MT V.s Human translation Essay

Introduction Today, computers are used in all fields, and even almost every field has it’s own software packages. Using computers to translate a text from one language to another refered to machine translation [MT]. Machine translation is an interesting technology for human translators. It is a fact that MT software can translate texts very quickly. The question is that: Are these machine translations perfect? Are these translation tools like Google valid? MT are somehow acceptable in technical and informative texts but how about literaral or expressive texts? According to Chapman† Literature is the art that uses language†(qtd.in Voigt and Jurafsky 1). So, literary translation represents the strongest formulation of machine translation problems. As MT quality continiues to improve, the idea of using MT to assist human translators becomes increasingly attractive, and human translators can correct mistakes in these machine translations. Translation is not only a linguistic act, but also a cultural one. It involves more than just a word-by-word representation of a text; translators also have to take double meanings, cultural subtleties and slang into accountContext of culture affects the specific meaning of the language. So the analysis of cultural context is essential for Machine Translation (MT). If the cultural context analysis of the source language is omitted in MT, ambiguity or mistranslation will be produced. At least nowadays when we compare MT with human translation, we claim that human say the last word. A Brief History Of Machine Translation The history of machine translation is as old as that of computers. It has been started in the 1950s. Georgetown –IBM experiment consisted of the automatic translation of Russian sentences in to English in a very speciallized field(Organic chemistry), and it was widely recognized as a successful demonstration. Documents in Russian gathered by the U. S. military and intelligence agencies during the 50’s and 60’s. Throughout this period university and government research funding drove the development of MT. However, the real progress was much slower, and in 1966 they found out that the ten years long research had failed to fulfill the expectations, so the funding was dramatically reduced until the late 70’s, at which time advances in theoretically linguistics and the growth of computing and language technology converged, resulting in the first practical MT tools for main frame systems. In the late 1980’s, as computational power increased and became less expensive, more interest began to be shown in statistical models for machine translation. Today there is still no system that provides the holy-grail of â€Å"fully automatic high quality translation† (FAHQT). However, there are many programs now available that are capable of providing useful output with in strict constraints; several of them are available onlin such as Google Translate and SYSTRAN system which powers Alta’s BabelFish. (Wikipedia 1) The importance of Human Translation Translation is not only a linguistic act, but also a cultural one and a prime channel of communication across cultures irrespective of geographic discrepancies. Cultural implications may be higher in cross-cultural translation and may range from lexical level to pragmatic level. More the gap between the source and target culture, the more serious difficulties would take shape. Translation between English and Hindi; which belongs to two different cultures and backgrounds is one of the best examples of such problems. In such situation, cross-cultural communication should be appropriately done using proper translation techniques to avoid ambiguity and miscommunication. Context of culture affects the specific meaning of the language. So the analysis of cultural context is essential for Machine Translation (MT). If the cultural context analysis of the source language is omitted in MT, ambiguity or mistranslation will be produced. Any attempt to replace Human Translation totally by machine translation would certainly face failure for, due to a simple reason, there is no machine translation that is capable of interpretation. For instance, it is only the human translator who is able of interpreting certain cultural components that may exist in the source text and that cannot be translated in terms of equivalent terms, just like what automatic translation does, into the language of the target text. In addition, it is widely agreed upon that one of the most difficult tasks in the act of translation is how to keep the same effect left by the source text in the target text. The automatic translation, in this regard, has proved its weakness, most of the time, when compared with a human translation. The human translator is the only subject in a position to understand the different cultural, linguistic and semantic factors contributing to leaving the same effect, that is left in the source text, in the target text. It is an undeniable fact that automatic translation is regarded as a tool for producing quick and great number of translated texts; nevertheless, the quality of the translation is still much debatable MT evaluation One way for people inorder to assess machine translation’s quality is kind of Back translation. I mean to translate from a source language to a target language and to the source language with the same engine. Although this way sounds good, it is a poor method. When we consider two variables â€Å"inteligibility† and â€Å"fiedelity† in our judgment, in most cases it is easy to separate translation by human from translation by machine. â€Å"Inteligibility is a measure of how understandable the sentence is and fidelity is a measure of how much information the translated sentence retained compared to the original†(Wikipedia 1). Although machine translation currently produces relatively unacceptable output compared to human translation, I do believe it will be much better in the future. Is machine translation output necessarily of lower quality than human translation? Some scholars believe that † Translators who work in technical domain will be increasingly require to interact with MT â€Å"(Pym 1). The need for technical translation has increased dramatically and in the future MT systems will continue to reduce the cost of translation. Advantages & Disadvantages of MT In the past when we had to find the meaning of a word from another language we used a dictionary . This was very time consuming. Moreover, when a paragraph or note had to translated, this could be very difficult because one word has several meanings. When time is crucial factor, with MT you don’t have to spend hours looking up dictionaries to translate the words. Instead, the software can translate it quickly . It is not costly but one of it’s disadvantages is that translation is not accurate and it can’t solve ambigiuity. It can’t produce translations for literary texts with good quality because translating literature requires special literary skills, but it doesn’t mean that machine translation is useless. The quality of translation which can get from an MT system is very low but we know human translator normally doesn’t produce a perfect translation. MT threats the job of translators. MT is an important topic sociolly, politically, commercially, scientifically, intellectually & philosophically. MT and Translating culture-Bound elements One of the most challenging tasks for all translators is how to translate culture-bound elements into a foreign language. According to Newmak: â€Å"Translation is a craft consisting in the attempt to replace a written message and/or statement in one language by the same message and/or statement in another language†(qtd in Armellino 1). When words in the source text are strongly rooted in the source culture that they are specific to the culture that produced them, therefore, they have no equivalent in the target culture because they are unknown, or because they are not yet codified in the target language. When cultural differences exist between the two languages, it is extremely difficult to achieve a successful translation. How can MT cope with problems of not only lexical expressions, but also with problems of register, syntactic order, dialects? MT has to decide on the importance of certain cultural aspects and to what extent it is necessary to translate them into the target language. Nida confers equal importance to both linguistic and cultural differences between the SL and the TL and concludes that â€Å"Differences between cultures may cause more severe complications for the translators than do differencs in language structure(qtd. in Glodjovic 2). Idioms are difficult to translate. It is sometimes hard to find the right equivalent for a single word without finding an equivalence for a sequence of words that convey one specific meaning. We know idiomds are culturally specific, which means that they may express a scene that doesn’t occure in the TL. Baker says: Idioms and fixed expressions which contain cultural specific items are not necessarily untranslatable. It is not the specific items an expressin contains but rather the meaning it conveys and it’s association with culture specific context which can make it understanable or difficult to translate. (qtd. in Muller 13) So translating the idioms mostly depends on the context in which it has occurred. Is it possible for Machine translator like Google translation to deal with such problems? What would be the best translation strategies for dealing with Idioms and culturally bound expressons? Human-Assisted Machine Translation Machine translation has faced many problems which can be solved by computer-assisted machine translation at the pre-editing and post-editing stages. As a result the final translation will be more acceptable if MT translation is edited by human inorder to generate more appropiate translation for some words in a sentence and as a result our translation could be semantically and pragmatically more proper and we discard odd and unnatural structures. In human-assisted translation the computer produce the first draft then the proffessional revises it. The question is that : Is machine and post-editing of MT output faster than human translation? To answer this question measuring time will be the main purpose, furthermore who should be doing post-editing? Should it be performed by translators, revisors, non-linguists, or trained specialists? According to Loffer-Laurian†Post-editing of machine-translated text is a task different from traditional HT and revision. Loffer-Laurian maintains that post-editing is not revision, nor correction rewriting. It is a new way of considering a text, a new way of working on it for a new aim†(qtd. in Martinez 23). Poetry and Machine Translation According to Oxford English Dictionary Poetry is â€Å"The art or work of poet†(qtd. in Hovhamisyan 1). Translating of poetry is one of the most difficult and challenging tasks for every translator. According to Robert Frost’s definition â€Å"poetry is what gets lost in translation†(qtd. in Hovhamisyan 1). To sum up the theoretical approaches, it is clear that poetry is the most difficult type of the text and can be considered to be untranslatable. Grammatical differences between the languages causes a lot of problems in translating poetry. Should we , then refrain from translating poetry. Where proffessional translators assumes that the translation of poetry is extremly difficult, is it possible for a machine softwares to translate poems among differen languages? In the following paragraph first sunnet of Hafez it’s English translation by and it’s Google translation are avalible. ? ? ? ? ? ? ? ? O beautiful wine-bearer, bring forth the cup and put it to my lips Path of love seemed easy at first, what came was many hardships. With its perfume, the morning breeze unlocks those beautiful locks the curl of those dark ringlets, many hearts to shreds strips. In the house of my beloved, how can I enjoy the feast since the church bells call the call that for pilgrimage equips. With wine color your robe, one of the old Magi’s best tips Trust in this traveler’s tips, who knows of many paths and trips The dark midnight, fearful waves, and the tempestuous whirlpool How can he know of our state, while ports house his unladed ships. I followed my own path of love, and now I am in bad repute How can a secret remain veiled, if from every tongue it drips? If His presence you seek, Hafiz, then why yourself eclipse? Stick to the One you know, let go of imaginary trips. Google Translation High boots or stands Casa Ella field and Novell I would be easy to love but difficult The smell of oak tress Saba Nafhay Kakhr open What was the blood from the heart twist lock Mshkynsh Janan how secure mirth at my house because the door JRS will scream that should tell concerning vehicle The mat is a colorful old Garrett says Taha The traveler did not know the way home. Fear of the dark night of the wave and vortex Heil We know where the loose banks All I took from his failure to end stigma Who would he have hidden the secret circle center Guardian of the person, do not be absent from his Hmykhvahy We produce p my Dunya Matthew invitation and Ahmlha In above two translations that one of them is done by human, but the other is produced by Google it is cristal clear that for translating poem MT is not acceptable. It is full of grammatical and lexical mistaks Works Cited Armellino, Elisa. â€Å"Translating Culture-Bound Elements in Subtitling. † Translation directory. N. D. Web. 16 June 2013. â€Å"Evaluation of Machine Translation. † Wikipedia. 15May 2103. Web. 2July 2013. â€Å"Evaluation of Machine Translation. † Wikipedia. 15May 2103. Web. 2July 2013. Fiederer,Rebecca and Sharon O Brien. â€Å"Quality and Machine Translation. † Jostran. Org/issue. 11January 2009. Web. 19June 2013. Glodjovic, Anica. â€Å"Translation as Means of Cross-Cultural Communication. † facta. junis. ni. ac. rs. June 2010. Web. 7July 2013. â€Å"History of Machine Translation. † Wikipedia. Web. 2July 2013. Hovhannisyan, Mariam. â€Å"The Art of Poetry and its translation. † Translation Directory. May 2012. Web. 27 June 2013. Martinez, Lorena. G. â€Å"Human Translation V. S Machine Translation. â€Å",Sceuromix. August 2003. Web. 3July 2013. Muller, Theo. â€Å"Translation of Idioms. † 17 September 2009. Web. 5July 2013. â€Å"Human Translation V. S Machine Translation. † Netmask. it/Products. 2003. Web. 5 July 2013. Pym, Anthony. † Translation Skill-sets in a Machine-Translation. † usuaris. tinet. cat/apym/on-line/training/2012_competence_Pym. May 2012. Web. 1 July 2013. Voigt, Rob and Dan Jurafsky. â€Å"Toward literary Machine translation. † Stanford. edu/jurafsky, N. D. Web. 1July 2013.