Abstract
In this paper, we will discuss simultaneous interpreters’ strategies for following the
source speech. In this research, we conducted a survey of the time delay, aka EVS
(ear-voice span), in the translations of simultaneous interpreters, and analyzed the
rollback as well as the timing of the start of the interpreter’s translation. We defined the
“Interpretation unit” as the summarization of a certain meaning in the context of the
source speech, and calculated the timing based on the number of interpretation units
(hereinafter the “delayed interpretation unit”) that had not yet been translated by the
interpreter during the course of the lecture. In this analysis, we used English-to-Japanese
interpretation data that had been stored in the CIAIR simultaneous interpretation
database at Nagoya University; specifically, data from two professional interpreters who
provided interpreting for the same English lecture. Based on the results of the analysis,
we observed that the interpreters could be divided into two types. The first type
demonstrates long delay times in translations, with large variance in those delay times.
This type often had one or more delayed interpretation units, and the timing of the start
of the translation was not uniform. The other type could be characterized as real-time
interpretation; that is, the delay times were short, and there were few delayed interpretation
units. Through this analysis, we confirmed that there are two different types of
strategies when following the original speaker’s utterance in interpreting settings.
