Research background

Chapter 1 Introduction

1.1 Research background

The main purpose of interpreting service is to facilitate communication between two parties that speak different languages. First introduced during the post-World War Two period, simultaneous interpretation has since played a vital role in cross-cultural communication, but interpreting has been here for thousands of years. Through various forms of interpretation, including consecutive interpreting, simultaneous interpreting, escort interpreting and sight translation, interpreters have been critical in various different situations that range from translating speeches given by a president to showing foreign guests around the interpreter’s home country. The act of interpretation is likened to building a bridge, connecting two sides that were originally isolated due to language barrier. Through this bridge, people on the two sides will be able to understand each other, which forms the basis for further discussion and communication.

In order to successfully facilitate understanding between the two participating parties, fluency in interpretation is very important as disfluency would impair the listener’s comprehension of the content. And pausing is one of the factors that influence the fluency of interpreting output. From the very beginning of interpretation training, trainee interpreters are instructed to control to the number and length of pauses in their interpreted output (楊承淑, 2000). From the author’s experience as an interpreting student, trainees are told to avoid uttering fillers like um and ah, which fall in the category of “filled pause.” In performing sight translation tasks, interpreters are even advised to

2

read and comprehend the source text continuously as they interpret so as not to leave awkward long pauses in the middle of the interpretation (Weber, 1990).

However, there seems to lack a systematic way of instructing how pauses in interpretation output should be controlled or avoided. During the author’s years as an interpreting student, trainee interpreters were sometimes instructed to avoid or shorten pauses, but sometimes also advised to add or extend pauses in their interpretations. These comments are usually made on a case-by-case basis, but a general rule for pausing in interpreting, such as where should pauses be avoided / inserted and how long the pause should be, does not seem to exist. Exactly how should pauses be managed in interpretation output? To find the answer for this question, another question needs to be answered first: what is the interpreter engaged in cognitively during pauses in interpreting? Or in broader terms, what is a person’s cognitive process during pause onset in speaking?

Pause, the silent gap between words in the flow of speech, has existed alongside speech since the very beginning of human discourse. However, the significance of pause in discourse analyses was not acknowledged until much later. While the academia focused on researching the hearable tangible human speech, pauses were simply treated as “meaningless or as hindrances to good communication” (Fox Tree, 2002). This was perhaps partly due to the lack of technology to accurately select and calculate pauses before the 1950s (Rochester, 1973). In the early days of pause studies, pauses were discerned from the rest of the text by human ear, which was not exactly accurate. Thus it was much easier to focus on researching the tangible oral output data instead.

As recording technology developed, scholars were finally able to record

3

actual speeches for analysis. With the help of automatic recording devices, they could also discern pauses as short as 1ms that were indiscernible to the human ears (Boomer & Dittman, 1962). As researches on pauses gain focus in the 1950s, scholars began to acknowledge that pauses do present specific meanings in human speech, and their functions can be exploited to indicate the discourse planning process.

Studies on pauses in spontaneous speech have mostly been global, quantitative analyses. Empirical data has been able to inform scholars where pauses are most likely to occur in spontaneous speech, and what incidences and conditions are likely to increase the onset of pauses. Researches on pauses in interpretation output also began alongside studies of pauses in spontaneous speech. In fact, early pause researchers such as Goldman-Eisler conducted studies both on pauses in speech and interpretation. The conditions that were found to incur pauses in speech also seem to increase the possibility of pause occurrence in interpreting. Aside from direct researches on pauses, pause is also used as a sub-parameter of fluency in interpretation quality analyses.

Though abundant empirical results from pause studies have been accumulated, there have been few qualitative studies that focus on investigating the speaker’s cognitive reactions right at the moment of specific pauses in the flow of speech. This is likely because analyzing only the pause itself and surrounding oral output, whether in spontaneous speech or interpretation, provides limited information that infers the speaker / interpreter’s cognitive process at the moment of pause onset. Another set of data, preferably one that has been used to indicate cognitive behavior, needs to be incorporated for triangulation. In recent years, psycholinguistic studies have yield results that established links between neurological activities and

4

different human cognitive processes. However, utilizing cognitive psychology tools in pause studies can be difficult. Most of these tools, such as EEG or MEG, collect electoral activities or magnetic fields created by the brain. Thus the prerequisite for experiments using these tools is usually total silence of the participant, as sound waves will interfere with the data collection. These devices have been used in single-word utterance / interpretation experiments, but for the analyses of pause, the speaker / interpreter needs to be able to speek freely for a certain period of time for natural pause data to be collected.

So far, fMRI, which detects blood flow changes in the brain, has been used in spontaneous speech pause studies (Kircher, Brammer, Levelt, Bartels,

& McGuire, 2004), as speaking does not interfere with the data collection process. Another viable device that can be used is the eye tracker, which records the eye movement data of the participant, since sound waves also do not jeopardize the recording process. However, in order to utilize the eye tracker in pause research, there must be a form of visual material relevant to the oral output for the participant to look at during the experiment, otherwise eye movement data could not be obtained and compared with pause data. This prerequisite is difficult to fulfill for spontaneous speech task, which usually involves only the speaker talking, and most interpretation tasks, which concerns listening and speaking, but not reading.

Fortunately, a form of interpretation does involve reading: sight translation. Sight translation is performed when the interpreter reads a written text in one language, then orally interprets the written text into another language. Though the source of input is different from simultaneous and consecutive interpretation, sight translation nonetheless shares similar general interpretation process and demand on efforts with other forms of

5

interpretation. Therefore, a close study of pause occurrence in sight translation output, triangulated with the interpreter’s eye movement data during the course of interpretation, should be able to bring certain insights into the interpreter’s ongoing cognitive process during pauses in interpretation. The results of this study would enable a systematic way of instructing pausing patterns in output to interpreting students based on the cognitive functions of pauses. The findings may also bring the possibility of utilizing pauses and their cognitive functions to analyze the interpreter’s cognitive processes during different stages of interpretation.