Examination of juncture pauses in the experienced interpreters’

eye-movement protocol

Examination of the protocol showed that most of the time, the experienced interpreters’ fixations stayed between the middle and the end of the segment they had finished translating. This is in line with the two location analyses, which showed that during 82 (71.3%) juncture pauses, the interpreters’ fixations stayed in the

just-translated segment, rather than moving to the next segment.

As explained in Chapter 3, the data of this current study was taken from the sight translation tasks of three Chinese paragraphs by experienced interpreters (Chen, 2013).

One example of each paragraph is given here. In the examples, the ROI for Parameter (a) “fixation’s location relative to output interruption” was defined as the segment of the sentence that was just concluded when the juncture pause occurred, and is marked in a box in the protocol examples. The ROI for Parameter (b) “fixation’s location relative to output resumption” was the first word or the first content word of the next sentence to be interpreted, and is marked with grey highlighting in the examples. The first content word, in addition to being highlighted in grey, is also underlined.

96

EX. 4-5

Source Text: 可是在旅館裡，好幾百個員工，每一個人都像是旅館的產品。

如果有哪一個員工對客人不友善，就會讓客人感到不開心。

Annotated Protocol: However, in a hotel, there are hundreds of employees, when every one of them is like the products of the hotel. ^[7: 28075-28479: 403ms;像是

→旅館] If any employee is not friendly to the customers,^[8: 31701-32057:

306ms;不→客人] the customers will not feel happy.

Eye Movement Protocol 1:

Ex 4-5 demonstrates the reading process during two consecutive pauses in one of the 18 experienced interpreter’s sight translation task. For the sake of clarity, the eye movement protocol is presented in two parts: Eye Movement Protocol 1 and 2.

97

In Ex 4-5, both Pause 7 and 8 were juncture pause, each occurring at the end of a syntactic segment. In the case of Pause 7, the interpreter had finished translating the segment “每一個人都像是旅館的產品。” (when every one of them is like the products of the hotel), but during the pause, instead of moving to the next segment

“如果有哪一個員工對客人不友善，”(If any employee is not friendly to the customers), the eyes stayed in the translated segment to fixate on the word “像是”

(Fixation 102) and “旅館” (Fixation 103). (See Eye Movement Protocol 1.) The eye only moved to the next segment when the oral output resumed. Being in the row below, the first word (如果) or content word (哪) of the next segment was not in the interpreter’s perceptual span during the pause. None of the words in the next-to-be translated segment had been previewed before the actual oral translation of the segment. However, as can be seen in the protocol, the first word “如果” was within the perceptual span during the previous two fixations (Fixation 96 and 100) on the word “產品”, because “如果” was less than three characters away from the fixated word “產品”.

Similarly, during Pause 8, the interpreter had finished translating the segment “如 果有哪一個員工對客人不友善” (If any employee is not friendly to the customer), but instead of reading the next-to-be translated segment, “就會讓客人感到不開心”, the eyes fixated on “不” and then regressed to “友善” in the just-translated segment (See Eye Movement Protocol 2). None of the words in the next-to-be translated segment had been previewed before the onset of the oral output.

Both pauses started with a forward saccade, and demonstrated how the eyes regressed at the end of the oral translation preceding a juncture pause, and moved forward at the onset of the pause. At the offset of the pause and the resumption of oral output, the eyes moved forward, indicating that the interpreters resumed to their normal reading pattern.

98

EX.4-6

Source Text: 事實上，即使沒有生病，也未必代表健康。

Annotated Protocol: In fact, ^[7:13704-14852:1147ms;健康→身體→健康→健 康] even if you’re not sick, it does not guarantee health.

Eye movement protocol:

Ex. 4-6 shows a similar pattern. Pause 7 occurred at the end of the segment “事實 上”, which meant it was a juncture pause. Interestingly, when translating the “事實 上”, the interpreter’s eyes were still in an earlier segment ”沒有病痛就代表身體很健 康” and stayed near the end of this segment during Pause 7 rather than moving to the next-to-be translated segment “即使沒有生病”. During the time, neither the first word “即使” or the first content word “沒有” was within the perceptual span. When the oral output resumed, the eyes first moved to the segment “事實上” before reaching the first word “即使” of the segment being translated. In line with the analysis results in the previous section, none of the words in the new segment “即使 沒有生病” had been previewed.

Ex. 4-6 is another example of how an experienced interpreters’ eyes regressed when approaching the end of the oral output and moved forward at the beginning of the following juncture pause. Again, the pauses ended and the oral production resumed with a forward saccade.

99

EX. 4-7

Source Text: 不過投資之前要了解自己的性格，有些人個性比較積極，有些 人則個性保守，不同的人適合不……。

Annotated Protocol: But before you invest, you must understand your own personality. ^[3:8662-8967:305ms;的] Some people are more aggressive by nature while others are more conservative.^[4:13280-13750:470ms;個性→人→個性]

Like Ex 4-5, Ex 4-7 also presents the reading process during two consecutive pauses. For the sake of clarity, the eye movement protocol is presented in two parts:

Eye Movement Protocol 1 and 2.

Finishing translating the segment ”不過投資之前要了解自己的性格” (But before you invest, you must understand your own personality), the interpreter paused (Pause 3) with the eyes fixating on the word “的” in the just-translated segment. The next-to-be translated was in the row below, so neither the first word “有” nor the first

100

content word “些” was within the interpreter’s perceptual span. When the oral output resumed, the interpreter first fixated on the last word of the previously-translated segment “性格” before moving to the next row to fixated the segment being translated.

None of the words in the new segment had been previewed.

Something similar happened to Pause 4. Moreover, when translating the preceding segment “有些人個性保守”, the interpreters mostly fixated on words in an even earlier segment “有些人個性比較積極”, and didn’t move to “有些人個性保守” until its translation was finished with Pause 4 (accompanied by Fixation 38, 39, 40). The interpreter’s eyes stayed in the just-translated segment during the pauses. No word in the next-to-be translated segment “不同的人適合…” was read during the pause or had been read before the pause. When the oral output resumed, the interpreter first fixated the last word of the segment, and then the period, before reaching the segment being translated.

Like in the previous two examples, the eyes regressed near the end of the oral output preceding the two pauses, and continued with a forward saccade at the onset of the pause. Both pauses ended with a forward saccade, suggesting that the eyes have resumed to normal reading.

During 82 (71.30%) of the juncture pause, the experienced interpreters fixated on the segment they just translated. 54 of the pauses started with a forward saccade (65.85%). Among the 54 pause, 31 (57.41%) immediately followed a regression at the end of the output. That is to say, there are 31 cases in like the three examples given above: regression was found not at the onset of the pause, but at the end of the

translation before the pause. Then the pause started with a forward saccade, but all the fixations during the pause were still in the just-translated segment N, rather than in the next-to-be translated segment N+1.

101

Previous researchers suggested that experienced interpreters’ semantic-based interpreting is always linked to the context, which allowed for more coherent output (Liu, 2009; Moser-Mercer et al. 2000). They see the over-all picture to make

successful predictions (Liu, 2009), moving from known to unknown information

(Moser-Mercer et al. 2000). Their suggestions may serve as a possible explanation for

“regressing-back” near the end of the oral translation the continuous reading of the already-translated segment by the experienced interpreters in this current study. When wrapping up the segment N, instead of reading forward, the experienced interpreters may have regressed to an earlier part of N to see if there’s anything to be added or fine-tuned based on the context. During the following juncture pause, they could have been double-checking their output against the original text. However, to better

understand this phenomenon will need further examination.

Without previewing the segment N+1 during the pause, the experienced

interpreters ended the pause to read and orally output N+1. Reading comprehension, reformulation (Chen, 2013) and possibly oral output took place at the same time.

Moreover, not previewing N+1 during the preceding juncture pause distinguishes the process of the experienced interpreters’ interpreting from that of spontaneous speech planning (Goldman-Eisler, 1958; Rochester, 1973) and from the novices’ (Su, 2013).

This seems to suggest that the experienced interpreters may have developed certain skills or strategies which allowed them to do so without hampering the quality of the immediate translation of N+1. Possible strategies will be proposed and discussed in Chapter 5.