Speech Recognition Software:Friend Or Foe For Trainees Of Chinese-English Simultaneous Interpretation?

Will the textual output of speech recognition software help trainee interpreters reduce memory load and improve the information integrity of their interpretation?To answer the question,the author conducted an experiment followed by a survey to explore the influence of speech recognition software on trainee interpreters.Four Chinese speeches different in register,speed of delivery and information were used.Two groups of trainees were invited for simultaneous interpretation with and without speech recognition output respectively.After the interpretation,all experiment participants were asked to fill in the designed questionnaires.The experiment shows iFlynote,the speech recognition application used in this research,had consistent output quality irrespective of speed and speech content and produced results easily readable and comprehensible,with an accuracy rate of over 98%.The textual output is especially helpful in improving the accuracy of figure interpreting.The delays of speech recognition,however,remain inconsistent:they vary with the speed of delivery.When a speech is delivered at high speed,the delays are generally too long for simultaneous interpretation as a time-sensitive event.In terms of the integrity of the information interpreted,the two groups do not exhibit significant differences in general.But individually,differences exist.In the experimental group,trainees with strong sight interpreting skills can benefit from the aid of the software.In contrast,trainees whose sight interpreting skills are underdeveloped are negatively influenced by the application:they either suffered from cognitive overload or became overly reliant on the textual output.The survey confirmed the above analysis.The research shows,speech recognition software,although in need of improvement,can already offer some assistance to simultaneous interpretation.However,students should receive trainings to fully tap the potential.