| The purpose of this study is to explore how to train post-editors to meet the market demands in the new age.Post-editing is defined as a process,whereby translators amend a machine-generated translation to meet a certain level of quality.In this process,a series of cognitive skills are involved,such as problem analysis,decision-making and evaluation.The post-editing training is mostly oriented towards empowering students with these cognitive skills to succeed in the marketplace.Although post-editing has been widely implemented by language service providers in recent years,post-editing training has been less investigated from both the research and training perspectives.There currently exists a small number of conceptual research on post-editing training,largely depending on personal experience.However,studies based on a social learning theory and adequate empirical evidence are surprisingly scarce,and those concerning on students’ cognitive process are even more scarce.Bearing this in mind,this study tries to propose a new pedagogical model for post-editing training.It attempts to apply interdisciplinary approaches into translation teaching to validate the model empirically,presenting an effective way to bridge the gap between theory and practice.This research initially constructs an online post-editing community model based on social constructivism.In this model,an online post-editing community will be generated around the authentic task.Members can interact online to solve problems,collect information and share knowledge to complete the task collaboratively.Following that,a teaching experiment is conducted to verify this model and,if necessary,adapt or modify it.It aims to answer the following three questions.(1)Does the online post-editing community model foster students’ problem-solving skills?(2)Does the online post-editing community model foster students’ self-assessment and self-reflection? Does it foster error mitigation in the post-editing output?(3)Does the online post-editing community model foster a positive attitude towards machine translation? What are students’ perceptions of the teaching model?To answer these questions,a quasi-experiment study with a mixed-method design was conducted.Two groups of students studying post-editing in the same course participated in the project.The experimental group was a virtual community of practice,using online collaborative tools for social learning,while the control group used the conventional face-to-face method.During the course,each group received the same instructional content and duration.The only independent variable was the teaching method.All the students were required to submit learning portfolios including online interactive reports,post-editing artefacts,self-assessment scores and reflection reports.Finally,attitude scales towards machine translation as well as the course were implemented to both groups.Major findings of the current study include:(1)The online post-editing community model fosters problem-solving skills.Inter-group comparisons showed that the experimental and control groups differed greatly in identifying,analyzing and resolving problems.Initially,the experimental group perceived more macro-level problems related to coherence,style,target reader and translation purpose.To solve these problems,they tended to use complex consultations searching domain-specific information more frequently.They also displayed a more advanced level of knowledge sharing in tacit knowledge of context,variations in meaning and understanding of post-editing purposes.Through online collaboration,they eventually found more effective solutions for these problems.Further,the experimental group worked in more harmonious ways,exhibiting more dynamics in the process of problem-solving.They had a stronger sense of group identity and self-identity.It indicated that the teaching method had the potential to facilitate a more positive collaborative problem-solving pattern.(2)The experimental group performed better on self-assessment as well as self-reflection,compared with the control group.Specifically,their mean scores of self-assessments on postediting quality were more consistent with those of tutor-assessment.Their overall level of reflection was 2.38,which was much higher than that of the control group(M=1.66).Deeper reflections could also be detected in the experimental group on six topics,including post-editing guidelines,problem-solving skills,information gathering,knowledge sharing,assessment skills,and collaborative skills.Moreover,great discrepancies in error mitigation could be found between the experimental and the control group.In the experimental group,the total number of errors and the number of syntactic and macro-level errors remaining in post-edited texts were significantly less.It further suggested the positive effect of the proposed model on post-editing quality,particular in the macro-level and syntactic accuracy.(3)The teaching model enhances positive perceptions of using machine translation.An independent-samples t-test showed that there was a significant difference in tendency-of-use of machine translation between the experimental and the control group(p=.022<.05).The experimental group demonstrated a stronger tendency to use machine translation and accept post-editing jobs.In addition,students’ attitudes towards the teaching model were,on the whole,very positive.93% of the experimental group agreed that the model of online post-editing community was a good way of running post-editing courses,which was a very positive result showing its usability.A high percentage thought that this model had the potential to foster collaboration,problem-solving and self-reflection.They also confirmed the benefits of the teaching method to individual learning motivation,confidence and self-directed learning.A refined model is presented finally,highlighting a more learner-centered mode as well as a dynamic,spiral process that evolves from novice to expertise.This study has the following implications for research on translator training.It primarily demonstrates an effective way to conduct a blended mode of translator training,empowering skill sets required for the future translation jobs.It also reflects that translator trainers should have fairly high-level facilitation skills in order to make online learning more engaging.They are also encouraged to use digital formative assessment tools such as e-portfolios and online interactive reports,to foster student’cognitive,behavioral and attitudinal development.Moreover,this study offers a promising path towards cognitive research in translation teaching,bridging theoretical and descriptive research to develop teaching methods that can later be used by practitioners.Nevertheless,concerning limitations in sample size,geographical areas,students’ levels of skills and the experimental environment,larger and randomly chosen experiments might yield more generalizable results.The online post-editing community model still requires further refinements,finally promoting its application in a large scale. |
PDF Full Text Request