Abstract

The study examines the precision of human and machine translation. Discussing these topics is crucial at this moment. Technology is an integral aspect of our daily lives, yet it has also led to some concerning issues. This essay examines the areas where machine translations excel and those that still necessitate human involvement. There are certain translations that machines are unable to execute. There are many places where only humans can translate texts perfectly. This study utilises ChatGPT as a translation tool, providing it with three types of random text: newspaper paragraphs, idiomatic expressions, and poetic verses. The best translation technique and the translation process of an AI tool are evaluated by cross-checking and examining the provided result. Is it reliable? It also discusses the role of machine translation and its impact on human translators. The study selected English and Urdu as its languages of focus. The target language is English, and the source language is Urdu.

Key Words

Chatgpt, Machine Translation, Human Translators, Translation

Introduction

The world is developing by leaps and bounds in every field. As it is developing, it is bringing easiness and some difficulties as well. Every new day scientists and scholars in their relative fields invent, innovate and discover something. The easiness they bring cannot be denied; however, they also bring some concerning difficulties or challenges. Here, in this research paper, the focus is mainly on machine translation. The role of machine translation is discussed in the field of translation. Moreover, its potential risks or challenges are given a glance and some ways forward are considered. It is undeniably true that machine translation has made herculean tasks very easy: within a few minutes a person can translate a bulky book with the help of different machine translators, but the challenges are also worth considering. The questions that are given thought are: is the quality of machine translation better than that of human translators? What are the impacts of machine translation on the role of human translators? And is the authenticity, validity and quality of machine translation trustable?

The paper tries to answer these questions. It also discusses some areas where machine translation cannot be ignored, whereas there are some areas where human translators perform very well. When it comes to substitution of words or word-to-word translation, machine translation performs awfully well but it becomes full of flaws when word-to-word translation is not needed. It has impacted the role of human translators but in some specific areas. Human translators are still considered the best and most reliable. Human translators can easily translate the ambiguities, idiomatic expressions etc. from the source language to the target language without any confusion, whereas machine translation cannot be trusted when ambiguities and idiomatic expressions need to be translated.

In this paper, some previous researches in the very field are cited in order to get the answers to the proposed questions. Different researchers in the past have discussed the topic and the quality of machine translation. Some are of the opinion that machine translation is better than human translators while some others oppose the view. Some are of the opinion that machine translation should be used by the students while some deny it. Some ignore its impact on the role of human translators while some are concerned about it. This paper tries to address these questions and views about machine translation and human translators.

Background

In translation, we have two languages: source language and target language. Source language is the language that is translated and target language is the language into which source language is translated. We also have two types of audience in translation: source and target audience. The target audience is the native people of the target language. Translation is not an easy job. It needs competency in two languages. A person should know well both the source language and the target language. The translation is not just the substitution of words or rephrasing rather it has to convey both the meaning and sense. Sometimes, a translator has to find out the exact meaning of some culture-bound expression in order to translate it from the source language to the target language.  

According to Ghazala (1995), "translation is generally used to refer to all the process and methods used to convey the meaning of the source language into the target language". most of the definitions are almost the same translation is the conveying of the meaning of one language into another, which is in translation called source language or text and target language or text. One another similar definition of translation is given by Catford in 1995 and it says that translation is the replacement of equivalent textual material of one language by another language. The definitions are almost the same but the wording is different. (Catford,1995)

The definition is not a big deal; anyone can define translation. When it comes to the process of translation, it becomes difficult to accomplish the task. How a translation should be done? What is the right way of doing translation? And how a message should be conveyed?

In order to find out the answers to the very questions, many translators proposed different theories of translation. Some of the theories focus on word-to-word translation, some focus on sense-to-sense translation while others focus on the effect that translation should be done in a way which should not minimize or diminish the effect of the meaning on the people. It means that translation should be done in a way that it should not lose its effect while conveying the message from the source language to the target language.

Keeping in mind different translator's views about translation, it can be said that machine translation is only capable of word-to-word translation. It cannot translate the sense from SL to TL. Neither can it translate idiomatic expressions, ambiguities and culture-bound expressions.

This study focuses on the role of machine translation and its impact on human translators. In addition to that, it also focuses on the quality of machine translation. It is obvious, that machines have impacted human beings to a great extent. Machine translation has put at risk a lot of jobs related to translation. It has jeopardized the situation of joblessness. This cannot be denied, however; there are some situations or translations where machine translation cannot be trusted. The paper tries to explore those situations and tries to find out whether the quality and validity of machine translation is trustworthy or not.

Statement of the problem:

In this technology-driven world, it is difficult for one to not use machines. On the one hand, machines have brought easiness. They do great work in no time. On the other hand, they have brought some concerning problems. Machine translation has made our work very easy, however; it is important to find out that for what type of work we can use it and trust it. Machine translation cannot be trusted all the time. This paper tries to explore the areas or the type of text where machine translation can be used and trusted. 

Objectives of the study

objectives of the study are:

§   To find out the impacts of machine translation on human translators

To check the quality and validity of machine translation

Research question:

Is machine translation as good as the translation of a human?

Research methodology:

This study has adopted a qualitative research method. To answer the research question, some instant prompts are given to an AI tool, Chatgpt, and it is asked to translate it. The source language is Urdu and the target language is English. A few sentences of Urdu language are randomly chosen and entered to find out the quality of the translation. Some idiomatic and poetic verses are also translated with the help of Chatgpt

Literature Review

This part of the study focuses on some previous papers or studies done in the field of translation which focused on the role of machine translation and the quality of machine translation.

One of the studies explored that machine translation can be faulty and there are a lot of difficulties and problems which machines cannot overcome. The problem stated is related to the process of translation. It says that translation is not just word for word or substitution of words. It is not just the replacement of words from the source language to that of the target language rather it should convey the sense and true meaning that it gives in the source language. there are some culture-bound expressions which can only be translated by a human expert and machines cannot translate them. Humans go beyond the surface structure and try to analyse the core meaning of a text, whereas machines cannot do this job.

In translation, we have to keep in mind the context, ambiguities and idiomatic expressions. In this article, Saleh tried to translate some sample sentences by three machine translators which are: Tarjum Ajeeb, Alwati and Almisbar. He found out that these machines were doing great when only word-to-word translation was needed but when they entered some culture-bound expressions and some idiomatic expressions, they were not giving the desired result.  When a human translates a text, he/she considers all these things and tries to give the right sense. Machines cannot do that. The study concluded that machines can replace humans in the field of translation but in certain areas. Otherwise, translation is an art and humans can handle it perfectly well. (Saleh, 2000)

A study was done in the field of translation in order to find out the differences in translation by Popovic in 2020. This study was purposed to check the differences in translation. To complete the very task, the researcher took three distinct language pairs: English-Croatian, German-French and English-Finnish. The data for the study was taken from newspapers and two questions were focused on in the study: are there differences in human translation? And how does the native language direct the translation? The translations of humans and machines were then cross-checked in order to get the result. The researcher got an outstanding result and found out that there were some minor or ignorable differences in the human translation. Such as the lexical varieties and sentence length.            He also discussed that machine translation needs improvement. (Popovic,2020)

An article published in 2020, entitled, "Comparison of Google translation with human translation" focused on Google Chinese translation and human Chinese translation. The study was destined to find out the differences between the translation of a human and that of a machine. The translated language was Chinese. The researcher took the data from the selected work of Mao Zedong and while analysing it he kept in mind the formality and cohesion in translation. He took 289 excerpts from spoken and written works of Mao Zedong. All the data was collected from one person's work in order to avoid differences and difficulties.

The researcher found out that machine translation is almost similar to human translation at different levels especially, the semantic and pragmatic level. But when it comes to translating complex sentences, it cannot give us the result the way human translators give us. (Li, et al., 2020)

Machines do not have the ability to feel and think but humans do have. Translating a work does not mean putting synonyms and rephrasing a text. Translation is a difficult job and it can only be done by humans when it comes to translating some specific works. Machines cannot replace humans in the field of translation because humans are capable of translating the tones, intonations and emotions. This was the finding of a study done by Benmansour and Hdouch. (Benmansour & Hdouch, 2023)

These were some of the previous works done in the field of translation which mostly focus on the role and impact of machine translation on human translators. This study aims to discuss the quality of machine translation and its trustworthiness.

Analysis and discussion

As it was stated earlier the data will be taken and collected with the help of chatgpt. Since the study focuses on the quality of translation, therefore; chatgpt is chosen. It is said that it is used a lot these days. The source language is Urdu and the target language is English. Three types of input are given to chatgpt and it is asked to translate it.

Firstly, some random poetic verses were given to it. It is believed that when poetry is translated, it loses its charm and beauty but the study that is undergoing is purposed to focus on the quality. Therefore, poetic verses are good options.

Secondly, some random idiomatic expressions are given to it in order to check whether it gives us word-to-word translation or gives us the sense as well. And lastly, a paragraph is selected from a recent online newspaper. Three different types of input are given for the purpose of checking which one is translated the way translation should be done. This will help us focus on the areas which should be improved by human translators.

The first input that was given to it is as under:

        ?? ???? ??? ???? ???? ? ??

?? ?? ??? ??? ???? ??? ????

 

 (Main Jo Hoon Jhon Elia Hoon, 2016)

translate it into English

 

Chatgpt

"Why don't I get peace? There was only one person in my world."

A poetic verse was written and chatgpt was asked to translate it into English. The translation is acceptable but it needs improvement. It has done a word-to-word translation. The verse has lost its true value, its charm and beauty. The first line of the verse in translation is written as, "Why don't I get peace? If a human tries to translate it, the first thing he will focus on is the use of don't and can't. it should have been written can't instead of don't. 'don't' means that the person wants to get peace and can get it but he is not getting it while the verse in the source text is something else. It says that I want to get peace but I cannot get it and the reason is unknown.

The other line of the verse also cannot be called a good translation because the original verse asks questions, whereas the translated line states a kind of statement. It should be translated like this, was there only one person in the world?

After a poetic verse, chatgpt was given an idiomatic expression which is as under:

 

User

??? ?? ???? ?????

 (Offline Idioms & Phrases Dictionary, 2023)

 translate it into English

 

Chatgpt

"The beats of distant drums."

The translation that Chatgpt gives of this verse can be called the replacement of words in one language by another. It did not even do this job well. The original idiom in the source language means that a person always likes or loves something he has not seen or something he has just heard of, but when he sees the thing that he used to love or like, he does not find it the way he expected it to be. The translation given by the chatgpt of the idiomatic expression is not good. It cannot be accepted. It has just written the meaning of the words and it has lost the sense of the idiom.

The third input that is given to it is also an idiomatic expression which is given below.

 

User

 ????? ??? ???? ???? ?? ????

 (Offline Idioms & Phrases Dictionary, 2023)

translate it into English

 

Chatgpt

"A blind person doesn't know the value of spring."

This translation of the chatgpt is acceptable because it has not only transferred the true meaning of the text but also has given the sense of the original text in the target language. a problem can be seen in the choice of words. It has not chosen the right word for the right place. That problem that was discussed in the first idiomatic expression is done again here. It has written the word doesn't instead of cannot. It gives the meaning that the person can know the value of the spring but he does not want to know. In the original text, it is something else. It is in the original text that the person wants to know the value of the spring but he cannot know it because he is blind and to know the value of spring one must see the greenery around which cannot be seen by this person.

The third input given to chatgpt is as under: it is also a poetic verse.

 

User

??? ???? ?? ?? ????? ????? ??? ???? ?? ??? ? ?? ????? ?? ????? ????? ?? ?? ???? ??? ???? ?? ???

 

translate it into English

 

Chatgpt

"You are truly excellent; I write you as priceless. I leave behind these verses and ghazels; I write a book about you."

 

This translation by chatgpt cannot be accepted. It has neither done a word-to-word translation nor a sense-to-sense translation. It has destroyed the true meaning of the original text. The original text says that you are a special/perfect person and I to make you unique; forget about poems and sonnets, I want to write a book on you. The translation has given a slight touch of meaning to the original text but the words that are chosen are totally different.

After translating the poetic verses and some idiomatic expression, chatgpt was given a paragraph from the newspaper. The paragraph was selected randomly from a random newspaper. The paragraph and the translation by chatgpt are given as under:

 

????? ???? ?? ??? ?? ?? ?????? ??? ??? ?? ??? ???? ????? ?????? ???? ???? ?? ?????? ?? ??? ??? ??? ????? ?? ???? ?????? ??? ????? ?? ????? ??? ??

?? ?? ??? ??? ????? ???? ???? ?? ????? ???? ?? ?? ???? ????? ???? ?? ??????? ????? ???

????? ???? ???? ???? ?? ???? ???? ???? ????? ??? ??? ??? ????? ????? ?? ???? ????? ????? ?? ?????? ????? ?? ????? ???? ??? ??? ??? ?????? ??? ?? ???( Google News, January 15, 2024).

translate it into English

 

Chatgpt

Justice Ijaz ul Ahsan, one of the three senior judges of the Supreme Court, has resigned from his position and submitted his resignation to the President. Prior to this, President Dr Arif Alvi accepted the resignation of Justice Mazhar Naqvi, another judge of the Supreme Court. The Islamabad High Court has lifted the stay order on the trial court proceedings against former Prime Minister Imran Khan and Shah Mahmood Qureshi.

A fact cannot be denied. This translation by chatgpt is awesome. It has focused on every single word. It has translated not only the words by has also translated the core meaning of the original text. It has even written the names in a correct way. The point that the original paragraph was explaining is stated in a clear and concise way. it has gone beyond the surface level and has explained the core meaning.

As discussed in the literature review by some previous researchers machine translation can translate simple sentences and day-to-day talk in the right way, but it cannot translate complex and culture-bound expressions in the right way.

Conclusion:

In the above discussion, it can be found that machine translation or any translation that is done by something else and not by humans can be trusted in some specific areas. Such as; simple translation, word-to-word translation etc. The translation apart these should be double-checked and it is better to translate expressions etc by humans.

References

• Benmansour, M & Hdouch, Y, 2023. The role of the latest technologies in the translation industry. Emirati Journal of Education and Literature
• Catford, J. (1995) A linguistic Theory of Translation. London. Oxford University Press.
• Ghazala, H, (1995) Translation as problems and solutions (4th ed.) Syria: Dar Elkalem ElArabi.
• Google News. (2024, January 15). Justice Ijaz ul Ahsan has resigned.
• Hutchins, J, (2001). Machine Translation and Human Translation: In Competition or in Complementation? International Journal of Translation, vol.13, no.1-2, Jan- Dec 2001
• Li, H, et al., (2020) Comparison of Google Translation with Human Translation.
• OpenAI. (2022, January). ChatGPT
• Popovic, M. (2020) On the differences of translation.Retrieved from
• Saleh, M, A, S, (2000). The effectiveness of Machine Translation. International Journal of Arabic English studies