论文部分内容阅读
摘要：口譯已经成为许多国际会议场合必不可少的服务。由于发言者在英文演讲过程中常带有口音，这对口译员的工作造成了诸多困难。该论文共展示并分析了包括共同作者网络、共同机构网络和关键词共现等三种使用CiteSpace形成的知识图谱，以及总体发文趋势图，以吉尔的认知负荷模型为理论指导，旨在调查带口音源语演讲对口译员的表现产生的影响，并着重分析其对口译员听辨过程产生的影响，最后为师生提出实践性的建议以提高口译质量。
关键词：带口音源语文本;口译;CiteSpace;共现分析;吉尔的口译模型
中图分类号：H315.9
Key words： accented source text; interpreting; CiteSpace; co-occurrence analysis; influence; strategies
引 言
As the development of globalization， the 21st century has become an age of flourishing international exchanges that require interpreters to facilitate communication between parties who speak different languages， while many speakers who speak English have a strong accent such as Indian accent， French accent， Spain accent， Japanese accent and even Scotland accent. According to Daniel Gile's Effort Models （2009）， interpreters confronted with source language speech with strong accents are forced to allocate more efforts on listening and analysis， which is the first step in the process of interpreting， resulting in production delay and a loss of information. Moreover， unfamiliar English accents can lead to stress， affecting the intonation and ability to make immediate response of interpreters. Therefore， to guarantee the quality of interpretation production， figuring out the specific impacts of accents and promoting practical strategies are of great significance.
This essay aims at analyzing the effects of accented source language speech on interpreters' performance particularly Listening and Analysis Effort based on bibliometric analysis and thus proposing practical suggestions for further improvement for both students and teachers. Hereinafter the bilbiometrics is a branch of scientometrics， used for developing and measuring information as well as transferring knowledge in traditional printed articles and electronic information environment （i.e. database）. And bibliometric analysis is frequently used in scientific researches to draw knowledge maps （i.e. visualization， co-author network and bibliometric coupling） for further quantitative and qualitative analysis of online database. In interpreting studies， bibliometric analysis is applied for explore the research frontier and present situation such as institutional cooperation.
1Literature
1.1The Effort Models of Interpreting
Printer （1969） has pointed out that there is some association between the deterioration of the interpreter's performance and some kind of overload of short-term memory. Some cognitive psychologists （Broadbent 1958; Moray 1967; Kahneman 1973; Norman 1976） have found out that several mental operations need attention （alternative name is 'processing capacity'） and time. And the interpreter's performance deteriorates if the requirements of attention is more than his/her available processing capacity. Three Efforts the mental operations involved in interpreting including the Listening and Analysis Effort， the Production Effort and the Memory Effort. The three terms are defined by Gile （2009） as following： The Listening and Analysis Effort or Comprehension Effort comprises all comprehension-oriented operations， from the subconscious analysis of the sound waves carrying the source-language speech which arrive at the interpreter's ears by means of the identification of vocabulary to the ultimate decisions about the 'meaning' of the text.
The Production Effort， the output part of interpreting， is the set of operations extending from the mental representation of the message to be delivered to speech planning and the performance of the speech plan， including self-monitoring and self-correction when necessary. The Memory Effort refers to the ability to keep information and process it over a short period.
1.2Listening and Analysis Effort in Interpreting
Firstly， As pointed out by Mendelsohn （2001）， in real-life listening we listen in different ways， depending on our purpose for listening. In China， English listening practice is test-oriented when listening to materials. It concentrates on the expression and usage of keywords， speech， intonation and other grammatical aspects. The content understood by listeners is superficial and usually lacks accuracy. However， interpreters aim at understanding the overall semantic meaning of the source speech or the intention of the speaker in the process of listening and analysis， and are guided by accurate restatement. As Gile （2009： 79） said， translation involves at least some degree of non-trivial comprehension of the source-language utterance， meaning that such comprehension is more than the simple recognition of vocabulary and linguistic structures.
Secondly， the degree of the listener's participation is not the same. Li Hui （2019） holds the view that listening practice for English testing is one-sided in receiving information and lags in response to information. It mainly activates the auditory system while understanding is just a passive and incidental process. However， listening comprehension is considered by some researchers as an interactive process between top-down and bottom-up processing and as a continuous modification of a developing interpretation in response to incoming information and pre-existing knowledge （Brownell， 1996; Buck， 1991; Harmer， 2001）. The translator is active in the process of listening and analyzing， accompanied by a large degree of prediction and judgment. Often， non-verbal factors need to be mobilized to analyze， organize， supplement and associate the content heard. Therefore， listening and analysis in interpreting is more difficult than listening practice for English testing， and thus requires more attention and energy. 1.3Conceptual Terms
Modularity Q is one measure of the structure of networks or graphs， which was designed to measure the strength of division of a network into modules （also called clusters）. Networks with high modularity Q are reasonably divided into loosely coupled clusters （Chen， 2014）.
Silhouette Score is a heuristic tool to see whether the parameters chosen by researchers are reasonably and estimate the level of uncertainty when interpreting a cluster’s culture， being bounded by 1. A silhouette score of 1 means that a cluster is distinct from other clusters.
Betweenness Centrality measures the extent to which a node connects to other nodes in a network. A high betweenness centrality value suggests that the node plays an essential role in the communication between other nodes.
2Methodology
2.1Data Source
As a key national information construction project under the lead of Tsinghua University， and supported by PRC Ministry of Education， China National Knowledge Infrastructure （CNKI） has become the gold standard for research discovery and analysis in China， linking publications and researchers to databases spanning various disciplines by citing and controlling indexes. This essay used the core set of the CNKI database as the data source and the search strategy of "Topic = Interpreting + Accent" to collect a total of 164 articles between January 1， 2005 and December 31， 2019. Article data were then stored in the form of "Refworks Exportation" using the "TXT Plain Text" format for subsequent analysis. The amount of data collected in this study meets the requirements for CiteSpace analysis （Chen， 2017）.
2.2Research Procedures
This study adopted a bibliometric method to perform an general trend analysis on the collected documents. In addition， a co-author network， co-institute network was conducted， and keyword co-occurrence was analyzed. According to Chen （2014）， the basic function of the CiteSpace is network analysis and visualization， enabling researchers to explore the intellectual landscape of a knowledge domain and discern what problems they have been trying to solving and what methods and tools they have developed to reach their goals.
3Research Results and Discussion
3.1Co-Author Analysis
The co-author network is used to analyze the joint research in a certain research field. As shown in Figure 2， the CiteSpace analysis results show a total of 105 nodes， indicating 105 highly prolific authors， and 5 bilateral cooperative lines， indicating that the collaboration rate between the authors is very low. No relatively stable and large academic teams have been formed. Referring to Figure 3， we can notice that only three authors including BAI （2018）， JI （2012） and SONG （2010） published more than one article. It can be concluded that the number of average article published per author is quite small. 3.2Co-Institute Analysis
The network of co-institute in Figure 4 indicates that there are 23 research institutes and two collaboration links. Since most of the nodes are isolated points， it can be inferred that nearly all the research is conducted by a single author. Only two pairs of institutes including the School of Interpreting and Translation Studies of Guangdong University of Foreign Studies， the College of Applied Foreign Languages of Shenzhen Polytechnic have cooperative relationship， the College of Foreign Studies in Guangxi University， and the School of Foreign Studies at Nanjing University. And the detailed information about the publications of seven top active institutes are shown in Figure 5， indicating that the top three universities are Heilongjiang University （19 articles）， Hebei Normal University （12 articles） and Beijing Foreign Studies University （11 articles）. Apparently， the universities which published most essays have no cooperative relationship with other institutes.
The source text mainly describes the macro-environment of climate change has quietly affected the operation of museums. As significant public facilities of educational publicity， museums need to take the initiative to assume the sense of corporate social responsibility. The administrators must carry on the green transformation to buildings and installations. And it is environmental entrepreneurship that guides this change. From an overall perspective， the author have identified four major areas of renovation， including suitable workspace， green technology and material selection， energy consumption and management and communicating environmentalism. By listing the experience of some museums， the author illustrates the respective advantages of these four kinds of areas and introduces the concrete practical methods.
3.3Keyword Co-Occurrence Analysis
Due to the close relation between the keywords and the cores of articles， the analysis of similar keywords can facilitate the identification of the cores of research on interpreting and accent. The terms are being grouped hereinafter， for instance， preparations， liaison interpreting and coping strategies can be divided into one category， and case study and Gile's models are another group. Figure 6 is a keyword co-occurrence network of 292 nodes generated from the core database， where nodes represent the keywords. Modularity Q reached 0.8659 > 0.3， and the structure is reasonable （Newman， 2004）， suggesting that the network has dense connection between nodes within modules. Silhouette score arrived at 0.5549 > 0.4， and the result is reasonable （Rousseeuw & Kaufman， 1990）， indicating that the node is placed in the correct cluster. As is demonstrated in Figure 6， the keywords which are most concerned by interpreting researchers include accent， Gile’s Effort Models， consecutive interpreting， resolutions and preparations. According to Gile （2009）， interpreters have more processing capacity and time for discourse production in consecutive interpreting than that in simultaneous interpreting. The interpreter's performance， in result， has greater chances to be improved. Furthermore， most companies that are about to have conference have insufficient time and fund to prepare professional facilities such as an interpretation booth or audience receivers for simultaneous interpreting. So consecutive interpreting becomes a more common choice especially in business field and draws attention from researchers.
4Conclusions and Suggestions
4.1Conclusions
In conclusion， this essay identifies general trend of annual publication， authors， institutions， and keywords that had significant influence on accent and interpreting study. It can be concluded based on the general trend analysis that studies on interpreting and accented source language speech are rare and has prospective future. Besides， mostly studies in this field are accomplished by individual researchers in China because the number of lines between nodes in the co-author network is very small. In addition， lines connected nodes in the co-institute network are also few， indicating that cooperation between institutions needs to be encouraged to save unnecessary efforts and resources and thus generate more comprehensive views. On the basis of high-frequency keywords， it can be deduced that strategies to deal with accent in interpreting is of top priority for scholars. And aspects like accent， consecutive interpreting， Gile’s Effort Models， preparations and resolutions are the hotspots and research trend in the recent years.
4.2B. Suggestions
Most importantly， interpreters should follow the instruction of the basic comprehension 'equation' put forward by Gile （2009）， which is C （Comprehension） = KL （Knowledge of Linguistics） + ELK （Extralinguistic Knowledge） + A （Deliberate Analysis）， to ensure comprehension. Linguistic knowledge is composed of various elements such as lexis， phonics， grammar and the Sender's mother tongue which can result in syntax， incorrect word usage and unnatural pronunciation that make the comprehension task harder. Gile （2009） argues that a good mastery of the Sender's mother tongue offers the interpreter more elements in support of analyzing errors in pronunciation， words and grammar （for instance in Japanese accent， /l//n//m/ are omitted or mispronounced， while /r/ is deviated into /l/. In Arabic accent， /p/ is deviated into /b/）. Extralinguistic knowledge contains encyclopedic knowledge stored in the interpreters' long-term memory as well as information about the occasion， background of main participants of the communication and the speaker's preference of quoting in favor of the anticipation and speech comprehension in terms of linguistic disambiguation.
Meanwhile， Vandergrift （2004） stated that interpreting students should integrate top-down process by using context and prior knowledge to build a conceptual framework for comprehension and bottom-up process by combining larger units of meaning from phoneme-level up to disclosure-level to enhance their comprehension ability.
References
[1] Gile. D. （2009）. Basic Concepts and Models for Interpreter and Translator Training. Philadelphia： John Benjamins Publishing Company.
[2] Vandergrift. L. （2004）. Listening to Learn or Learn to Listen. Annual Review of Applied Linguistics， 24， 3-25.
[3] 鲍刚. （2005）. 口译理论概述. 北京： 中国对外翻译出版社.
[4] 陈悦， 陈超美， 刘泽渊， 等. （2015）. CiteSpace 知识图谱的方法论功能. 科学学研究， 33（2）， 242-253.
[5] 李辉. （2019）. 浅析口译中的听辨能力. 才智， 33， 241.
[6] 雷中华. （2018）. 吉尔谈口译与教学. 学术访谈， 06， 61-65.
[7] 田甜. （2013）. 吉尔模式下口音对口译员口译表现的影响及应对策略. 硕士学位论文. 上海： 上海外国语大学
[8] 谢瑜. （2019）. 口譯课堂中语音听辨能力的培养. 海外英语， 17， 61-62.
关键词：带口音源语文本;口译;CiteSpace;共现分析;吉尔的口译模型
中图分类号：H315.9
Key words： accented source text; interpreting; CiteSpace; co-occurrence analysis; influence; strategies
引 言
As the development of globalization， the 21st century has become an age of flourishing international exchanges that require interpreters to facilitate communication between parties who speak different languages， while many speakers who speak English have a strong accent such as Indian accent， French accent， Spain accent， Japanese accent and even Scotland accent. According to Daniel Gile's Effort Models （2009）， interpreters confronted with source language speech with strong accents are forced to allocate more efforts on listening and analysis， which is the first step in the process of interpreting， resulting in production delay and a loss of information. Moreover， unfamiliar English accents can lead to stress， affecting the intonation and ability to make immediate response of interpreters. Therefore， to guarantee the quality of interpretation production， figuring out the specific impacts of accents and promoting practical strategies are of great significance.
This essay aims at analyzing the effects of accented source language speech on interpreters' performance particularly Listening and Analysis Effort based on bibliometric analysis and thus proposing practical suggestions for further improvement for both students and teachers. Hereinafter the bilbiometrics is a branch of scientometrics， used for developing and measuring information as well as transferring knowledge in traditional printed articles and electronic information environment （i.e. database）. And bibliometric analysis is frequently used in scientific researches to draw knowledge maps （i.e. visualization， co-author network and bibliometric coupling） for further quantitative and qualitative analysis of online database. In interpreting studies， bibliometric analysis is applied for explore the research frontier and present situation such as institutional cooperation.
1Literature
1.1The Effort Models of Interpreting
Printer （1969） has pointed out that there is some association between the deterioration of the interpreter's performance and some kind of overload of short-term memory. Some cognitive psychologists （Broadbent 1958; Moray 1967; Kahneman 1973; Norman 1976） have found out that several mental operations need attention （alternative name is 'processing capacity'） and time. And the interpreter's performance deteriorates if the requirements of attention is more than his/her available processing capacity. Three Efforts the mental operations involved in interpreting including the Listening and Analysis Effort， the Production Effort and the Memory Effort. The three terms are defined by Gile （2009） as following： The Listening and Analysis Effort or Comprehension Effort comprises all comprehension-oriented operations， from the subconscious analysis of the sound waves carrying the source-language speech which arrive at the interpreter's ears by means of the identification of vocabulary to the ultimate decisions about the 'meaning' of the text.
The Production Effort， the output part of interpreting， is the set of operations extending from the mental representation of the message to be delivered to speech planning and the performance of the speech plan， including self-monitoring and self-correction when necessary. The Memory Effort refers to the ability to keep information and process it over a short period.
1.2Listening and Analysis Effort in Interpreting
Firstly， As pointed out by Mendelsohn （2001）， in real-life listening we listen in different ways， depending on our purpose for listening. In China， English listening practice is test-oriented when listening to materials. It concentrates on the expression and usage of keywords， speech， intonation and other grammatical aspects. The content understood by listeners is superficial and usually lacks accuracy. However， interpreters aim at understanding the overall semantic meaning of the source speech or the intention of the speaker in the process of listening and analysis， and are guided by accurate restatement. As Gile （2009： 79） said， translation involves at least some degree of non-trivial comprehension of the source-language utterance， meaning that such comprehension is more than the simple recognition of vocabulary and linguistic structures.
Secondly， the degree of the listener's participation is not the same. Li Hui （2019） holds the view that listening practice for English testing is one-sided in receiving information and lags in response to information. It mainly activates the auditory system while understanding is just a passive and incidental process. However， listening comprehension is considered by some researchers as an interactive process between top-down and bottom-up processing and as a continuous modification of a developing interpretation in response to incoming information and pre-existing knowledge （Brownell， 1996; Buck， 1991; Harmer， 2001）. The translator is active in the process of listening and analyzing， accompanied by a large degree of prediction and judgment. Often， non-verbal factors need to be mobilized to analyze， organize， supplement and associate the content heard. Therefore， listening and analysis in interpreting is more difficult than listening practice for English testing， and thus requires more attention and energy. 1.3Conceptual Terms
Modularity Q is one measure of the structure of networks or graphs， which was designed to measure the strength of division of a network into modules （also called clusters）. Networks with high modularity Q are reasonably divided into loosely coupled clusters （Chen， 2014）.
Silhouette Score is a heuristic tool to see whether the parameters chosen by researchers are reasonably and estimate the level of uncertainty when interpreting a cluster’s culture， being bounded by 1. A silhouette score of 1 means that a cluster is distinct from other clusters.
Betweenness Centrality measures the extent to which a node connects to other nodes in a network. A high betweenness centrality value suggests that the node plays an essential role in the communication between other nodes.
2Methodology
2.1Data Source
As a key national information construction project under the lead of Tsinghua University， and supported by PRC Ministry of Education， China National Knowledge Infrastructure （CNKI） has become the gold standard for research discovery and analysis in China， linking publications and researchers to databases spanning various disciplines by citing and controlling indexes. This essay used the core set of the CNKI database as the data source and the search strategy of "Topic = Interpreting + Accent" to collect a total of 164 articles between January 1， 2005 and December 31， 2019. Article data were then stored in the form of "Refworks Exportation" using the "TXT Plain Text" format for subsequent analysis. The amount of data collected in this study meets the requirements for CiteSpace analysis （Chen， 2017）.
2.2Research Procedures
This study adopted a bibliometric method to perform an general trend analysis on the collected documents. In addition， a co-author network， co-institute network was conducted， and keyword co-occurrence was analyzed. According to Chen （2014）， the basic function of the CiteSpace is network analysis and visualization， enabling researchers to explore the intellectual landscape of a knowledge domain and discern what problems they have been trying to solving and what methods and tools they have developed to reach their goals.
3Research Results and Discussion
3.1Co-Author Analysis
The co-author network is used to analyze the joint research in a certain research field. As shown in Figure 2， the CiteSpace analysis results show a total of 105 nodes， indicating 105 highly prolific authors， and 5 bilateral cooperative lines， indicating that the collaboration rate between the authors is very low. No relatively stable and large academic teams have been formed. Referring to Figure 3， we can notice that only three authors including BAI （2018）， JI （2012） and SONG （2010） published more than one article. It can be concluded that the number of average article published per author is quite small. 3.2Co-Institute Analysis
The network of co-institute in Figure 4 indicates that there are 23 research institutes and two collaboration links. Since most of the nodes are isolated points， it can be inferred that nearly all the research is conducted by a single author. Only two pairs of institutes including the School of Interpreting and Translation Studies of Guangdong University of Foreign Studies， the College of Applied Foreign Languages of Shenzhen Polytechnic have cooperative relationship， the College of Foreign Studies in Guangxi University， and the School of Foreign Studies at Nanjing University. And the detailed information about the publications of seven top active institutes are shown in Figure 5， indicating that the top three universities are Heilongjiang University （19 articles）， Hebei Normal University （12 articles） and Beijing Foreign Studies University （11 articles）. Apparently， the universities which published most essays have no cooperative relationship with other institutes.
The source text mainly describes the macro-environment of climate change has quietly affected the operation of museums. As significant public facilities of educational publicity， museums need to take the initiative to assume the sense of corporate social responsibility. The administrators must carry on the green transformation to buildings and installations. And it is environmental entrepreneurship that guides this change. From an overall perspective， the author have identified four major areas of renovation， including suitable workspace， green technology and material selection， energy consumption and management and communicating environmentalism. By listing the experience of some museums， the author illustrates the respective advantages of these four kinds of areas and introduces the concrete practical methods.
3.3Keyword Co-Occurrence Analysis
Due to the close relation between the keywords and the cores of articles， the analysis of similar keywords can facilitate the identification of the cores of research on interpreting and accent. The terms are being grouped hereinafter， for instance， preparations， liaison interpreting and coping strategies can be divided into one category， and case study and Gile's models are another group. Figure 6 is a keyword co-occurrence network of 292 nodes generated from the core database， where nodes represent the keywords. Modularity Q reached 0.8659 > 0.3， and the structure is reasonable （Newman， 2004）， suggesting that the network has dense connection between nodes within modules. Silhouette score arrived at 0.5549 > 0.4， and the result is reasonable （Rousseeuw & Kaufman， 1990）， indicating that the node is placed in the correct cluster. As is demonstrated in Figure 6， the keywords which are most concerned by interpreting researchers include accent， Gile’s Effort Models， consecutive interpreting， resolutions and preparations. According to Gile （2009）， interpreters have more processing capacity and time for discourse production in consecutive interpreting than that in simultaneous interpreting. The interpreter's performance， in result， has greater chances to be improved. Furthermore， most companies that are about to have conference have insufficient time and fund to prepare professional facilities such as an interpretation booth or audience receivers for simultaneous interpreting. So consecutive interpreting becomes a more common choice especially in business field and draws attention from researchers.
4Conclusions and Suggestions
4.1Conclusions
In conclusion， this essay identifies general trend of annual publication， authors， institutions， and keywords that had significant influence on accent and interpreting study. It can be concluded based on the general trend analysis that studies on interpreting and accented source language speech are rare and has prospective future. Besides， mostly studies in this field are accomplished by individual researchers in China because the number of lines between nodes in the co-author network is very small. In addition， lines connected nodes in the co-institute network are also few， indicating that cooperation between institutions needs to be encouraged to save unnecessary efforts and resources and thus generate more comprehensive views. On the basis of high-frequency keywords， it can be deduced that strategies to deal with accent in interpreting is of top priority for scholars. And aspects like accent， consecutive interpreting， Gile’s Effort Models， preparations and resolutions are the hotspots and research trend in the recent years.
4.2B. Suggestions
Most importantly， interpreters should follow the instruction of the basic comprehension 'equation' put forward by Gile （2009）， which is C （Comprehension） = KL （Knowledge of Linguistics） + ELK （Extralinguistic Knowledge） + A （Deliberate Analysis）， to ensure comprehension. Linguistic knowledge is composed of various elements such as lexis， phonics， grammar and the Sender's mother tongue which can result in syntax， incorrect word usage and unnatural pronunciation that make the comprehension task harder. Gile （2009） argues that a good mastery of the Sender's mother tongue offers the interpreter more elements in support of analyzing errors in pronunciation， words and grammar （for instance in Japanese accent， /l//n//m/ are omitted or mispronounced， while /r/ is deviated into /l/. In Arabic accent， /p/ is deviated into /b/）. Extralinguistic knowledge contains encyclopedic knowledge stored in the interpreters' long-term memory as well as information about the occasion， background of main participants of the communication and the speaker's preference of quoting in favor of the anticipation and speech comprehension in terms of linguistic disambiguation.
Meanwhile， Vandergrift （2004） stated that interpreting students should integrate top-down process by using context and prior knowledge to build a conceptual framework for comprehension and bottom-up process by combining larger units of meaning from phoneme-level up to disclosure-level to enhance their comprehension ability.
References
[1] Gile. D. （2009）. Basic Concepts and Models for Interpreter and Translator Training. Philadelphia： John Benjamins Publishing Company.
[2] Vandergrift. L. （2004）. Listening to Learn or Learn to Listen. Annual Review of Applied Linguistics， 24， 3-25.
[3] 鲍刚. （2005）. 口译理论概述. 北京： 中国对外翻译出版社.
[4] 陈悦， 陈超美， 刘泽渊， 等. （2015）. CiteSpace 知识图谱的方法论功能. 科学学研究， 33（2）， 242-253.
[5] 李辉. （2019）. 浅析口译中的听辨能力. 才智， 33， 241.
[6] 雷中华. （2018）. 吉尔谈口译与教学. 学术访谈， 06， 61-65.
[7] 田甜. （2013）. 吉尔模式下口音对口译员口译表现的影响及应对策略. 硕士学位论文. 上海： 上海外国语大学
[8] 谢瑜. （2019）. 口譯课堂中语音听辨能力的培养. 海外英语， 17， 61-62.