RIYADH — The debate over machine translation (MT) vs. human translation (HT) continues, with the question of whether MT will eventually replace HT in an era when MT is rapidly advancing and improving. A group of Saudi and international translation and interpretation academics and experts believe MT will replace HT, while another group believes MT will not replace HT. Meanwhile, a third group of those academic and experts agree that the relation between MT and HT is complementary. In an interview with the Saudi Press Agency (SPA), Dr. Abdelhamid Elewa, Associate Professor of Linguistics and Translation at Imam Mohammad Ibn Saud Islamic University in Riyadh, says: "I think we will not have human translators in the future. In the future, the machine translation will overcome human translation." He notes that "If you compare the machine translation product in the past and today, you will find a huge difference; you will find a great pace has been taken in that dimension." Dr. Elewa further says: "20 years ago, there was Google Translation and if you use it then you will find that you have to do the job again. At that time, there was no awareness of the texts and proper names. It just transferred the meaning of words from a dictionary to another dictionary without looking at the context." Dr. Mohammed Albarakati, Associate Professor of Translation and Comparative Linguistics at King Abdulaziz University in Jeddah, Saudi Arabia; Mr. Nawwaf Albeedani, a Saudi Researcher of languages and translator; and Dr. Abrar Mujaddidi, Assistant Professor at the Department of Foreign Languages and Literature, King Abdulaziz University, do not see eye to eye with Dr. Elewa. Dr. Albarakati says, "I don't think machine translation is going to replace the translator." However, he adds, "I want to emphasize this very much: a translator who is good at machine translation is going to have a value-added aspect of his expertise, rather than those who are not coping with new technologies of translation." He notes that the new technologies in translation are going to be more efficient, faster, and more accurate in some texts because they are using memories. So, using machine translation is going to give translators who use it a bonus over others who don't use it. Dr. Albarakati, who is also the Chairman of the Board of Directors of the Saudi Translation Association, points out that "machine translation has actually made the life of those who use it easier. It has been a great help to them, but they need to master it properly, to keep up with new trends in it. Because this is like a technological product that is being produced by different companies, so they are still improving it, adding some aspects to it every now and then; the translators need to cope with the new advancements in machine translation." Mr. Albeedani, agrees with Dr. Albarakati. He adds: "I am one hundred percent with human translation." He notes that MT "will be very effective in translating very clear scientific text." However, Albeedani adds, "when it comes to literary texts, when it comes to articulated speech, and when it comes to text where we have deep cultural references it will fail drastically." Mr. Albeedani voices a strong opposition to the proposed idea that MT will entirely dominate and replace HT and says, "It (MT) will be an aid and it will not replace human translators at all." Dr. Abrar Mujaddidi supports Mr. Albeedani and says, "Machine translation will never take the place of human translators." However, she continues, "It does assist human translators. Human translators should embrace technology, not fear the use of technology. The technology improves the quality, speed and consistency of human translation." Dr. Mujaddidi adds: "There has been some worry or concern that technology or machine translation will take the place of humans but in my opinion and according to what data and researches showed, it will not; however, it will assist humans." CAT tools, which stands for computer-assisted translation, is one type of technology that can assist translators. However, Dr. Elewa adds, "Since 2007, different steps have been taken in developing machine translation into a more statistical form. Statistics are used as the main programming tool for Google Translation. So, we have got a statistics-based translation which doesn't pay attention only to the meaning of the word but also how words occur and how frequent a phrase is used in the target language." According to Dr. Elewa, "In 2016, a major change happened — people now use neural machine translation. Currently, machine translation is being developed into something close to human translation. If you use Google Translation today, you will find it very effective in some texts like media and general translation. You will only have to do some editing quickly and then you will find the product very well, not like five or six years ago." In April 2006, Google launched its translation service. Prof. Nike Kocijancic Pokorn, Professor of Translation Studies at the University of Ljubljana in Slovenia, agrees and says: "In the field of machine translation there is tremendous progress made lately with the introduction of neural networks." Prof. Pokorn adds, "The translations that are created by machines are now indeed comparable to what humans can do, especially for the languages that are more central. For example, between French and English this communication, machine translation, works quite fabulously." Prof. Abdulrahman Alswliaman, Associate Professor at KU Leuven in Belgium, says that MT "now is used in companies, between related languages, like for example, English and German. You can get administrative and economic translations with an accuracy rate of up to 80%, and a human translator must finish it." However, he adds, "We call it post editing. You must do post editing to finalize the translation, whereas the machine cannot do the work alone. You need a human translator to complete and finish the translation." Dr. Joseph Dichy, Professor of Arabic Linguistics in Lumière University Lyon 2 in France, notes that MT "In the future will be very important, even more and more important; but there is one essential element for MT which is that it needs to be helped by terminological databases." He adds, "On the other hand, machine translation can be very useful in order to have a low-cost translation but you need human translators to review what comes from the machine translation." Prof. Catherine Way, Associate Professor of Translation at the University of Granada in Spain, agrees that MT is improving. Nevertheless, she points out: "The problem is that each word in any language has many meanings and the machines could not understand the context yet. The creativity or the intuition of the meaning, the machines still cannot understand them. I think our job may change slightly, but I do not think we will lose our job." Prof. Way adds, "Machines still work with words and translation is not about words, it is about meaning." Prof. Martha Lucia Pulido Correa, Associate Professor at University of Antioquia and Visiting Professor at Federal University of Santa Catarina in Brazil, says that MT "is helping the translators a lot, but cannot do it alone." She explains, "For example, if a machine has to communicate a fear somebody is having, it won't do it because it cannot feel the fear." Prof. Correa emphasizes, "The relation between MT and HT is complementary. We are happy to have Google Translate; it helps a lot and it has replaced dictionaries in a sense." Dr. Kristijan Nikolic, Lecturer and Researcher at University of Zagreb in Croatia, says that "machine translation is going to be used and it is used already, but always there will be a need for a human who is going to assess the quality of that translation at least to verify that everything is correct. Machines can speed up the translation but they can't be the only solution and it will never be the final solution." Dr. Nikolic stresses that "Machines will never translate literary productions; poetry will never be translated by machine. What is going to be translated by machine more is legal translation because legal texts are sort of codified; they are very repetitive so machine translation can help. But at the very end, there will be a need for humans to succeed." Dr. Asma Alqunayir, Translation Studies Assistance Professor at Princess Nourah bint Abdulrahman University in Riyadh, Saudi Arabia, emphasizes the need to work with technology, adding, "We have so many translation tools to use. This is the reality of translation professional." She adds, "Will machine translation replace humans? I have no idea, but what I know is that we need to work with technology. Therefore, it is not machines versus humans. It's machines with humans." Dr. Alqunayir notes that "In China, the translation profession is not actually translation in the classical sense, where we have translators working on a source text and then they produce their target text, actually, they are editors." She explains, "They have translation institutions that hire editors. These institutions hire editors to use machine translation purely. Their job is to edit the machine translation. Therefore, in some parts of the world, they do not have translation services anymore. They have editing services over translated text." On her side, Prof. Pokorn says, "Still, there will always be professions and situations where a human translator will be needed. One of those situations is what we now see in hospitals, police stations, courts; one should not rely on machines. Literary translation and translation of religious and spiritual texts again cannot be done by machines because they are so late with cultural knowledge, cultural meanings, illusions and metaphors that the machine cannot really solve." Prof. Pokorn continues, "The machine translation is not created out of nothing. It feeds itself on human translation and for the fact that machine translation works well, you need humans to create translations and feed that to the system. The humans are always there creating machine translation and correcting it." In short, Elewa adds, "I am with machine translation. Actually, there is work around the clock to develop it, but the problem is in our products. If you know that the online English content is 56% of all languages, everything in English, while the Arabic content online is only 0.8%. Once we increase this percentage, neural machine translation will develop by default, because it is based on the content and how equivalent it is." He continues, "MT currently doesn't focus only on single words, now it is like a human being. If we find a big context here, it will be more significant to the machine to understand than the small content. If the percentage of the Arabic content increases and be in a comparable form, we will have no more human translators; when the Arabic context is close to 5%, we will only have editors." Dr. Joseph Dichy says, "Machine translation will not replace human translation. Suppose you have an international political conference, a head of state declares something, and if you have a machine translation, you can have a catastrophe at the other end. In between the speaker and the final result of the translation, the first step can be machine translation and then it has to be reviewed by a human expert. Because it's too difficult and dangerous to leave it to machine translation." Dr. Alqunayir concludes by saying, "I think, at least in the near future, humans will not be replaced, the need for humans will be there, as just I can't obviously, predict the future but I feel like we are indispensable, at least for the next 10 to 15 years."
