Syntactic Structure Differences Across Interpreted And Original Oral Texts

Interpreting is a cognitively demanding task that requires dealing with multiple cognitive tasks under intense temporal pressure.Previous studies have explored cognitive load differences to distinguish between two major interpreting task types,namely,simultaneous interpreting and consecutive interpreting,by analyzing linguistic features regarding lexical sophistication,syntactic sophistication,and language sequence organization.However,the existing quantitative syntactic inquiry of interpreting mainly investigated the syntactic complexity of interpreting products,but some other profound aspects of syntactic features,such as those regarding the distribution of syntactic relations,remain unexamined.To fill the gap,dependency relation and dependency direction are introduced into the present study.Dependency relation reflects the syntactic relationship between the governor and the dependent,and dependency direction denotes the linear order difference between the governor and dependent.This thesis selects the distribution of dependency relation and dependency direction as metrics to characterize syntactic features and language styles quantitatively between interpreted text and original oral text.Dependency grammar annotation treebanks,including three treebanks of interpreted texts: consecutive interpreting（CI）,simultaneous interpreting（SI）,and read-out translated speech（TR）output texts and two treebanks of original oral texts: texts of the State of the Union Address（SU）and texts of Presidential Election Debate（ED）.Results indicate some systematic differences regarding the distribution of dependency relations and the dependency direction between interpreted text and original oral text,reflecting some syntactic preferences regarding the two types of texts.There is no significant difference between written,spoken,and interpreted texts in the examined features of the subject,object,and adverbial structures;simultaneous interpreted and written texts show similar features concerning attributive structures;consecutive interpreted and spoken texts demonstrate similarity regarding cohesion and coordination structures.Three possible explanations are provided for the results.First,since it is a central norm of interpreting to minimize information loss,interpreters generally prioritize retaining key information in the source language as much as possible,even when faced with tremendous cognitive pressure.Second,simultaneous interpreters usually adopt a sentence-by-sentence interpreting strategy which tends to retain most of the syntactic structures of the source language,while consecutive interpreters can make use of the time interval between the comprehension and the production stages to express in the target language more “freely” due to less syntactic constraints of the source language.Third,the face-to-face working environment makes consecutive interpreters more oriented towards the audience’s perspective,which is inductive to the choice of more oral expressions in interpreting.As a quantitative inquiry of interpreting,the present study considers the distribution of syntactic relations in the deep structure of sentences for the first time,and explores the syntactic features of written,spoken,and interpreted texts.The results are plausibly associated with the cognitive load factors and pragmatic social factors that have potentially influenced these syntactic features.