Quasi-Translation" of the Pun: The Specificity of Using Algorithms

The article is devoted to the problem of the translation of pun, which poses a challenge not only for beginners, but also for experienced translators and interpreters. The leading approach to solving this problem is to build translation algorithms for various types of puns. The need to develop algorithms is caused by the variety of operations that a translator performs to meet the challenge. The experiments on the translation of jokes containing puns based on the use of homonyms, paronyms, polysemants and phraseological units (more than 500 examples have been considered) revealed the most common types and sequences of the operations performed by a translator. The revealed regularities formed the basis of algorithms in which the cognitive processes occurring in the brain of a translator are simulated. The algorithm includes three main steps: recognition of signs used to create ambiguity in the source language; search for matching signs in the target language; the re-creation of the pun with the use of the matching signs in the target language. Each step involves a particular sequence of operations consisting of conditional and non-conditional steps. The translation and didactic experiments were carried out in order to identify the specificity of using algorithms. The translation experiment consisted in using algorithms by translation practitioners. At the first stage, the participants had to draw on their own experience when translating puns; at the second stage, they were offered the stimulus material. The survey revealed that more than 91% of participants stated that the algorithms simplify the translation task; more than 83% noted that they also help to reduce the time of translation. Under the didactic experiment the students of the translation department were proposed to translate four pun-based jokes from English into Russian. At the first stage, the participants of the experiment had to translate the jokes by using their own knowledge; at the second stage, they were given a course of lectures on the application of algorithms when creating a "quasi-translation". The experiment demonstrated that the use of algorithms improves the quality and the variety of translations (the number of "quasi-translations" increased by 72%, the variability by more than 30%). It also helps to avoid stalemates and develops a creative approach that does not allow to go beyond the scope of the translation. The materials of the article are of practical value for translation teachers, as well as translation practitioners. The results of the study may also have an applied nature. The algorithms associated with searching for matching signs in the target language can be used to build training and testing sets for machine learning when composing language pairs. This will increase the amount of teaching material and improve the quality of neural machine translation.

Keywords

Authors

References