Chapter One of Sign Language Interpreting continued...
banner 120 words per minute. The role of lag time in the segmentation of incoming text is addressed by Goldman-Eisler (1972), who finds that frequently lag time consists of syntactic units (such as adverbial expressions). In this study, Goldman-Eisler compares interpreters' segmentations within target output with the original speakers' segmentations in the source message. She finds that very few of the interpreters' chunks match the original segmentation (identity), and that almost half the time interpreters began to translate before a chunk in the source text had been completed (fission). Just over a third of the interpreters' segments involved the linking of two or more chunks from within the source message (fusion). Thus, studies of segmentation and chunking indicate that interpreters influence the structure of the target text.

Research indicates that pauses often serve as unit breaks for interpreters in the attempt to chunk incoming information (Barik 1969; Gerver 1971). Kade and Cartellieri (1971) suggest that interpreters use pauses and redundancies in the original presentation as a time to catch up with the presenter, and Barik (1973) finds that, in practice, interpreters do so. In a study of English-ASL interpretation, Cokely (1992) finds that 87 percent of pauses are used for this purpose. Pauses have not only been viewed as unit markers, however. Goldman-Eisler (1967, 1968) suggests that within utterances, interpreters use pauses for planning upcoming productions.

Several studies have addressed the ways in which interpreters handle information overload. Interpreters face potential overload problems as a result of the physical and mental demands of interpreting (Brasel 1976). Studies indicate that interpreters do have "adjustment procedures" (Chernov 1969) to assist in such instances of overload. For example, Miller (1964) examines the strategies used by interpreters faced with continuous visual and auditory stimulation. Interpretations include omissions, interruptions of the input, errors, delayings (queueing), systematic omissions (filtering), and reduction in preciseness of output (approximation). Similar categories are identified by Gerver (1969) and Barik (1973). Gerver finds that differences between source and target texts consist of omissions of words, phrases, and longer stretches of text, as well as substitutions of words and phrases. He also finds that target messages include corrections of words and phrases. Barik also identifies specific types of omissions, additions, and substitutions, such as comprehension and delay omissions.

While these studies are experimental in design, Cokely (1982, 1992) has identified similar categories in analyses of interpreters in interaction. In an experimentally designed study of interactive interpreting, Cokely (1982) analyzes the performance of two ASL-English interpreters interpreting medical interviews between a nurse and patient. He identifies four categories of miscues: perception errors, memory errors, semantic errors, and performance errors. In a larger study of ASL-English conference interpreting, Cokely (1992) identifies a taxonomy of interpreter miscues that include not only omissions, additions, and substitutions, but also intrusions and anomalies. Whether experimentally designed or based on natural interaction, studies indicate that information overload influences an interpreter's renditions.

Previous Page
Back to information about this book
Next page
Catalog Submissions Permissions Ordering Home