Corpus-Based Translation Studies Research Papers (original) (raw)

Just a few years ago, the term “cloud computing” had little meaning to many people. Nowadays it has come to be associated not only with computer technology, but also with translation tools.
Over the past 20 years the field of computer-assisted translation (CAT) tools and technologies has witnessed substantial progress and, with the advent of cheap desktop computing and the internet it has been possible to see translation barriers falling down. Currently, almost the entire translation market is run by small and medium-sized organisations competing with larger ones. Thus, this situation has acted as a barrier to CAT tools adoption since large translation companies can afford to build automated workflow and CAT tools, but for many small enterprises this is less feasible. Thus, they have to rely on cloud computing translation technology. Things are changing dramatically since Google and the social networks have shown that the future is not on the desktop, but on the internet. One of the benefits of this solution is that translation companies supply their translators with tools that require no licences and, most importantly, do not require any installation or training. In addition, the staff support has direct access to any problem that might arise and is able to provide support. Another advantage is that multiple translators can work simultaneously on the same file, obviating the additional onerous task of cutting up files into smaller parts and then stick them back together after the translation process. Using a browser-based system means that there is no need to send files via mail. All the assets are updated and shared in real time, and data remain secure on the server. Another interesting feature of browser-based systems is that the design tends to be cleaner and functional than with desktop packages. Translators can concentrate only on what they have to do: translate.
The aim of this research is to point out the significant role of cloud translation in a brand new translation environment. The entire work is subdivided into five chapters, each focusing on the different methodologies used in the field of text analysis and translation, including the traditional methods as well as the latest related to cloud computing. The first chapter offers an overall review of traditional dissections in linguistic analysis, and especially in Corpus Linguistics, as well as the main linguistic theories worked out by major corpus scholars. Corpus Linguistics, indeed, offers effective methods of language analysis, taking into consideration its use in real linguistic environments since it relies on empirical data. The second chapter provides a useful guide as well as a theoretical background for corpus compilers, indicating the most efficient techniques to build a Do-It-Yourself (DIY) corpus. It starts taking into account the most important considerations by scholars such as Tognini Bonelli, Halliday, Sinclair and Leech not only on corpus building processes, but also on lexis structure, pragmatics and semantics. Then, it offers reflections about Sinclair’s main areas of investigation when analysing a corpus, i.e. collocation, colligation, semantic preference and discourse prosody. At the end of the chapter a detailed user guide to AntConc, a very useful free concordance tool for corpus investigation, is provided. The third chapter presents a detailed corpus-based investigation and analysis of veterinary research articles dealing with muscular diseases. The investigation offers meaningful personal deductions about the most frequent words occurring in all texts and the relations with their collocates, colligates, semantic preference and useful considerations about semantic prosody. The fourth chapter covers new approaches in Corpus Linguistics since it focuses on new theories which rely on the web in order to build corpora. As a matter of fact, the web is also considered as an excellent corpus source since it provides the largest text collection in the world. One of the tools available for online corpus building and investigation is the Sketch Engine. It is run by humans, who should establish themselves the parameters of research in advance. Once the corpus has been created, the Engine provides efficient functions for corpus analysis. In the case of the present research, the Sketch Engine has been used as a corpus builder uploading target language files, so that abundant considerations on texts belonging to the same field of research of the source texts could be made. Then, the results of this preliminary investigation have also been used as a valid help and support, for research purposes, to translate a scientific article. The last chapter, the fifth, is the most relevant one since it deals with terminology management systems and translation memory systems; on how translators can work on existing TMSs or how they can create new ones. Then, the chapter offers a review of the most popular computer-assisted translation (CAT) tools, from Trados up to the most recent one, KantanMT, a cloud-based translation tool. The last part of the chapter, indeed, copes with the latest advantages in translation technology in terms of cloud translation. With the KantanTM tool translators can work on their tasks directly on the web, without having to worry about downloading software on their computers. In order to prove the effectiveness of these new cloud tools, KantanMT was used to translate automatically the same text that had been previously translated using traditional methods. The last step of the present research offers an appendix with the first draft translation of the research article, the text translated by KantanMT, and at the end, the final and correct version of the translation (which can be considered as a combination of the two translations) completed with the aid and support of the findings acquired with the Sketch Engine.