Integrated Journal for Research in Arts and Humanities

ISSN (Online): 2583-1712

Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

Investigating Machine Translation Errors in Rendering English

Literary Texts into Arabic
Wesam Mohsen Tahseen1 and Dr. Shifa'a Hadi Hussein2
1
Department of Translation, College of Arts, Tikrit University, IRAQ.
2
Professor, Department of Translation, College of Arts, Tikrit University, IRAQ.

1
Corresponding Author: [email protected]

www.ijrah.com || Vol. 4 No. 1 (2024): January Issue

Date of Submission: 07-01-2024 Date of Acceptance: 09-01-2024 Date of Publication: 18-01-2024

ABSTRACT

Machine translation is a machine that employs artificial intelligence (AI) to translate texts between languages without
human intervention. Machine translation approaches translate text or speech from one language to another, including the
contextual, idiomatic and pragmatic issues of both different languages. The present study aims to analyze the translation of
literary texts selected from different novels, plays, and poems and clarify the method for translating them from English into
Arabic. This study also aims to discover machine translation errors in rendering English literary texts and clarify the
translator's role in transferring the rhetorical impact on the reader who reads the (TT). This study hypothesizes that
translators(students) face difficulties regarding words and structures when translating literary texts from English into Arabic
because they misunderstand rhetorical devices. So they tend to use machine translations that translate literally, such as (Google
Translate, Reverso translation and Bing Microsoft translation). This study adopted two models: First, Newmark's translation
model (1988b), which includes two basic types of translation: semantic and communicative. This model is used widely in the
analysis of literary texts. Second, Harris (2018) linguistic model theory of rhetorical question and the general purpose of the
rhetorical devices to analyze the data. Finally, the study ends with the conclusions that all machine translation programs (Google
Translate (GT), Reverso Translation (Reverso. T), Bing Microsoft Translation (Bing. M.T) in rendering English literary texts
from English into Arabic are unacceptable and have more problems because these programs are just machines and cannot think
or feel as well as all these machines renderings are meaningless and ambiguous. So Human translation is better than Machine
Translation because the first uses communicative translation while the other uses semantic translation.

Keywords- Translation, Machine Translation, Machine Translation Problems.

I. INTRODUCTION languages.MT is one of the things that Natural Language

Processing can be used for (NLP). It is also called
Praise be to Allah the Lord of the worlds and Automatic Translation. This means that a computer can
may the blessings and peace of Allah be upon the most be used to translate from the source language (SL) into
honored of messengers our master Muhammad and upon the target language (TL).
all his family and companion. Arabic is one of the languages that has been
Translation is not easy because languages and worked on since the beginning of machine translation.
cultures are different. As long as computers have been However, only some systems have worked with Arabic
around, people have thought about using them to because its syntax differs from Latin. Arabic is the first
translate languages or help translate them. language of more than 300 million people. Because
Machine translation (MT) is a computer science Arabic is a "non configurational" language, many (NLP)
and applied linguistics field that translates human applications have trouble with it.

68 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

II. TRANSLATION 2.3.1.1. Intranlingual translation or “ rewording”:

This kind refers to the interpretation of the verbal signs
Due to the importance of translation in in one language by a set of verbal signs in the same
linguistics, many researchers studied it. There are many language. It means (within one language), e.g.
different linguists' definitions of translation. restatement or paraphrase from one lyrical shape into
Etymologically, “translation” means “bringing another. So, it means when you use different words to
across” or “ carrying across.” The Latin term “ translatio mean the same thing in the same language, paraphrase,
” is derived from )“ trans,” “ across,” and “ ferre,” “ to or change an idiom like “pass away” to “die”.
carry,” or “ to bring ”). In general, the modern Romance, 2.3.1.2. Interlingual translation: This type describes
Germanic, and Slavic European languages have the verbal signs by means of some other language. It
developed their terms for this concept based on the Latin means (between two languages) from one language into
term “ transfer ” or the related term “ traducer ” (“to another . It means when some code units in the(SL) are
bring across” or “to lead across”). In addition, the Greek changed to code units that are the same in the (TL).
word for “translation” is “ metaphrasis,” which means 2.3.1.3. Intersemiotic translation: This kind of
“speaking across” (Kasparek: 1983: 83-84). translation means interpreting certain verbal signs by
Catford (1965:20) , Bassnett (1991:25) , and means of signs of nonverbal sign system ”. It means
House (2015:63) describe translation as the substitution (between sign systems) one system of signs into another,
of a text in the source language with a semantically and e.g. from verbal art to music, dance, cinema or painting.
pragmatically equivalent text in the target language. It means the use of signs or signals to communicate.
Also, translation is transmitting a linguistic idea Human language is the most crucial semiotic system
into another language. It is a method that helps convert compared to sign language and traffic lights. This type
written content in one language into another. (Pinchuck, fits into Jakobson's framework, which says that
1977:9; Wills, 1982; Newmark, 1991). translation is changing one sign into another alternative
2.1 Translation Principles or equal sign, whether spoken or not. (Jakobson
Translation principles are great use translators (1959:232 cited in Schulte and Biguenet, 1992:145);
and give much help to them. According to Mc. Quire Shuttleworth and Cowie, 2007 : 85).
(1980: 3) translation is governed by general principles: 2.3.2. Translation Types according to Mode: (Written
1) Translation must present a perfect transfer of ideas vs. Oral)
conveyed in the original text. Nida and Taber (1969: 12) define interpreting
2) The style and manner of translation must be as the reproduction of the closest equivalent of the (SL)
comparable to the original text. message in the (TL). This is a common ground or
3) The translator should have perfect knowledge of interface between translating and interpreting. The
both languages. former is not mainly or only concerned with the correct
4) The translator should generally utilize a particular semantic transference. The translated text should, at least
mode of speaking. in theory and in an ideal world, have the same meaning,
5) The translator must avoid literal translation. correct grammar, effective style, and textual cohesion as
6) The translator must carefully select and organize his the source text.
or her words and ideas.
III. MACHINE TRANSLATION
2.2. Translation Steps (Process)
The translator must follow specific steps to Machine Translation has many definitions.
ensure the translation is correct and complete. Barihi According to Garvin (1963:223) and Hutchins (1985: 1),
(2020:19-20) says that there are seven steps to machine translation (MT) is the process of translating
translation: any text from one language (SL) to another (TL).
a) Establishing the project. Lawson (1982:5) and (1989:280) defines (MT)
b) The Explanation. as an automatic translation with or without human
c) Transfer and Draft One. intervention. Also, Sippl (1985: 359) defines (MT) as the
d) Evaluation. automatic transmission of information from one
e) Revised Draft. representation to another. The translation may involve
f) Consultations. codes, language, or other systems of representation.
g) Final Draft. House (2018:18) says that the term "machine
2.3 Types of Translation translation" refers to computers doing some or all of the
Translation are classified into many types: translating. Professional translators and academics have
2.3.1. Translation Types according to Code: recently come to accept (MT) systems because these
The Swedish- American structuralist Roman systems are constantly improving in consistency and
Jacobson(1959/2004 as cited in Munday, 2001: 5) effectiveness, and more high-quality translation becomes
classifies translation into three types: available, posing a risk to human translators (Way, 2018;
Bowker, 2019; Vieira & Alonso, 2020).

69 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

3.1. Types of Machine Translation languages presently supported by Google Translate. It

According to Nichole (2018: 25-27) , there are presents word-level alternatives if a translation seems
four types of machine translation: incorrect. This technique is also utilized as feedback to
3.1.1. Statistical Machine Translation (SMT): SMT is improve the quality of Google Translate's machine
the first type of machine translation. It works by translation(Google Translation, 2013: 1).
referring to statistical models based on the analysis of
the bilingual text. It tries to determine if a word in the
source language is the same as a word in the target
language. Google Translate (GT) is an excellent example
of this. SMT works well for simple translations, but its
biggest flaw is not considering the context. This means
that translations often need to be corrected. In other
words, it is expected that translations is of low quality.
3.1.2. Rule-Based Machine Translation (RBMT):
RBMT is the second type of machine translation. It 3.2.2. ATLAS: ATLAS is one of the oldest computer
works by following the rules of grammar. It conducts a programs that can translate from and into Arabic. It was
grammatical analysis of the source language and the made by the Hong Kong-based FTC (First Trading
target language to generate the translated sentence. But Company). In the Arab world, ATLAS electronic pocket
RBMT needs much proofreading, and its heavy reliance dictionaries are more popular than other electronic
on lexicons means that it does not work well until a long dictionaries that translate from and into Arabic. There
time has passed . have been many different kinds of ATLAS translators,
3.1.3. Hybrid Machine Translation (HMT) : HMT is like the SM series and the SD series. ATLAS 's most
the third type of machine translation. It is a mix of recent version is the ATLAS Modern Dictionary, found
RBMT and SMT, which is clear from the name. It uses a online and in ATLAS Dictionary L519. The company
translation memory, which makes it a lot better at what it has made many different versions of the program in a
does. However, even HMT has some problems, the paper, online, and electronic forms (Atlas, 2013).
biggest of which is that it needs much editing. There is a
need to have human translators.
3.1.4. Neural Machine Translation (NMT): NMT is
the fourth kind of machine translation. NMT depends on
neural network models based on how the human brain
works. The main benefit of NMT is that it uses a single
system to decipher the source and target text. So, it does
not rely on specialized systems like SMT, which are
used by other machine translation systems .
3.2.3 Babylon: Babylon is a program which contributes
3.2 Programs of Machine Translation to machine translation. It translates from and into Arabic
There are many programs that are considered as
by using a machine. Dictionaries and Encyclopedias
Machine translation systems, as the following:
(2013) says that Babylon (Ltd) has made 36 English-
3.2.1. Google Translate (GT) : GT is one of the most
based proprietary dictionaries in 21 languages, including
popular online translation tools today. Och (2006:3) says
Arabic, and has given them away for free to software
that Google has put its Arabic-English and English-
users. These dictionaries have 60,000 and 200,000 terms,
Arabic systems online. He says, Arabic is a very hard phrases, acronyms, and abbreviations. They also have a
language to translate to and from because it has a very
morphological engine that makes it easy to recognize all
rich morphology and requires a long-distance reordering
forms of single words and phrases that have changed
of words. Many people have used Google Translate
over time, give all forms of terms with prefixes and
lately, including academics, students, people just starting
suffixes, and offer a solution for all types of writing.
to learn how to translate, professional translators, and so (Wikipedia,2023)
on.
Google made Google Translate (GT) in 2007.
Also, Google Translate used Systran at first, but in
October 2007, Google switched from Systran to its
machine translation system for all 25 language pairs on
the site. It has used its machine translation system in
Arabic, Chinese, and Russian (Chitu & Schwartz, 2007;
Korosec, 2011).
Google Translate is a free translation service 3.2.4. Microsoft Translation: Microsoft translation
that gives instant translations between dozens of includes programs such as (Microsoft Translator, Bing
different languages. Arabic is one of the seventy one translator, and Reverso translator). These programs

70 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

consider the modern programs people use to translate have a list of categories that cannot be found next to
texts from English into Arabic. The translation results each other (Bourbeuea ,1988:67).
for these programs are the same regarding accuracy and (4) Ahmed sprigs by his hands
errors (www.microsoft.com). This sentence is translated by human as :
In relation to this, Handschuh (2013) studied ‫ يَ ْغ ِرس أَحْ َمد ِبيَدَيْه‬.
German-English translation using four online (MT) Also by (MT) as : ‫أَحْ َمد يُ ْنتَزَ عُ مِ ْن يَدَ ْي ِه‬
systems: GT, SYSTRAN, Bing, and Babylon. He found
that MT's output was primarily wrong and inaccurate. So, machine translation cannot give the real
Keshavarz (1999) said that the errors were lexico- meaning of the verb “sprigs”
semantic errors, like using the wrong tense, verb group,
word order, use of prepositions, and use of active and
passive voice, and errors with the use of articles.

3.3.3. Homographs: Once the grammatical category is

known, the homograph can be separated based on the
type of the text as long as the homograph set is made up
of words from different grammatical categories. In this
3.3 Problems of Machine Translation way, Nagao (1989:24) says that a sentence like "Time
Machine translation has problems. The flies like an arrow" could mean more than one thing. The
following problems are some of which: word "flies" could be a third-person singular form of the
3.3.1. Semantic Ambiguity : Semantic ambiguity is verb "to fly" or a plural form of the noun "fly." The word
thought to be more difficult than syntactic ambiguity "like," on the other hand, can be a conjunction or a verb
because only a human's intelligence can translate it. For that means "to love."
example: The following example shows how important
(1) He went to the bank it is to identify the type of grammar: When these
In the previous sentence, it is not clear whether sentences are translated into Arabic, the word "use"
the word "bank" means the "mound of sand" or the could mean more than one thing(S. Riyad ,2013:350).
"financial institution”. The choice can only be made in (5) They use their cars;
this sentence based on the situation. This is what the (6) The use of abrasive cleaners on the printer casing is
machine thinks this sentence means: ِ‫َب ِإلَى ْال َمص ِْرف‬
َ ‫( ذَه‬see not recommended.
Palmer, 1980:106; Shaalan et al., 2004: 4). "Use" is a verb in (5), but it is a noun in (6).
However, consider the following: An English- Arabic dictionary gives ُ)‫ ( يُ ْستَ ْع َمل‬for the
(2) He went to the bank and sat down under a tree. verb and) ‫ ( ا ْستِ ْع َما ُل‬for the noun. So, it is up to the reader
(3) He saved money by going to the bank. or an automatic parser to determine whether a verb or
can be easily translated into: noun should be translated when it appears in a sentence
based on how the source language's grammar works .
another example of the verb “ use ” :
(7) They use their pens
(8) They use Ahmed in their problem
Here, human translation is : , ‫( يَ ْست َْخ ِد ُمون أَ ْق ََل َم ُهم‬
) ‫يَ ْستَعِينُون بِأَحْ َمد فِي مشَا ِك ِل ِهم‬, But, MT can’t give the real
meaning of the verb “use” in (8) .

However, the problem is that when you type (2)

and (3) into the machine translation (reverse.t), you only
see the word "‫ش َج َرة‬ َ َ‫س تَحْ ت‬ َ َ‫َب ِإلَى ْال َمص ِْرفِ َو َجل‬
َ ‫ "ذَه‬and ‫لَقَد َوفَ ّر‬
" ِ‫ب إلَى ْال َمص ِْرف‬
ِ ‫ " ْال َمال بِالذَّهَا‬.This means that even if there
is a context, the computer cannot tell the difference
between the two meanings of (2) and (3). 3.3.4. Problems of Translating Verb Phrase in MT:
3.3.2. Syntactic Ambiguity: Syntactic ambiguity Even when there is a context, it is hard for the machine
happens when a word with a particular form can be to figure out what a particular word means. The
understood differently. For instance, "sprigs" can be following examples of the verb "hold" show how it can
considered a plural noun or a verb in the simple present be used differently:
(third-person singular). Machine translation (G.T) will
71 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

(9) He holds silence.

(10) He holds silence.
(11) He holds a theory.
The machine (reverso .t ) turns them into: (
َّ ‫يُحْ َم ُل ال‬, ‫) يُحْ َم ُل بَ ْعض ْال ُكتُب‬
َ َ‫يُحْ َم ُل ن‬, ‫ص ْمت‬
‫ظريَّة‬

3.3.6. Idioms : Idioms are hard for computers to

translate because machine translationis mainly based on
finding the best word for each word in the source text
and then using rules to translate this word into the target
But the word "hold" does not mean the same language. This method only works for idioms where the
thing in all three sentences, so it should not be translated literal translation of the parts of the idiom sometimes
the same way. If you want to translate(9), (10) and (11), shows what the idiom means. Some English idioms have
you should use (‫ب‬ ِ ُ ‫ض ْال ُكت‬
ِ ‫يُحْ َمل بَ ْع‬,‫ت‬ َّ ‫ يَ ْلت َِز ُم ال‬,‫ظ ِريَّة‬
ِ ‫ص ْم‬ َ َّ‫) يُ َؤ ِ ّم ُن ِبالن‬ an Arabic proverb that means the same thing(Meryem,
(see Ba'lbakki, 1990:430) . 2010:13-15) . For example :
3.3.5. Lexical Ambiguity: This happens when the (14) Out of sight, out of mind ‫ب‬ ِ ‫ع ْن ْالقَ ْل‬
َ ٌ‫ع ْن ْال َعي ِْن َبعِيد‬
َ ‫ ْال َبعِي ِد‬.
machine cannot determine the right word to replace a (15) Forbidden fruit is sweet ‫غوب‬ ُ ْ‫ ُك ِّل َم ْمنُوع َمر‬.
term or word used in the source text. Since machines
cannot understand the "meaning" of what they are
translating, they cannot choose the correct equivalent
independently.
Balkan et al. (1994:105) discuss this problem
by saying, “A word is lexically ambiguous when it can
mean more than one thing”. For instance: The word
"spring" refers to (A season of the year), (A coiled wire 3.4. Machine Translation V.S Human Translation
object) and (A natural source of water) for example: Machine translation(MT) and human
(12) spring is a nice season. translation(HT) are important translations of languages,
(13) Water at the bottom of the spring especially in the literary aspect. According to Sijeta
Here, spring is a season in (12) but spring is a (2016: 6-7-8),there is a difference between machine
natural source of water in (13) translation (MT) and human translation (HT) as the
َّ ,. ‫) ْال َماءِ فِي قَاع ْالبِئْر‬, but machine (GT)
ْ َ‫الربِيع ف‬
( ‫ص ٌل َجمِ يل‬ following :
translates the sentences as: ( ‫ص ٌل‬ ْ َ‫الربِيع ف‬
َّ ,. ‫الربِيع‬ َّ ‫ْال َماءِ فِي قَاع‬ 3.4.1. Machine Translation (MT)
‫ ) َجمِ يل‬.

Machine translation is the application of computers in translating texts from one

language into another .A computer translates text without the help of a person.
The Conception
This is called "machine translation." It is also called automated translation
(Hutchins ,1986:15).
Some of the advantages of using (MT) are:
First : It is free, which is an advantage in business that everyone likes. However,
if the translator uses it through an (API), (He\she) has to pay a small fee.
Second: It works right away and is online, making it easy to use because
translations are done immediately and can be done anywhere at any time.
The Advantages
Third : Page and text translation are easy and quick, no matter how long or
complicated the page or text is. This includes web pages, emails, documents,
MT
and more.
Fourth :There are many different languages available and ready to be translated
(Saba, 2015)..
Using (MT) has several disadvantages :
First : It is limited, so it cannot translate everything and limits what could be
translated.
Second: These services offer many languages, but quite a few still need to be
The Disadvantages
included.
Third : You cannot expect natural, easy-to-read translations because the
services are mechanical and automated and only do direct, mechanical
translations of words.
72 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

Fourth : Context is not part of the equation, meaning that (MT) will translate
the text as a text without considering the context of the content. It does not
consider who the text is for or why it was written (Saba, 2015).

3.4.2. Human Translation (HT)

Humans do the process of translating, which is called "human translation." It is

still the best way to translate since it is a complicated process that is not a
scientific calculation or a mathematical equation that could be programmed into
The Conception
an intelligent machine. The translator will turn the original text into a version
that stays true to the original's spirit and meaning and has the right tone and style
for the person reading it (Day Translations, 2015).
Human translations outweigh their drawbacks. First, consider language's
complexities. Human language includes grammar, syntax, connotations, senses,
and the speakers'/writers' cultural background. Translators must locate the target
material's correct terminology, grammar, and culture. There is no standard
translation format, so the translator must be trained and innovative. As the
HT
The Advantages context is vital, human translators are suggested for formal correspondence. To
express a brand's identity and message, a human must interpret web material,
marketing text ,etc (Fletcher, 2016).
Also, Humans are needed to make sure that a translation makes sense
in terms of grammar, slang, and subtleties of language so that the tone and
meaning of the original text are kept in the translation (Fletcher, 2016).
The only real problem with using a human translator is when someone needs to
understand or communicate the main idea of something in a non-official letter.
The Disadvantages In these situations, the translators want to save money and time on something
other than a human translator and do not want to spend the money or time on it
(Saba, 2015).

IV. DATA ANALYSIS

SL Text ( 1 ) :
“ In the teeth of every difficulty ”.
(Orwell,1954: 63)
TL Texts ( 1 ) :
‫صعُوبَة‬ ُ ‫ان ُك ّل‬ ِ َ‫سن‬ ْ َ‫فِي أ‬ 1
‫صعُوبَة‬ ُ ‫ان ُك ّل‬ ِ ‫سن‬َ ْ َ‫فِي أ‬ 2
‫صعُوبَات الَّتِي واجهتها‬ ُّ ‫الر ْغ ِم مِ نْ ال‬
َّ ‫علَى‬
َ 3
‫صعُوبَة‬ ُ ‫ان ُك ّل‬ ِ َ‫سن‬ ْ َ‫فِي أ‬ 4
‫الر ْغ ِم مِ نْ جَمِ يعِ ا ْل َم َخاطِ ر والصعوبات‬
َّ ِ‫ب‬ 5

Discussion: that the animals rebuild the windmill despite the

“In the teeth of every difficulty ” is a literary obstacles they have encountered and their lack of
text taken from the novel (Animal Farm) in (1954). The experience. Thus, Orwell compared the obstacles to
interpretation of this expression is that Orwell intended sharp teeth.

Table ( 1 ) : Text Analysis No. ( 1 )

Method of Translation Problem Of Translation

Appropriateness
Translator No.
SL Text NO.

Communicati
Type

Syntactical

Rhetorical
Semantic

Stylistic
Lexical
ve

73 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

1 + - + - + + -

2 + - + - + + -
Novel
One

3 - + - - - - +

4 + - + - + + -

5 - + - - - - +

The translators (1),(2), and (4) fail in writer wants to describe it as a metaphor that has
rendering this expression because they rely on (MT) outlived its utility, but (MT) cannot comprehend the
(Reverso.T), which causes numerous problems in its word's intended meaning (teeth), which is that the
translation, including the following: First, there is a difficulties have sharp teeth. Therefore, it translates it
lexical problem because (MT) gives the literal meaning semantically , which focuses on the structure, not the
of the word (teeth \ ‫)أَ ْسنَان‬. Second, there is a rhetorical content of the (SLT). Third, stylistic problem because of
problem because this expression is a metaphor; the the style which is meaningless.

Only the translators (3) and (5) succeed in " ‫" برغم جَمِ يع ا ْل َمصَاعِب‬
their renderings of the (SLT) because they transfer the (Mohamed, 2013:13)
metaphor's meaning and sense. They translate the
metaphor “ teeth of every difficulty” as the difficulties of
all the obstacles that the animals have faced. Also, they
translate it communicatively in order to get an Also, there is a difference between the two renderings of
acceptable translation like the proposed translation by the (SLT) as the following table shows:
Adel Mohamed Ahmed :

Table ( 2 ) : (The Difference Of Renderings) Text No. ( 1 )

The The The The The
Item
Translator Procedure Model Rendering Reason
In appropriate
Reverso.T

Semantic

Word for word MT

MT translation cannot comprehend the word's intended
meaning.
Communicative
Adel Mohamed

Appropriate
Ahmed

The translator
metaphor
HT transfers the metaphor's meaning and
to metaphor
sense.

74 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

SL Text ( 2 ) :
“ Flavius: Hence! Home, you idle creature, get you home.”
(Shakespeare. W ,Julius Caesar , 1623, Act I ,Scene I, Line I)
TL Texts ( 2 ) :

‫ أوصلك إلَى ا ْل َم ْن ِز ِل‬، ‫ أَ ُّيهَا ا ْل َم ْخلُوق ا ْلعَاطِ ل‬، ‫ثَ َّم ! ا ْل َم ْن ِزل‬ ْ‫ َومِ ن‬: ‫فالفيوس‬ 1
‫ أوصلك إلَى ا ْل َم ْن ِز ِل‬، ‫ أَيُّهَا ا ْل َم ْخلُوق ا ْلعَاطِ ل‬، ‫ثَ َّم ! ا ْل َم ْن ِزل‬ ْ‫ َومِ ن‬: ‫فالفيوس‬ 2
‫ أوصلك إلَى ا ْل َم ْن ِز ِل‬، ‫ أَيُّهَا ا ْل َم ْخلُوق ا ْلعَاطِ ل‬، ‫ثَ َّم ! ا ْل َم ْن ِزل‬ ْ‫ َومِ ن‬: ‫فالفيوس‬ 3
‫ أوصلك إلَى ا ْل َم ْن ِز ِل‬، ‫ أَ ُّيهَا ا ْل َم ْخلُوق ا ْلعَاطِ ل‬، ‫ثَ َّم ! ا ْل َم ْن ِزل‬ ْ‫ َومِ ن‬: ‫فالفيوس‬ 4
‫ أوصلك إلَى ا ْل َم ْن ِز ِل‬، ‫ أَيُّهَا ا ْل َم ْخلُوق ا ْلعَاطِ ل‬، ‫ثَ َّم ! ا ْل َم ْن ِزل‬ ْ‫ َومِ ن‬: ‫فالفيوس‬ 5

Discussion:
The (ST) above is the an expression taken from Shakespeare's tragedy (Julius Caesar) in (1623). The
interpretation of this expression is that Flavius, whose duty it is to defend the rights of common people against nobles, is
angry that the working people have taken a holiday to welcome Caesar home after he triumphed over Pompey's two sons.

Table ( 3 ) : Text Analysis No. ( 2 )

Method of Translation Problem Of Translation

Appropriateness
Translator No.
SL Text NO.

Communicative

Syntactical

Rhetorical
Type

Semantic

Stylistic
Lexical

1 + - + + + + -

2 + - + + + + -
Drama
Two

3 + - + + + + -

4 + - + + + + -

5 + - + + + + -

In rendering this expression, all the translators as well as the meaning is not clear. Third, rhetorical
failed in their renderings because they used (GT), which problem because (MT) cannot understand how to use the
causes problems: First, Lexical problem because it gives figurative language of compensation. it is just a tool; it
the literal meaning of the words (idle creature\ ‫ْال َم ْخلُوق‬ has no feelings or sense. Fourth, stylistic problem
ْ which is not related to the (ST). Second,
‫)ال َعاطِ ل‬ because the style of (MT) is ambiguous as well as in
Syntactical problem because word orders are not rendering this expression, compensation by merging is
acceptable, the structure of the sentence is not accurate needed to express the idea to the (TL) readers.

75 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

In this instance, to obtain a correct translation, not convey the negative connotation loaded in the (SL)
the translator must use the compensation procedure by expression, and compensation is necessary here.
merging two lexical items (idle creature) into one َ ‫ تَفَ َّرقُوا إلَى بُيُو ِتكُم أَ ُّيهَا ا ْل ُك‬: ‫" فالفيوس‬
"‫سالَى عُودُوا إلَى ِدي َِاركُم‬
(‫سالَى‬ ْ which preserves the aesthetic function of the
َ ‫)ال ُك‬, (Amin, 1998)
source language term. So, in the proposed translation
below, the translator conveys the negative connotation of Here ,there is a clear deference between these two
the (SL) expression. In contrast, the literal translation of renderings as the following table:
this text would be an awkward rendition because it does

Table ( 4 ) : (The Difference Of Renderings) Text No. ( 2 )

The The The The The
Item
Translator Procedure Model Rendering Reason

In appropriate
Semantic
Word for word MT cannot understand how to use the
GT

MT translation figurative language of compensation

because it is just a tool.
Communicative

Appropriate
The translator preserves the aesthetic
Amin

Compensation by
HT function of the (ST) by using
merging
compensation.

SL Text ( 4 ) :
“ Twit twittwit
Jug jugjugjugjugjug
So rudely forc’d.
Tereu.”
(The Waste Land (The Fire Sermon), 1922, Section III, lines(202-206),p: 30)
TL Text ( 4 ) :
‫ت َُويْت تويتويت‬ ‫ت َُويْت تويتويت‬
‫إب ِْريق جوج‬ ‫إب ِْريق جوج‬
2 1
‫قَس ًْرا بوقاحة‬ ‫قَس ًْرا بوقاحة‬
‫تيريو‬ ‫تيريو‬
‫ت َُويْت تويتويت‬ ‫ت َُويْت تويتويت‬
‫إب ِْريق جوج‬ ‫إب ِْريق جوج‬
4 3
‫قَس ًْرا بوقاحة‬ ‫قَس ًْرا بوقاحة‬
‫تيريو‬ ‫تيريو‬
‫ت َُويْت تويتويت‬
‫إب ِْريق جوج‬
5
‫قَس ًْرا بوقاحة‬
‫تيريو‬

Discussion: poem, some relate them to an allusive reference to the

This short excerpt above is an expression story of Philomela, -a Greek mythical figure who is
taken from the poem (The Waste Land), the third section violently raped and whose tongue is cut by her brother-
of (The Fire Sermon) by Eliot in (1922) . The in-law (King Tereu) contending that the words 'twit' and
interpretation of this excerpt is that Eliot's lines "Twit 'jug' are onomatopoeic of the sound of the nightingale to
twit twit / Jug jugjugjugjugjugjug" are nothing more than which Philomela is transformed as a means of liberation
"random noises." Though neither of the neologisms from the injustice inflicted upon her.
seems to have a meaning concerning this part of the

76 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

Table ( 5 ) : Text Analysis No. ( 3 )

Method of Translation Problem Of Translation

Appropriateness
Translator No.
SL Text NO.

Communicative

Syntactical

Rhetorical
Type

Semantic

Stylistic
Lexical
1 + - + + - + -

2 + - + + - + -
Poetry
Three

3 + - + + - + -

4 + - + + - + -

5 + - + + - + -

All the translators failed in their renderings because the sentence structure is inaccurate, and the
because they depended on (Bing. M.T ), which caused word order is unacceptable. Third, Stylistic problem
problems: First, Lexical problem because translates the because( MT) deserts the poetic style in its rendering,
following words literally ( twit \ ‫ ت ُ َويْت‬, jug \ ‫ إب ِْريق‬, rudely there is no coherence between the sentences, and it has
forced \ ‫ )قَس ًْرا بوقاحة‬which are word for word translation no idea about how to use neologism in phonemic
and it is mistranslation. Second, syntactical problem translation. So, its rendering is meaningless.

As can be seen in the translations below, the ‫ق‬

ّ ‫ش‬ِ ‫ق‬ّ ‫ش‬ ِ ‫ق‬ ّ ‫ش‬ ِ
translator Lulu opts for 'phonemic translation,' which ‫ق‬
ّ ‫ق ِز‬
ّ ‫ق ِز‬ّ ‫ق ِز‬ ّ ‫ق ِز‬ ّ ‫ق ِز‬ّ ‫ِز‬
reproduces the (ST) language sounds into the (TT), thus ‫شيَّة اغت ُصبت‬ ِ ْ‫ِبغَايَة ا ْل َوح‬
subscribing to an'allusive 'interpretation of the ‫تيريو‬
neologisms rendered as " ‫ " شق شق‬and "‫" زق زق‬, (Lulu , 1995:44)
respectively, deriving his translation from the quadruple
Arabic verb "‫ " شقشق‬meaning " ‫ص ْوت العصفور وزقزق نحوه‬ َ
". Finally ,the differences between the renderings of (HT)
and (MT) are shown in the following table :

77 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

Table ( 6 ) : (The Difference Of Renderings) Text No. (3 )

The The The The The
Item
Translator Procedure Model Rendering Reason

In appropriate
MT cannot understand the(ST), it
Bing.M.T

Semantic
Word for word deserts the poetic style phonemic
MT translation translation
(neologism), it translates word for
word translation.

Communicative
The translator understands

Appropriate
the(ST); he uses the poetic style
Lulu

Phonemic translation
HT Phonemic translation
(neologism)
(neologism), his rendering is
meaningful.

V. FINDINGS OF THE STUDY 1. Communicative translation is more accurate than

semantic translation for this linguistic phenomenon
The present analysis study concludes with the because semantic translation distorts the meaning of the
following findings: (SL) texts. Therefore, this study validates this issue by
contrasting the communicative and semantic translation
percentages in the table below.

Table ( 7 )
Numbers and Percentages of
(Translators, Communicative translation and Semantic translation)

Renderi
Percent

Percent

Total
Communicative Semantic
age

age

ngs
Translators NO.
Translation Translation

1 1 0.8% 24 19.2% 25
2 1 0.8% 24 19.2% 25
3 3 2.4% 22 17.6% 25
4 1 0.8% 24 19.2% 25
5 2 1.6% 23 18.4% 25
Total 8 6.4% 117 93.6% 125
Final Renderings ( 8 + 117) = 125
Total Percentage ( 6.4% + 93.6% ) = 100%

2. The renderings of the translators (students ) are renderings as the translators(students) fail to transfer
classified into two types: First, appropriate renderings literary texts' impact on the (TL) reader by adopting the
because the students succeed in producing the same semantic translation. The following table illustrates that
impact on the (TL) reader in their renderings by adopting clearly:
the communicative translation; Second, inappropriate

Table ( 8 )
Percentage of Appropriate and Inappropriate Renderings in this study
literary texts

Percentage

Percentage

Percentage
renderings

Number
Texts No.
Type of
English

Number
Total

Of
Of Appropriate
Inappropriate
renderings
renderings

78 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

Novel 12 8 13% 52 87% 60 48%

Drama 10 0 0% 50 100% 50 40%
Poetry 3 0 0% 15 100% 15 12%
Total 25 8 6.4% 117 93.6% 125 100%

3. The translation of English literary texts requires a vocabulary(lexical words The following table indicates
high level of linguistic proficiency. The translator must the percentage of each translator in rendering English
proficiently employ the proper syntax (structure)and literary text into Arabic.

Table ( 9 )
Numbers and Percentages of accurate and inaccurate Translators
Inappropriate

Renderi
Percent

Percent
Appropriate renderings

Total
renderings

age

age

ngs
Translators No. No.
No.

1 1 0.8% 24 19.2% 25
2 1 0.8% 24 19.2% 25
3 3 2.4% 22 17.6% 25
4 1 0.8% 24 19.2% 25
5 2 1.6% 23 18.4% 25
Total 8 6.4% 117 93.6% 125
Final Renderings ( 8 + 117) = 125
Total Percentage ( 6.4% + 93.6% ) = 100%

4. Many translators (students) rely on machine many problems. The following table shows the types,
translation in their renderings as checked by the numbers and percentages of each machine used in this
researcher himself practically one by one , which causes study.

Table ( 10 )
(Types, Numbers and Percentages) of Errors by Machine Translation Programs that are used in this Study
Types
Types of Literary

Renderings by all
Total renderings

of
Classification of
Translators No.

Renderings by
Translators

MT
Percentage

Percentage
Total S.T
(ST)No.
Texts

(MT)
(ST)

Reverso.T

Bing.M.T
GT

Novel 12 7 4 1 27 20 5 60 8 6.4% 52 41.6%

5 Drama 10 5 4 1 25 20 5 50 0 0.0% 50 40%
Poetry 3 1 1 1 5 5 5 15 0 0.0% 15 12%
Total
57 45 15
5 3 25 125 8 6.4% 117 93.6%
117
GT 45.6%
Percentage of Reverso.T 36%
Bing.M.T 12%
Final Percentage of MT 93.6%
Final Percentage of HT 6.4%
Final Total Percentage (HT) 6.4% + (MT ) 93.6%= 100%
79 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)
Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

5. At the end ,while analyzing the students' renderings, different types, as shown in the following figure:
errors are classified in rendering all the (SLT) into

LINGUISTIC ERRORS

grammatical errors Pragmatic error

syntactical errors Addition

Morphological errors Omission

In appropriate word form Too free translation

Too literal translation

Wrong lexical choice

Wrong focus of attention

Distorted
meaning of the source text

VI. CONCLUSIONS 5. The communicative method is the most appropriate

method to render English literary texts especially in
The current study reaches to the following rhetorical devices because it is an appropriate
conclusions: translation that produces the same rhetorical impact
1. Machine Translation programs often function as a on the (TL) reader. So, the appropriate renderings
bilingual dictionary rather than a proper (MT) by (students), despite their small percentage, are
program, resulting in a more or less word-for-word considered the best, most correct, and most
translation. So, using the literal translation in understandable renderings than machine translation.
rendering literary texts distorts the intended The reason is that they translate the texts
meaning and fails to produce a rhetorical impact on communicatively, relying on themselves to deliver
the (TL) reader. those translations to the (TL) readers, and they
2. According to the data analysis, Arab translators know that machine translations have more errors in
(Students) face many problems and difficulties in renderings
rendering English literary sentences such as (lexical\ 6. Human Translation is better than Machine
syntactical \ rhetorical \ stylistic) problems. These Translation because the first uses communicative
problems are caused by transferring the (SL) text translation while the other uses semantic translation.
into the (TL) text semantically using machine
translation without ensuring the rendering if it is Our last supplication is Praise be to Allah, Lord
true or false. of the Worlds, and may blessings and peace be upon our
3. In this study, many translators(students) use Noble Messenger Mohammad, his family, and all his
different machine translation programs in rendering companions.
the (ST) expressions, and the most widely used REFERENCES
programs are Google Translation, Reverso
Translation, and Bing Microsoft Translation. [1] Abdelaal, N. M., & Alazzawie, A. (2020).
4. The programs ( Google, Reverso, Bing) cannot deal Machine Translation: The Case of Arabic-English
with the semantic phenomenon of ambiguity, a Translation of News Texts. In Theory & Practice
common characteristic feature of all machine in Language Studies ,Vol.10,(No.4),pp. pp. 408-
translation programs. This ambiguity is transferred 418.
from the (SL) text into the (TL) text semantically [2] Abdulaal, M. A. A. D. (2022). Tracing machine
using machine translation without ensuring the and human translation errors in some literary
rendering if it is true or false. texts with some implications for EFL translators.

80 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)

Integrated Journal for Research in Arts and Humanities
ISSN (Online): 2583-1712
Volume-4 Issue-1 || January 2024 || PP. 68-81 https://doi.org/10.55544/ijrah.4.1.11

In Journal of Language and Linguistic Studies, [15] Alsohybe, N. T., Dahan, N. A., & Ba-Alwi, F. M.
Vol.18,(No. S1),pp. 176-191. (2017). Machine-translation history and
[3] Abdullah, A. H. (2020). The Linguistic Problem- evolution: Survey for Arabic-English translations.
based Effect of Using Machine Translation in In Current Journal of Applied Science and
Mobile & Computer Apps: English and Arabic as Technology, Vol.23,(No.4),pp.1-19.
a case study. In Al-Qalam [16] Al-Timen, A., & Abbas, F. (2021). Analyzing
journal,Vol.4,(No.8),pp.325-349. Error Analysis in Google Translate: A Case
[4] Abdullah, T. A., & Thanoon, S. I. (2019). The Study of a Medical Text. In Journal of College of
Treatment of Lexical Ambiguity in Machine Education/Wasit, Vol. 43,(No.2),pp.737-752.
Translation. In Adab AL Rafidayn, Vol. [17] Arnold, D., Balkan, L., Humphreys, R. L.,
49,(No.79),pp. 1-12. Meijer, S., & Sadler, L. Machine Translation:
[5] Abdullah, Y. N., & Nasser, L. A. (2021). Text An introductory guide. Originally published in
Typology and Lexical Problems in Arabic- UK: NCC Blackwell Manchester/Oxford and
English Machine Translation. In Adab AL USA: Cambridge ,London.
Rafidayn, Vol.51,(No.87),pp. 29-54. [18] Arono, A., & Nadrah, N. (2019).
[6] Abulhassan ,B. (2014). Between English and STUDENTS’DIFFICULTIES IN
Arabic: A Practical Course in Translation, TRANSLATING ENGLISH TEXT in JOALL
United Kingdom: The British Library. (Journal of Applied Linguistics and Literature),
[7] Adetunji , B.(2020). Translation (Arabic- Vol.4,(No.1),pp. 88-99.
English). Unpublished M.A Thesis .Victoria – [19] Carl, M., & Báez, M. C. T. (2019). Machine
Island , Lagos: National Open University of translation errors and the translation process: a
Nigeria. study across different languages. In Journal of
[8] Al Saleh, R. A.(2019). An Application of Specialised Translation, Vol.31,pp. 107-132.
Newmark's Procedures to Muhammad Abul [20] DROBOT, I. A. (2021). Translating Literature
Quasem's English Translation of Mohammad Al- Using Machine Translation: Is It Really
Ghazali's Islamic Guidance.In World English Possible?. In Scientific Bulletin of the Politehnica
Journal,pp.1-59. University of Timisoara. Transactions on Modern
[9] AL-ANI, O. A. H., & AL-ANI, A. A. N.(2021( Languages/Buletinul Stiintific al Universitatii
.The Accuracy of Google Translation system in Politehnica din Timisoara. Seria Limbi Moderne,
Translating Referential Pronoun(that) within the 20(1).Vol.20,( No.1),pp.57-64.
Literary Texts .In International Journal of [21] ElBeheri, N. (2020). Collaboration between
Humanities and Educational Research, Vol.3 Machine Translation and Human Translation for
(No.5), pp. 192-201. Higher Quality and More Production in
[10] Alawi, N. (2010). Intertextuality and literary Translation. InTextual Turnings: An International
translation between Arabic and English. In An- Peer-Reviewed Journal in English Studies,
Najah University Journal for Research-B Vol.1,(No.1),pp. 446-473.
(Humanities), Vol.24,(No.8), pp.2437-2456.
[11] Alawneh, M., Omar, N., Sembok, T., ONLINE REFERENCES
Almuhtaseb, H., & Mellish, C. (2011). Machine
translation from English to Arabic. In [22] Bowker, L.(2021). Machine translation use
International Conference on Biomedical outside the language industries: a comparison
Engineering and Technology.Vol.11,pp.95-99. of five delivery formats for machine translation
[12] Alkhawaja, L., Ibrahim, H., Ghnaim, F., & literacy instruction. Retrieved from
Awwad, S. (2020). Neural machine translation: (https://doi.org/10.26615/978-954-452-071-
Fine-grained evaluation of Google translate 7_004 ) ( 12th May,2023).
output for English-to-Arabic translation. In [23] Burchardt, A. et al. (2016). Machine translation
International Journal of English Linguistics, quality in an audiovisual context. Retrieved
Vol.10,(No.4),pp. 43. from
[13] Al-khawalda, M., & Al-Oliemat, A. (2014). (http://dx.doi.org/10.1075/target.28.2.03bur )
Machine Translation: Deficiency in Translating (15th June ,2023).
English Sentences with Different Temporal [24] Irfan , M. (2017). Machine Translation .
References into Arabic. In Journal of Education Retrieved from
and Practice ,Vol. 5,(No.5),pp.47-52. (https://www.researchgate.net/publication/3207
[14] Al-Samawi, A. M. (2014). Language errors in 30405 ) (14th July ,2023).
machine translation of encyclopedic Texts from
English into Arabic: the case of Google
Translate. In Arab World English Journal,
(No.S3),pp.182-211.

81 Attribution-NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0)