Abstract
This study aims to identify the appropriate length of original speech per segment of consecutive
interpretation from English to Japanese. It is customary in Europe for an original utterance to
continue for about five minutes before interpretation starts. While this duration is believed to be
much shorter in Japan, there is no consensus in practice or education based on scientific
evidence. The longer the segment interpreters must listen to at one time, the greater the memory
burden and harder it will be to remember the content. This could result in errors such as
omission or substitution. To investigate how the burden on interpreters’ memory affects the
precision of interpretation, an experiment on consecutive interpretation was conducted using
segments of three different lengths: 1, 2.5, and 5 minutes. The results indicate that errors
proportionately increase with the length of the segment.
