With the rapid improvement of the virtual international, the busywork of translation is reduced, permitting more time for the exploration of translation as an arc shape. Repetitive phrases are inserted routinely, dictionaries are accessed with the press of a mouse, and content precise terminology is move-referenced to ensure accuracy. The question is, with such a lot of positives, where are the negatives? And is it realistic that a freelance translator can compete in brand new task marketplace without the assist of translation software? In the subsequent pages the advantages and drawbacks of the translation memory programme Wordfast, created through Yves Champollion, will be compared with the ones of the parallel concordancer, Paraconc, mainly the unfastened beta model created via Michael Barlow. The programmes may be tested on many stages, some very realistic, along with usability, performance, and fee effectiveness, and a few greater theoretical, along with their effect on the paintings and language comprehension stage of the translator.
Before entering into the strengths and weaknesses of Paraconc and Wordfast, I will deal with the query of necessity. The move towards computer-assisted translation equipment has been picking up speed with the arrival of low fee or loose software. With so many translators taking advantage of those offers and thereby increasing their productivity, the ones who’ve now not taken this step should ask themselves if they’ll be left at the back of. With a leaf through online job postings for freelancers, it leads one to consider that there are only paintings to be had for the ones the usage of translation reminiscence. Most agencies are more likely to give a job to someone who can do it in less time and with greater consistency. The strain to apply TM may not have an effect on the freelancers who have mounted direct customers offering regular work and trust simplest them to do it. The well-known consensus appears to be that without these tools the company work might be scarce 1.
The first issue that has to be addressed is that of constructing a parallel corpus. The most crucial stage in any construction is building a robust basis. In the case of translation software program, this is the development of the corpus. This starts offevolved with locating parallel texts, texts that have already been translated and are available in both the supplied language and the goal language. Depending on the concerned area and the language pair, these parallel texts can be hard to return by means of. As Maeve Olohan points out in her book Introducing Corpora in Translation Studies, the Internet offers some full-size resources for building a corpus but while they are published through global businesses and written by way of several authors it could be difficult to envision which one is the source text.2 Being that texts published on the Internet are not always edited or proofread before being made public, the source texts and their translations ought to be scrutinized for accuracy. The availability of relevant cloth can pose issues due to the fact the character of texts varies significantly relying on the language pair. Texts translated from Italian to English are commonly written at a higher stage, while those from English to Italian are greater regularly of the first-class-supplier variety.3
The next step in constructing the corpus is an alignment of the supply and goal texts. Precision is crucial right here; misalignment due to a further paragraph ruin or punctuation that splits a sentence in the wrong vicinity throws off the whole system and makes it not possible to match units. The translation reminiscence database is depending on accurate matching. Due to the inherent traits of the language pair that I work in, German and English, there are frequent occurrences of misalignment. The style of written German frequently consists of longer sentences than in English. Because of this, it is advocated inside the Paraconc commands that the shorter unit, usually the English version, be merged with the subsequent unit instead of breaking up the longer, generally German, textual content unit. Another problem that results in misalignment is abbreviations. German and English do not necessarily have corresponding abbreviated forms. Being that an automated alignment breaks up devices at every full stop, the English abbreviation ex., for ‘example’, could smash off after the total stop whereas the German equal, z.B., for ‘zum Beispiel’, might spoil off two times. Another example that has become a hassle for me became the way the date is written. In German, there is a full stop after the day and month while English uses the diminish to split these. When aligning a huge text, that information can add up to a sizable amount of correction.
Depending on the software program used, aligning the texts can take days or hours. Using Paraconc, the translator ought to complete all of the steps manually. This is speeded up with Microsoft Word’s “find and replace” tool but is quite time-consuming in spite of this. The supply textual content and goal textual content must be divided into units by way of inserting paragraph marks at the quit of every sentence. This is finished with the aid of finding and changing each complete forestall, exclamation point, or query mark with the equal mark accompanied by means of a paragraph mark. The hassle is, relying on the texts being labored with, full stops may be scattered generously all through the text, no longer necessarily denoting the give up of a sentence. This is, as stated above, common in German. With this being the case, paragraph marks have to be inserted manually as opposed to routinely. Microsoft Word then numbers each unit and the textual content is considered as an extended list of devices, throughout which the alignment must be checked and edited. The subsequent steps are to remove the numbers, spell check the files, and store them as textual content files.
Wordfast, on the other hand, has a fixed of tools known as PlusTools that offers a quicker and greater efficient technique of alignment. Compared to Paraconc, this is a relatively easy and straightforward procedure that we could the computer do the work. After commencing Wordfast, multiple source files may be selected and the gadgets extracted concurrently. After extraction the interpretation devices are counted saved; the equal is done with the goal texts. Following this, PlusTools is opened and by using clicking on ‘Align’, the supply and goal texts will be processed and appear either facet by using aspect in two separate files or as one document in desk format. The texts can then be manually changed and demonstrated the use of keyboard shortcuts. During the exploration of these shortcuts, I did come across a few problems; Alt M (merge), Alt W (delete cell), and Alt I (insert cell) did now not paintings, consequently, I used the mouse and the menu bar. The ‘locate and replace’ device in MS Word can also be used to correct repeated reasons of misalignment. The documentation that accompanies the PlusTools download gives recommendations for the alignment procedure.
As the paintings of the translator modifications with time, it is of a hobby to take into account the impact of these programmes on this work. The quality of a translator’s paintings relies upon numerous factors, one of the maximum critical being the intensity of his or her expertise of the languages labored with. It is essential to have giant assets accessible, with which you will confirm questionable phrases or terminology. In the beyond those resources were made from hard-reproduction dictionaries or private glossaries. With the improvement of parallel concordancing software, the translator is now provided with a device that is as dynamic as language itself. Texts applicable to the theme of the translation are aligned and can be used as an electronic aid; terms can be searched and regarded contextually, presenting a greater complete explanation of use. Parallel concordancing software program is greater regularly used as a coaching or studying device in language development than as a translation tool. As Olohan points out, expert translators not often permit themselves the time to end up familiar with software in order to no longer at once speed up their output. Parallel corpora are useful in exploring precise phrasing intensive; this exploration takes time, which is not continually available. Translators are much more likely to invest time in studying a translation reminiscence programme.4 Another helpful utility of corpora is in the purchase of LSP, or languages for particular purposes. Parallel texts may be loaded and used as an aid to developing a greater tremendous vocabulary when running in unique fields, together with technical, criminal, or clinical. The point wherein parallel concordancing may additionally have a bonus over TM is in the translation of more innovative texts. There are little terminology repetition and an extra consciousness on writing style. Here, the corpus is looked for styles or oddities that distinguish one character from some other or make a creator’s work their personal.5 When a translator takes the time to research and employ parallel corpora, this intensity of research and precision will, in reality, be pondered in the equivalence level of the translation and the “local-sound”.
Translation memory programmes are used widely via freelance translators. They offer immediate outcomes that are visible inside the boom of productivity. This relies upon loads on the sphere that the translator works in but whilst there is a demand for excessive output, TM can really assist. If the texts are of a more technical nature or have a massive quantity of repeated phrases or terminology, the TM will mechanically insert them into the interpretation. Wordfast, like many different TM programmes, leads the translator through the source text unit with the aid of unit. This ensures that every unit is translated and nothing is missed within the rush to satisfy time limits. As Frank Austermühl factors out in his e-book Electronic Tools for Translators, 3 critical factors should be evaluated earlier than selecting the usefulness of TM. These are as follows: textual content kind, TM is better proper to technical texts, re-usability, or how a great deal repetitive content material exists, and thirdly, quantity, longer texts are more likely to have repetitive content.6 Another sizeable benefit of a TM includes revisions. In her master’s thesis on translation reminiscence, Lynn Webb talks about the quantity of needless work that can be prevented when a customer sends a revised model of a text that the translator has already begun operating on. Rather than having to look for the changes or additions, the revised model may be run through the TM and the software reveals changes at once.7