General findings
The 4 elements (backtracks, fillers, pauses, and average time spent) calculated
in this section say something interesting about the influences of language proficiency
and professional training on delivery. For instance, when two groups possess similar
language proficiency (Group A and Group B), training does help improve all 4
On the other hand, when two groups possess different language proficiency but
both go through ST training (Group A and Group C), the one with higher language
proficiency performs a little better in the usage of fillers, but much better in the rest of
the 3 elements.
As for Group B and Group C, one with higher language proficiency and the other
with ST training, their comparisons on delivery are of mixed results. Group C does
better in terms of backtracks and fillers, while Group B does better in terms of pauses
and time. This indicates that backtracks and fillers can be reduced through ST
training regardless of language proficiency, while pauses and time are mainly
influenced by language proficiency.
By the way, there are few subjects in Group B and Group C that would repeat the
original text during pauses (especially one Group B subject) like they were “thinking
aloud.” No subjects in Group A did this. This says something about ST training on
delivery. However, because the subjects doing this were so few that the researcher
didn’t calculate but only mentioned them here for future studies.
4.3.7 Implications for ST Training
From the findings above, it could be inferred that ST training has stronger effect
on avoiding backtracks and fillers for delivery. Also, this is so far the only section
that doesn’t show such strong dominance of language proficiency over ST training.
This manifests delivery doesn’t have so much to do with language proficiency as
fidelity and ST chunk ing. In the future, the ST instructor might want to put more
emphasis on avoiding pauses, time control, and helping the students to improve their
renditions by enhancing their language proficiency.
4.4 Discussion
This section describes the varied influences of language proficiency and ST
training, along with the possible impacts of different degrees of training on ST
performance.
4.4.1 The Influence of Language Proficiency
The comparisons between Group A and Group C can help us see the influence
of language proficiency. The two groups both possess certain (though different
levels of) knowledge of ST; however, the results in all sections show that Group A
obviously does a much better job. This then shows the influence of language
proficiency, because that’s the greatest difference between the two groups. With
much higher language proficiency, Group A is more capable of practicing the ST
skills they know and can be more flexible in interpreting. In other words, the higher
language proficiency enables Group A subjects to understand the speech text better,
so they could spare more efforts for performing ST skills.
4.4.2 The Influence of ST Training
Following the same rationale described above, the comparisons of Group A and
Group B can help us see the influence of ST Training. They both possess sufficient
language proficiency for the speech text, but the results tell us that Group A still does
a better job. This affirms the value of ST training.
For example, in fidelity, Group A has fewer units of miss- interpretation, more
omissions and substitutions, and much fewer units of translationese. In chunking,
Group A again outperformed Group B by moving fewer chunks too far or
unsuccessfully. In delivery, Group A made much fewer backtracks, used very much
fewer fillers, made fewer pauses, spent less average time. These all display the
effects of ST training, which is to enable students to communicate the message
faithfully, naturally, and in an orderly fashion. The students are taught to add or omit
information into the target language to make the renditions more comprehensible.
They are also taught to control the flow of their delivery. This study shows that the
ST training in GITI is effective.
4.4.3 Language Proficiency versus ST Training
According to the analyses mentioned earlier in this chapter, when language
proficiency and professional training are put together for comparison or choice in ST,
the former would be more fundamental and it turns best to be the foundation stone of
the latter. If we compare the results of Group B and Group C, we can find that
Group B does better in fidelity (far fewer units of miss-interpretation and
translationese, though more omissions, which is not necessarily a bad thing) and ST
chunking (more chunk- moving and fewer unsuccessful chunk-connections, but the
latter is more important than the former). As for delivery, Group B and Group C are
about even, because Group B does better in pause and time, while Group C does
better in backtracks and fillers.
Through the comparisons, it is obvious that language proficiency comes before
ST training. That is, if the ST training is to be effective, certain level of language
proficiency is required in the first place, and that is why Group A subjects do the best
job in this case study.
4.4.4 Mere Concepts versus Intensive Training
Although Group A and Group C both possess some ST knowledge through
lectures and drills, their trainings are actually of different levels. Group A, coming
from the Graduate Institute of Translation and Interpretation, receives more intensive
training, while Group C, coming from a high intermediate class that has translation
and interpretation lesson only once a week, receive less training in ST with easier text.
This is also mainly because they posses different levels of language proficiency.
At the time of the case study, most subjects in Group C only know the concepts
of ST but cannot really apply the skills at will. In contrast, subjects in Group A are
more familiar with ST and have greater ability to make good use of the skills needed.
These differences (ST training as well as language proficiency) create the results
mentioned of Group C and Group A. The former is observed to have applied the
skills of ST, only at very superficial level, so there is as much failure as is success, or,
more failure than success, as in fidelity. The latter, on the other hand, does more
successful ST performance. We can’t say that it is mere concepts and intensive
training that made all the difference for Group C and Group A, but, with different
language proficiencies, the different trainings certainly widened the gap of
performance.
4.4.5 Implications for future ST Training
From the case study, it can be inducted that language proficiency should always
be fundamental in ST training. Sufficient language proficiency plus ST skills create
the best synergies. Insufficient language proficiency hinders the application of ST
skills.
However, we should also be aware that the material used in this case study is for
training professional interpreters. Therefore, the “sufficient language proficiency”
mentioned in the case study refers to the sufficiency for this type of material. The
researcher therefore assumes that those who with lower language proficiency level
still can have their opportunity to receive ST training. They just need to be in a class
with different objectives and materials. If the class uses materials that fit their
language proficiency level, they can then focus on the absorption of basic ST skills.
Only that they have to know this will not equip them to be an interpreter due to their
language ability. Nevertheless, if they can work on enhancing language proficiency,
they may be able to take on greater challenges in the future. However, this is only an
assumption so far, more investigation on ST teaching is needed to prove it applicable.
4.5 Suggestions for Future Improvement
This is a one-time case study, without follow-ups, interviews, whatsoever.
Therefore, a lot more aspects that could have been explored further cannot be covered.
The following are some suggestions for future improvement:
4.5.1 The Difference within the Same Group
Even in the same group, the subjects are very different. For example, subjects
in Group A are supposed to have similar language proficiency and ST training, but
they don’t. This is because they range from year 1 in GITI to after graduation, and
thus are actually of different language proficiency and in different stages of training.
Other differences include majors in college, age, etc.
It is ideal to keep the same group totally congenial, but in this case study it is not
done due to the feature of scarcity of the subjects.
4.5.2 Their Notes, Slashes, and Psychology
In ST, what the interpreter marks on the text could say a lot about their mindset
and efforts. In the case study, the subjects are allowed to mark anything on their text
sheets and they are collected afterwards, some completely clean, some with arrows,
slashes, or even words. However, if the researcher intends to find out about the
meaning behind their notes, slashes, and psychology, more needs to be done, such as
interviews, cooperation with an expert in the field of psychology, etc.
