Simplification in Interpreted Speech: Evidence from an entropy-based algorithm

Activity: Talk or presentationInvited talk

Description

Motivated by advancements in corpus-based translation studies(CBTS), the field of corpus-based interpreting studies(CBIS) has experienced considerable growth over the past two decades. This data-driven methodology enables the systematic analysis of extensive machine-readable data derived from authentic interpreting events, fostering a more coherent and comprehensive understanding of interpreting. Mirroring CBTSCBIS has demonstrated a keen interest in identifying distinctive linguistic properties that differentiate interpreted speech from other language varieties. One such property frequently explored in CBIS is simplification which posits that interpreting exhibits linguistic simplification compared to native speech. However, research focusing primarily on individual linguistic features has not provided sufficient evidence to substantiate the existence of this simplification universal.
The present study investigates the simplification hypothesis in interpreting and its cognitive implications by examining the linguistic complexity of English speech simultancously interpreted from Chinese,original English speech produced by native speakers (L1speech), and original English speech produced bynon-native speakers(L2 speech). Two entropy-based measures, specifically unigram and POS entropy,were employed to assess whether interpreted speech exhibits greater simplification than the other two language varieties from an information-theoretical perspective. Our findings indicate that interpreted English tends to be lexically simpler than non-translated native speech,but not simpler than L2 English speech, as measured by unigram entropy. Notable differences were also observed at the syntactic level among the three language varieties as measured by POS entropy. In particular interpreted speech is less simplified than native speech, but more complex than L2 speech.The findings of this study have enriched our understanding of simplification phenomenon in interpreting studies from a quantitative linguistics viewpoint.
Period20 May 2023
Event title第四届数字人文研究论坛
Event typeConference
LocationQufu, ChinaShow on map
Degree of RecognitionNational