Download as pdf or txt
Download as pdf or txt
You are on page 1of 214

Uploaded by S. M.

Safi
Noureldin Abdelaal

Translation between English and Arabic

Uploaded by S. M. Safi
A Textbook for Translation Students and
Educators

Uploaded by S. M. Safi
Noureldin Abdelaal
University of Nizwa, Nizwa, Oman

ISBN 978-3-030-34331-6 e-ISBN 978-3-030-34332-3


https://doi.org/10.1007/978-3-030-34332-3

© The Editor(s) (if applicable) and The Author(s) 2020

This work is subject to copyright. All rights are solely and exclusively
licensed by the Publisher, whether the whole or part of the material is
concerned, speci ically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on micro ilms or in
any other physical way, and transmission or information storage and
retrieval, electronic adaptation, computer software, or by similar or
dissimilar methodology now known or hereafter developed.

The use of general descriptive names, registered names, trademarks,


service marks, etc. in this publication does not imply, even in the
absence of a speci ic statement, that such names are exempt from the
relevant protective laws and regulations and therefore free for general
use.

The publisher, the authors and the editors are safe to assume that the
advice and information in this book are believed to be true and
accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with
respect to the material contained herein or for any errors or omissions
that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional af iliations.

This Palgrave Macmillan imprint is published by the registered


company Springer Nature Switzerland AG

Uploaded by S. M. Safi
The registered company address is: Gewerbestrasse 11, 6330 Cham,
Switzerland

Uploaded by S. M. Safi
This book is dedicated to the soul of my father.

Uploaded by S. M. Safi
Acknowledgements
I offer my heartfelt thanks to my father, who spared no effort in
supporting me, and to my great mother, who has always kept me in her
prayers and prayed for my success. I am also deeply indebted to my
mother who has always been immensely supportive.
My deepest thanks go to my wife and my children for their
solicitude and love.
To all who have supported me in the preparation of this book, I
extend my gratitude.

Uploaded by S. M. Safi
Abbreviations
SL Source language
ST Source text
TL Target language
TT Target text

Uploaded by S. M. Safi
Contents
1 Translational Concepts
1.1 De initions of Translation
1.2 Translation Unit
1.3 Meaning in Translation
References
2 Translation Theory
2.1 Stages of Translation Theories
2.1.1 Linguistic Stage
2.1.2 The Communicative Stage
2.1.3 The Functionalist Stage
2.1.4 The Ethical/Aesthetic Stage
2.2 The Notion of Equivalence in Translation Theories
2.2.1 Direct and Oblique Translation (Vinay & Darbelnet,
1958/2004)
2.2.2 Van Leuven-Zwart’s Comparative-Descriptive Model of
Translation Shifts (1989)
2.2.3 Overt and Covert Translations (House, 1997)
2.2.4 Jakobson’s Equivalence (1959)
2.2.5 Dynamic Equivalence and Formal Equivalence (Nida,
1964)
2.2.6 Communicative and Semantic Translation (Newmark,
1981, 1988)
2.2.7 The House, Nida, and Newmark’s Theories in a
Nutshell
2.2.8 Form-Based and Meaning-Based Translation (Larson,
1998)

Uploaded by S. M. Safi
2.2.9 Halliday’s Typology of Equivalence
2.2.10 Catford’s Typology of Equivalence
2.2.11 Mona Baker’s Typology of Equivalence
2.2.12 Koller’s Notion of Equivalence
2.2.13 Popovič (1976) Types of Equivalence
2.2.14 The Cognitive Approach to Translation
2.2.15 Functionalist Approach in Translation (Non-
equivalence Approach)
2.2.16 Darwish’s Notion of Equivalence (2010)
2.2.17 The Polysystem Theory
References
3 Grammatical Problems in Translation
3.1 Arabic Tense as a Problem in Translation
3.2 Translating English Tenses and Aspects into Arabic (Based
onCollins COBUILD English Grammar , 2005)
3.3 Gender as a Problem in Translation
3.4 Grammatical Category as a Problem in Translation
3.5 Syntactic Order: Foregrounding and Backgrounding as a
Problem in Translation
3.6 Shifting (Iltifat ) as a Problem in Translation
References
4 Lexical and Semantic Problems in Translation
4.1 Lexical Gaps at the Semantic Field Level (Lack of Equivalent
Problem)
4.2 Improper Selection of Vocabulary
4.3 Lexical Ambiguity: Polysemy and Homonymy
4.4 Synonymy

Uploaded by S. M. Safi
4.5 Problems in Translation of Rhetorical Devices
References
5 Culture as a Problem in Translation
5.1 Culturally Bound Terms
5.2 Idioms and Fixed Expressions as a Problem in Translation
5.3 Collocations
5.4 Strategies to Translate Culturally Bound Terms
5.5 Pym’s Typology of Translation Solutions
5.6 Conclusion
References
Bibliography
Index

Uploaded by S. M. Safi
List of Tables
Table 2.1 Transeme and architranseme relationship (based on van
Leuven-Zwart’s comparative-descriptive model of translation shifts,
1989)

Table 2.2 Main categories of van Leuven-Zwart’s comparative model


(from van Leuven-Zwart 1989, pp. 159–169)

Table 2.3 The gradual erosion of the notion of equivalence in


translation studies

Table 2.4 Terms used in the translational action approach

Table 5.1 The collocations of ‘bend’

Table 5.2 Typology of translation solutions (Pym, 2018, p. 45)

Uploaded by S. M. Safi
© The Author(s) 2020
N. Abdelaal, Translation between English and Arabic
https://doi.org/10.1007/978-3-030-34332-3_1

1. Translational Concepts
Noureldin Abdelaal1
(1) University of Nizwa, Nizwa, Oman

Noureldin Abdelaal

Overview
This chapter explains the main concepts related to translation. It
provides theoretical de initions of translation. It also explains the
concept of the ‘translation unit’, and how scholars of translation
disagree on identifying the unit of translation. Some scholars
consider a word to be the unit of translation; others believe that a
unit of translation may be a sentence, piece of text, or culture.
Moreover, the chapter sheds light on the thorny notion of meaning
in translation studies.
In particular, this chapter covers the following topics:
A.
De inition of translation
B.
Translation unit
C.
Meaning in translation

1.1 De initions of Translation


Translation is a controversial concept that is hotly debated. There are
many de initions of translation that revolve around meaning, and the
different notions of equivalence. Halliday et al. (1965) describe

Uploaded by S. M. Safi
translation as the establishment of textual equivalents, rather than
lexical or grammatical equivalents. Thus, as stated by Halliday et al.
(1965), translation is not a mere word-for-word rendition but, rather,
translation is seen as a whole text-to-text transplanting. Catford (1965,
p. 20), similar to Halliday, de ines translation as ‘the replacement of
textual material in one language, i.e. the source language (SL) by
equivalent textual material in another language, i.e. target language
(TL).’ House (2001) perceives translation as a reproduction of a text in
an SL in an equivalent text in a TL. Thus, Halliday et al. (1965) see
translation in terms of textual equivalence, and not word-for-word
equivalence. Widdowson (1978) sees translation from a different
perspective; he views translation from a communicative perspective.
Widdowson states that translation should neither operate at the word
level, the sentence level, the lexical level, or the grammatical level;
translation should be only at the communicative level. As for Hatim and
Munday (2004), they state that there are two distinctive senses of
translation: translation as a process, and translation as a product. They
view translation, or to use their words, the ambit of translation as: (1)
the process of transferring a ST into a TT in a speci ic socio-cultural
context; (2) a product which is the result of the previous step; and (3)
the cognitive, linguistic, visual, cultural and ideological phenomena
that are a principal component of (1) and (2). This de inition seems
have greater care for the socio-cultural aspects of translation.
Other scholars and researchers (e.g. Venuti, 2004; Ahmed, 2006)
see translation as a process that not only implies conveying meaning in
a TL, but that should retain the same style and tone of a ST: the
translated text should not appear to be a translation but, rather, should
appear to be an original text, wherein no translator is visible. This
de inition focuses on the equivalence level between a ST and a TT; it
goes beyond lexical equivalence to include style and tone, which are
dif icult to achieve. However, another perspective sees translation as
reproducing a text in one language in another TL to make it accessible
to a larger audience (Ordudari, 2007). Ordudari’s de inition is more
concerned with the aim of translation; that is, to reach a greater
audience, which applies to many types of texts. Levý (1967) sees
translation as a process of decision making, whereby the components
of this decision are:

Uploaded by S. M. Safi
1. The situation: sometimes the SL expresses one lexical item using a
certain item where the TL has two equivalents for the same word.
For example, a translator has to make a decision when translating
the English word ‘eclipse’ because it has two equivalents in Arabic
(i.e. one is related to the moon and the other related to the sun).
2.
Instruction I: This implies de ining the class of possible
alternatives.
3.
Instruction II: This denotes making a selection from the available
class alternatives. This selection is context-based. For example, to
translate the word ‘eclipse’ into Arabic, a translator should refer to
the context to understand whether it is a ‘lunar eclipse’ (relating to
the moon), or a ‘solar eclipse’ (relating to the sun).
Levy’s (1976) view of translation is related to the process of the
translation, which sounds practical. Levy’s perspective of translation
sounds comprehensive, as it encompasses the notion of ‘equivalence’
without disregarding the role of a translator in selecting the most
appropriate equivalent. Another perspective of translation is that of
Reiss (2004), who sees translation as a process of producing a text in
the TT that is functionally equivalent to the ST. However, she goes on to
say that, during the communication process, the message will be
altered, perhaps by a translator’s views, or experience and knowledge.
These changes can result in two types of message changes:
1.
Unintentional changes: Changes may result from the differences
between the structure of a language, or from the degree of the
translator’s competence.
2.
Intentional changes: These changes can affect the functionality of
the ST. This kind of change occurs if the aim of the ST is rendered
differently in the TT.
Reiss’s view of translation equivalence is in terms of functional
equivalence. Overall, most of the aforementioned views of translation
are centred on certain basic notions—lexical equivalence and meaning

Uploaded by S. M. Safi
in translation; and the functional equivalence in translation—that
re lect the different approaches and theories of translation.

Exercise
‘Translation’ has been given various de initions by a variety of
scholars. Elaborate on this, explaining the implications behind each
of these de initions.

1.2 Translation Unit


A translation unit, as de ined by Manfredi (2014), is the linguistic level
employed by a translator during their act of translation. Theorists hold
a variety of positions in regard to the translation unit, depending on
what they consider a translation unit to be. For example, Vinay and
Darbelnet proposed the terms ‘lexicological unit’ or ‘unit of thought’ as
a translation unit. They rejected the notion that a word can be a unit of
translation (Manfredi, 2014). Newmark (1988, pp. 66–67), on the other
hand, regards a ‘sentence’ as the best unit of translation. He justi ies his
view by stating that a sentence is a unit of thought and a means
presenting objects. He adds, ‘All lengths of language can, at different
moments and also simultaneously, be used as units of translation in the
course of the translation activity’ (Newmark, 1988, pp. 66–67).
However, Newmark mentions that in some texts such as expressive
texts, a ‘word’ should be deemed as the unit of translation as it can
better convey the inest nuances. Bassnett (2005), however, states that
a text should be the unit of translation, especially in relation to literary
prose texts. Strangely, Snell-Hornby considers the notion of culture as
the unit of translation (Hatim and Munday 2004). Manfredi (2014), in
line with Halliday and Matthiessen (2014), argues that a clause is the
most proper unit of translation. However, Manfredi states that in
written translations, especially literary ones, a sentence should be
considered as the unit of translation. Sentence, in this context, refers to
a graphological unit that begins with a capital letter and ends with a full
stop. ‘Word, in its context, can be the proper unit of translations,
especially authoritative and sacred texts such as the Holy Quran
because nuances between words give different meanings.’ In addition,

Uploaded by S. M. Safi
faithfulness to the ST requires carrying out the translation process at
the word level, rather than at sentence level. In sum, different scholars
revealed different understandings and perspectives of the unit of
translation. Whatever the case may be, translating these units of
translations (whether word, sentence, clause, or culture) poses many
problems, which will be discussed in due course. However, irst, we
should discuss meaning in translation, as translation is a process of
conveying meaning.

Exercise
One important notion in translation theory is the ‘unit of
translation’. Why do you think that this notion is important? How
does it affect the theory and practice of translation?

1.3 Meaning in Translation


Halliday (1992) states that a main feature of translation is that it is a
process of meaning-making and that without such creation of meaning
there is no translation. He also states that this creation of meaning is a
guided activity. Halliday further comments that, for a language theory
to be related to translation, it should be concerned with functional
semantics. He explicates his notion of function by stating that it does
not refer to the vague sense of use; rather, it refers to ‘metafunction’.
Metafunction, as de ined by Halliday, is ‘function as the fundamental
organizing concept around which all human language has evolved’
(Halliday, 1992, p. 15). He has made it clear that his notion of functional
semantics does not imply discarding the formal patterns (e.g.
phonological or syntactic patterns), but that these formal patterns
should keep the semantic relations in place. He states that complete
semantic equivalences between any two languages cannot be absolute.
They can be contingent on the contextual meaning of an item. He
makes a clear deduction of the meaning of ‘equivalence of meaning’ as
being ‘equivalence of function in context’ (Halliday, 1992, p. 16).
Halliday recommends that, in a translation process, target contextual
equivalents should be found. A translator, then, should select from
among the variants of equivalents. A translator should also identify the

Uploaded by S. M. Safi
relevant context that conditions their choice in order to translate the
given ‘item’ in the most relevant way.
Halliday (1992) believes that linguistics cannot offer a ‘theory of
translation equivalence’; rather, it can offer a ‘theory of context’. Such a
theory of context that can be offered to a translator is driven by the
functional notion of ‘constituency’. By the use of the term ‘constituency’
(which refers to one of Halliday’s contexts) Halliday refers to the part–
whole hierarchical relationship between grammatical constituents. Put
more simply: a clause is made up of phrases, phrases are made of
words, words are made up of morphemes, and so on. Following this
model, a translator needs to move one or more levels up the scale.
Halliday gives an example of the morpheme ‘-er’, which can be moved
to a word, such as in the word ‘driver’. The ‘-er’ morpheme in ‘driver’,
depending on context, means either a person who works as a driver, or
someone who is currently driving. Thus, a translator should select the
proper equivalent based on the meaning of this morpheme in context.
It is noteworthy to highlight that, for example, the context of a
morpheme such as -er comprises the words with which it may occur.
Halliday (1992) states that any piece of discourse represents the
mapping of three simultaneous structures with three different
constituents of meaning (i.e. ideational, interpersonal and textual).
Halliday recommends examining all these meanings, including the
‘writer’s construction of his or her own subjectivity and that of the
audience, of attitude to and distance from the subject-matter and so on’
(Halliday, 1992, p. 20). Another context mentioned by Halliday is
‘discourse semantics’, which, for example, includes grammatical
metaphor. Other contexts are those of situation (Halliday, 1992, p. 21),
and culture (Halliday, 1992, p. 23), as one cannot decide on the
meaning of a word without considering the situation and culture as
factors that contribute to meaning-making. To return to the de inition
of translation as a ‘guided creation of meaning’, Halliday argues that
such creation of meaning is constructed through the context of a
situation, which results from analysis of the text. Thus, such context of
situation will ‘guide’ the creation of the new translated text.
However, there are various types of meaning, and a variety of
classi ications according to a number of semanticists and linguists. For
example, Cruse (1997) identi ied four types of lexical meaning:

Uploaded by S. M. Safi
propositional meaning, expressive meaning, presupposed meaning
and evoked meaning. The term ‘propositional meaning’ is used to
describe the relation between a word and its real or imaginary
meaning. For example, socks are ‘a kind of cloth worn on feet’. This kind
of meaning can be judged in terms of true or false. This meaning is
referred to by other semanticists (e.g. Palmer, 1981; Hurford et al.,
2014) as ‘denotational meaning’. ‘Expressive meaning’ refers to the
speaker’s feelings or attitudes. Thus, this meaning cannot be judged in
terms of true or false. For example, ‘cruel’ and ‘unkind’ are words that
denote the meaning of disapproval of someone’s attitude: of the two
words, ‘cruel’ has the stronger and most negative meaning. Some
words have propositional and expressive meaning (e.g. whinge); some
have expressive meaning only (e.g. bloody); and others have
propositional meaning only (e.g. book).
The third type of meaning is ‘presupposed meaning’, whereby
meaning arises from restrictions occasioned by co-occurrence. These
restrictions include selectional restrictions and collocational
restrictions. Selectional restrictions are always observed, with the
exception of the igurative use of language. For example, the verb
‘speak’ is expected to refer to a human being, while ‘meow’ to non-
human creatures. Collocational restrictions, on the other hand, refer to
arbitrary semantic co-occurrences. For example, a law is broken in
English, whereas in Arabic it is contradicted, not broken. The last type
of meaning is the ‘evoked’ meaning, which arises from differences in
dialect and register. Propositional meaning is the only type of meaning
to fall into the true/false category.
Leech identi ied seven types of meaning in semantics: conceptual,
associative, affective, collocative, connotative, re lected and thematic.
‘Conceptual meaning’ is sometimes referred to as ‘denotative’,
‘designative’, ‘cognitive’, or ‘descriptive’ meaning. It is the primary
meaning of a lexeme, and is based on contractiveness and constituent
structures. This meaning can be found in dictionaries—wherein
certain lexical features are identi ied (constituent structures), and
other lexical features are excluded (contractiveness structures)—and
can be syntactically analysed. ‘Associative meaning’ refers to the
individual mental understandings of the speaker. It is subdivided into
six categories: connotative, collocative, social, affective, re lected and

Uploaded by S. M. Safi
thematic. ‘Connotative meaning’ is the real-world meaning; it is
associated with a speci ic lexeme, based on the user’s experience, and
it is subjective. It includes ‘social meaning’, which differs from one
person to another depending on factors such as age, gender and so on.
For example, the word ‘home’ may encompass meanings for the
language user that differ from those of another. For some, ‘home’
implies ‘love’, ‘wife’, ‘family’; for others, it may imply ‘boredom’ and
‘monotony. ‘Affective meaning’ is the meaning that communicates the
emotions of the language user. For example, ‘I am awfully sorry for
doing that’ is an expression of regret, or feeling sorrow. Intonation and
voice timbre can affect this kind of meaning. ‘Re lected meaning’ is the
meaning that arises from the use of word in a speci ic context; for
example, ‘pray’ can have a variety of meanings, depending on the
context. ‘Collocative meaning’ is the meaning that collocates with
speci ic words; for example, ‘pretty’ and ‘handsome’ refer to the same
denotative meaning but they are used in a collocatively different
manner. As for ‘thematic meaning’, it depends on the order of words
and how they affect meaning. In other words, it depends on the theme
(i.e. what is being talked about).
Transferring meaning from an ST to a TT is a complex process due
to the complicated nature of the notion of meaning, on which there is
no particular agreement. Meaning has no precise de inition; Ogden and
Richards (1923) listed 16 different meanings for the word ‘meaning’
itself. Meaning is an ambiguous and fuzzy concept that lacks clear-cut
understanding. Such ambiguity of meaning is a characteristic of all
languages (Bock, 1986). Ghazala (2008) identi ies meaning as the
linguistic components: grammar, vocabulary, style and phonology.
Grammar, by turn, includes sentence, clauses, word order, tenses and
such matters. Vocabulary includes the sense relations (e.g. synonymy,
antonymy and so on), idioms, collocations, proverbs, metaphor, culture
and such. Style includes ambiguity, repetition, redundancy,
nominalization, verbalization, fronting, formality and so forth.
Phonology includes rhyme, rhythm, assonance, alliteration and so on.
Together, these elements shape meaning. Though, in translation,
grammar may not be translated in a straightforward manner, it
nonetheless affects the meaning and the message conveyed (Ghazala,
2008). Hence, as Ghazala (2008) states, only meaning—which includes

Uploaded by S. M. Safi
grammar, style, vocabulary and phonology—is translated; there are a
few exceptions, however, such as in poetry, in which prosody may be
the primary aim, rather than meaning. Meaning is regarded as the total
components of words, grammar, style and sounds (Ghazala, 2008).
Ghazala mentions that it is dif icult to translate meaning and form
simultaneously. Ghazala’s perspective on the dif iculty of preserving
meaning and form simultaneously is consistent with Larson’s (1998)
perspective of translation. Quine (1959) proposed a different notion of
meaning in translation. He introduced what he called ‘empirical
meaning’. Empirical meaning is de ined as ‘what remains when, given
discourse together with all its stimulatory conditions, we peel away the
verbiage’ (Quine, 1959, p. 94). As for Cruse (1997), each word is
assumed to have canonical traits that cannot be discarded. For
example, a bird has the canonical trait of lying.
Translating such meaning from one language to another is
challenging. Nugroho (1999) argues that rendering meaning is a
process that involves aspects such as diction, grammatical structure,
communication setting, and cultural context of the ST. He adds that
meaning in an ST should be equivalent to meaning in a TT. In short,
translation is basically about translating meaning from one SL to a
different TL. This process of transferring meaning is complicated, and
many problems must be faced.

Exercises
1.
According to Leech, what types of meaning describe the
following words? Provide appropriate translations for these
words.
A.
Girl: [+HUMAN-ADULT+FEMALE]
B.
Secondhand in: (a) a second-hand car
(b) a second-hand smoker
C.
Mother vs. mama
2. Meaning is an important concept in translation studies and has

Uploaded by S. M. Safi
been de ined in different ways. Explain.
3.
Explain, with examples, the seven types of meaning proposed by
Leech.
4.
Cruse identi ied three types of meaning. Explain these types,
with examples.

References
Ahmed, M. F. (2006). Investigating some semantic problems in the translation of the Holy Quran.
Adab al-Ra idayn, 2(43), 61–72.

Bassnett, S. (2005). Translation studies. Vasa (3rd ed.). New York: Routledge; Taylor & Francis.

Bock, J. K. (1986). Syntactic persistence in language production. Cognitive Psychology, 18(3), 355–
387. https://doi.org/10.1016/0010-0285(86)90004-6.

Cruse, D. A. (1997). Lexical semantics. Cambridge: Cambridge University Press.

Ghazala, H. (2008). Translation as problems and solution. Beirut: Dar El-Ilm Lilmalayin.

Halliday, M., Macintosh, A., & Strevens, P. (1965). The linguistic sciences and language teaching.
London: Longman Publishing House.

Halliday, M. A. K. (1992). Language theory and translation practice. Rivista Internazionale Di


Tecnica Della Traduzione, 1(1), 15–25.

Halliday, M. A. K., & Matthiessen, C. (2014). Halliday’s introduction to functional grammar (4th
ed.). New York: Routledge; Taylor & Francis. https://doi.org/10.4324/9780203431269.

Hatim, B., & Munday, J. (2004). Translation: An advanced resource book. London and New York:
Routledge; Taylor & Francis.

House, J. (2001). Translation quality assessment: Linguistic description versus social evaluation.
Meta: Journal Des Traducteurs, 46(January), 243. https://doi.org/10.7202/003141ar.

Hurford, J. R., Heasley, B., & Smith, M. B. (2014). Semantics: A coursebook. Cambridge: Cambridge
University Press.

Larson, M. (1998). Meaning-based-translation. Oxford: University Press of America.

Levý, J. (1967). Translation as a decision process. In L. Venuti (Ed.), To honour Roman Jakobson on
the occasion of his seventieth birthday (Vol. 2, pp. 1171–1182). The Hague: Mouton.

Levy, J. (1976). Translating as a decision process. In L. Venuti (Ed.), The translation studies reader
(pp. 148–189). London: Routledge.

Uploaded by S. M. Safi
Manfredi, M. (2014). Translating text and context: Translation studies and systemic functional
linguistics. Volume 2: From theory to practice. In Quaderni del CeSLiC. Functional grammar
studies for non-native speakers of English (p. 158). Bologna: Centro di Studi Linguistico-Culturali
(CeSLiC). https://doi.org/10.6092/unibo/amsacta/3219.

Newmark, P. (1988). A textbook of translation. Hertfordshire: Pearson Education Limited.

Nugroho, A. B. (1999). Meaning and translation. Journal of English and Education, 2(3), 94–112.

Ogden, M., & Richards, I. A. (1923). The meaning of meaning. New York and London: Harcourt
Brace Jovanovich.

Ordudari, M. (2007). Translation procedures, strategies and methods. Translation Journal, 3(5),
781–789.

Palmer, F. (1981). Semantics: A new outline. London and New York: Cambridge University Press.

Quine, W. V. O. (1959). Translation and meaning. In R. A. Brower (Ed.), On translation. Cambridge,


MA: Harvard University Press (Reprinted in L. Venuti (Ed.). (2000). The translation studies reader
(pp. 94–112). London: Routledge).

Reiss, K. (2004). Type, kind and individuality of text: Decision making in translation. In L. Venuti
(Ed.), The translation studies reader (2nd ed., pp. 168–179). New York: Routledge.

Venuti, L. (2004). The translation studies reader. London and New York: Routledge.

Widdowson, H. (1978). Teaching language as communication. Oxford: Oxford University Press.

Uploaded by S. M. Safi
© The Author(s) 2020
N. Abdelaal, Translation between English and Arabic
https://doi.org/10.1007/978-3-030-34332-3_2

2. Translation Theory
Noureldin Abdelaal1
(1) University of Nizwa, Nizwa, Oman

Noureldin Abdelaal

Overview
This chapter brie ly explains the stages of translation theory: the
linguistic stage, the communicative stage, the functionalist stage
and the ethical/aesthetic stage. It also presents the notion of
equivalence in translation theories, with reference to the most
prominent theories in translation, supported by examples.
The chapter covers the following topics:
Stages of translation theories
The notion of equivalence in translation theories
The following research works support the contents of this
chapter:
1.
Direct and Oblique Translation (Vinay & Darbelnet,
1958/2004)
2.
Van Leuven-Zwart’s Comparative-descriptive Model of
Translation Shifts (1989)
3.
Overt and Covert Translations (House, 1997)
4.
Jakobson’s Equivalence (1959)
5.
Dynamic Equivalence and Formal Equivalence (Nida, 1964)

Uploaded by S. M. Safi
6. Communicative and Semantic Translation (Newmark, 1981,
1988)

7.
Form-Based and Meaning-based Translation (Larson, 1998)
8.
Halliday’s Typology of Equivalence
9.
Catford’s Typology of Equivalence (1965)
10.
Mona Baker’s Typology of Equivalence (1992/2011)
11.
Koller’s Notion of Equivalence
12.
Popovič (1976) Types of Equivalence
13.
The Cognitive Approach in Translation (Bell, 1991)
14.
The Functionalist Approach
15.
Non-Equivalence Approach (Skopos Theory)
16.
The PolySystems Theory

2.1 Stages of Translation Theories


In his discussion of translation theory, Munday (2009) explains that
translation theory was controlled by the West until recent times. He
adds that, in Western Europe, the topic of word-for-word or sense-for-
sense translation was the subject of heated debate until the twentieth
century. Further, Munday (2009) states that ‘translation studies’ as a
discipline did not emerge until the second half of the twentieth
century; it arose from the branches of applied comparative linguistics
and modern languages. The concept of translation studies was irst
introduced by James Holmes as a substitute for ‘translation science’, or

Uploaded by S. M. Safi
‘translatology’, in 1972. However, Newmark (2009) favours ‘translation
theory’ over ‘translation studies’. He views theory as an important
framework that should be taught to translation students, though he
states that learning a theory is not fundamental to being a good
translator. Peter Newmark (2009) identi ied four stages of translation
theory: linguistic, communicative, functionalist and ethical/aesthetic.
Each stage is marked with a unique approach.

2.1.1 Linguistic Stage


Covering the period up to 1950, this stage was basically concerned
with literary texts—that is, poetry, short stories, plays, novels and
autobiographies. This stage was predominantly concerned with the
discussion of the word-for-word translation (literal), as opposed to
sense-for-sense translation (natural, liberal, or idiomatic). During this
period, there was preference for sense-for-sense or contextual
translation over word-for-word translation. This, as Newmark states,
marks the interpretive theory of translation. The most prominent work
of translation theory in this period was Essay on the Principles of
Translation by Alexander Tytler (1790). Tytler (1797, pp. 14–15), as
cited in Newmark (2009), de ined a good translation as one in which
‘the merit of the original work is so completely transfused into another
language as to be as distinctly apprehended, and as strongly felt, by a
native of the country to which that language belongs, as it is by those
who speak the language of the original work’. Newmark adds that what
can be inferred from Tytler’s statement is that a good translation
should completely convey the message of the ST; it should also follow
the same style and manner of the original, and should have all the ease
of the original composition (p. 23). George Steiner’s After Babel (1975)
marks the end of this linguistic stage.

2.1.2 The Communicative Stage


Beginning in around 1950, this stage marked the application of
linguistics to translation studies; it mainly covered non-literary and
literary texts. It was concerned with the categorization of text
registers, the participation of a range of readership groups (from the
less well-educated to the expert) and the identi ication of types of
procedures for translating various segments of a text.

Uploaded by S. M. Safi
2.1.3 The Functionalist Stage
Commencing in around 1970, this stage covered mainly non-literary
texts—that is, ‘the real world’. It was concerned with the intention of a
text and its essential message, rather than the language of the ST.
Translation in this period was concerned with how to translate a text
functionally.

2.1.4 The Ethical/Aesthetic Stage


Since around 2000, this stage has been concerned with authoritative
and of icial or documentary texts, and includes serious literary works.
It highlights translation as a truth-seeking profession. The truth is
essentially twofold: the correspondence of a factual text with reality;
and the correspondence of an imaginative text with a meaningful
allegory—and, consequentially, the correspondence of the translation
with the respective type of text. Newmark concludes that these
translation theory stages are cumulative; in other words, they overlap,
or, in Newmark’s words, they ‘absorb without eliminating each other’
(2009, p. 21). Having shed light on the different stages of translation
theories, we shall move on to the unit of translation.

Exercise Explain the stages of translation theory.

2.2 The Notion of Equivalence in Translation


Theories
Discussing the concept of ‘equivalence’ brings into the discussion the
perspectives of concepts. There are two main perspectives of concepts:
the universality of concepts, as proposed by Chomsky (1977), and the
relativity of concepts. According to Chomsky, all humans share the
same basic brain structures and, thus, there are deep similarities
between all languages, even if these are not obvious in surface
grammar. Universalists believe all languages have a commonality, or
universal concepts, that are shared by all languages. Relativists believe
that languages are too disjointed and, hence, concepts are not common
among languages (Steiner, 1998). In his book After Babel (1998),

Uploaded by S. M. Safi
Steiner rejects Chomsky’s universality of concepts. Steiner believes
that language is relative and that, thus, languages are too disjointed.
Steiner takes the stance of the Relativists and opposes that of the
positivistic Universalists (Steiner, 1998). These variant stances of
Universalists and Relativists bring different understandings of the
notion of equivalence. Simply put, if we were to adopt the Universalist
stance, we would say that equivalence is achievable between languages
because they are similar in deep structure, at least. However, adopting
the Relativist stance, it can be argued that real equivalence does not
exist between languages. Those different stances created considerable
debate in relation to the concept of equivalence, which has always been
a source of disagreement among scholars and theorists of translation
and linguistics (Munday, 2009).
According to Munday (2009), equivalence is a thorny issue in the
realm of translation studies; it is fuelled by the debate among theorists
and scholars; some scholars more or less reject the notion (e.g.
Gentzler, 2001; Snell-Hornby, 1988/1995), while others ind it useful
and helpful (e.g. Baker 1992; Kenny 1998). By contrast, some scholars
perceive that translation without equivalence is impossible (e.g. Koller,
1989, 1995; Nida and Taber, 1974/1982). However, Munday concludes
that equivalence is a principal issue in the world of translation, and
that it will remain essential to the practice of translation (Munday,
2008, p. 49). There is clear evidence of the necessity for equivalence in
translation; irst, the de initions of translation mainly revolve around
the notion of equivalence (e.g. Catford, 1965; Newmark, 1981, 1988);
second, translation is basically a kind of communication, hence
equivalence between ST and TT is a requirement; third, dif iculty of
translation and untranslatability are always discussed with respect to
inding equivalent items in a TT (Yinhua, 2011). The concept of
equivalence was dominant in the discussions of translation during the
period during the 1960s and 1970s (Venuti, 2004).
Many scholars and theorists, adopting a variety of perspectives,
discussed the notion of equivalence. The notions of equivalence of
Vinay and Darbelnet, Mona Baker, Jakobson, Nida, and Newmark,
together with the strategies proposed by them, will be discussed in the
following sections.

Uploaded by S. M. Safi
2.2.1 Direct and Oblique Translation (Vinay & Darbelnet,
1958/2004)
Vinay and Darbelnet (2004) were basically in luenced by Catford’s
(1965) shifts. They identi ied two strategies of translation: direct and
oblique. They argued that changing the syntactic order and lexis of the
ST in the TT is sometimes necessary in order to transpose certain
stylistic effects of the ST, so as to ill the gap in the TL: oblique
translation. Sometimes it is possible to transpose the ST message
elements into the TT individually, due to structural or metalinguistic
parallelism between the ST and the TT: direct translation. These
strategies are subdivided into seven procedures; three for direct
translation and four for oblique translation. Those for direct
translation include: borrowing, calque, literal translation,
transposition, modulation, equivalence, adaptation.
A.
Borrowing, is where an SL word is transferred to the TT to ill a
semantic gap in the TL. One of the advantages of this strategy is
that it keeps the same connotations of the SL (Ni, 2009). Moreover,
this method adds the lavour of the SL culture to the TL. Some of
the borrowed items became a central core of the repertoire of
lexicons in the TL. For example, menu, coup d’état, café, alcohol,
sheik and Islam are part of the English language, though they
basically belong to other cultures and language. Similarly, Arabic
words such as ‫ أﻧﺘﺮﻧﺖ‬،‫ ﺳﻮﺑﺮﻣﺎرﻛﺖ‬and many others were borrowed
from English. Also, many words were borrowed from Arabic to
English, such as: cotton, falafel, algebra, sheriff, Mujahidin,
Fedayeen, caliph, sheik, halal and many others. Using borrowing as
a translation strategy should observe the style and message to be
conveyed accurately.
B. Calque, whereby an SL expression or structure is transferred with
minimum adaptation, is a special kind of borrowing and is
subdivided it into two types: lexical calque and structural calque.
1. In lexical calque, the SL lexis are transferred into the TT
without violating the syntactic structure of the TT; for
example, translating the English expression ‘compliments of
the season’ into French as ‘Compliments de la saison!’. Other
l ‘S t G l’ hi h i t l t d i t ‫اﻷ ﻦ‬

Uploaded by S. M. Safi
examples are ‘Secretary General’, which is translated into ‫اﻷﻣﯿﻦ‬
‫ اﻟﻌﺎم‬and ‘life is a journey’, which can be rendered as ‫اﻟﺤﯿﺎة رﺣﻠﺔ‬.
Other examples, respectively, include translating cornerstone,
feedback, play a role as ‫ ﺗﻐﺬﯾﺔ راﺟﻌﺔو ﯾﻠﻌﺐ دورا‬,‫ﺣﺠﺮ اﻟﺰاوﯾﺔ‬.
2.
In structural calque, a new structure is introduced into the TL,
translating the ST lexicons literally. For example, translating
the English expression ‘Science iction’ into French as ‘Science
iction’. In this example, the English structure is introduced into
the French language. To clarify, calque is a kind of literal
translation that sometimes observes the lexical features of the
ST (i.e. lexical calque), and at other times observes the
structural features of the ST (structural calque). Examples of
structural calque between English and Arabic rarely exist
because the two languages belong to two different families.
C.
Literal translation is a word-for-word translation; it is described as
the most common procedure between related or close languages
and cultures (e.g. French and Italian). For example, translating
‘Ahmed is a student’ as ‫ أﺣﻤﺪ ﯾﻜﻮن ﻃﺎﻟﺐ‬is a literal translation that can
be used for a didactic purpose only. However, literal translation is
sometimes possible at the lower level of language. For example,
translating ‘I love Rabiaa’ as ‫ اﻧﺎ أﺣﺐ رﺑﯿﻌﺔ‬is an acceptable literal
translation. Other examples are:

Example
I drink tea.
‫أﻧﺎ أﺷﺮب اﻟﺸﺎي‬
I speak English.
‫أﻧﺎ أﺗﺠﺪث اﻷﻧﺠﻠﯿﺰﯾﺔ‬
I bought a villa.
‫أﻧﺎ اﺷﺮﯾﺖ ﻓﯿﻠﻼ‬

According to Vinay and Darbelnet (2004), if all the direct or literal


translation procedures, mentioned above have not yielded acceptable
translations, oblique translation offers an alternative. The

Uploaded by S. M. Safi
unacceptability of translation as identi ied by Vinay and Darbelnet
refers to cases where ‘the message translated:
1.
gives another meaning;
2.
has no meaning;
3.
is structurally impossible;
4.
does not have a corresponding expression within the
metalinguistic experience of the TL; or
5.
has a corresponding expression, but not within the same register’
(p. 87).
Oblique translation procedures include: transposition, modulation,
equivalence and adaptation.
D. Transposition implies changing a part of speech (i.e. word class)
without altering the meaning. There are two types of
transposition: obligatory and optional.
1.
Obligatory transposition, which is sought when the TL does
not allow anything other than a speci ic form. For example, the
French expression ‘Dès son lever’ must be transposed into the
English expression ‘As soon as he gets up’. This is the only
permissible form in English. A further example is translating
‫ أﻏﺘﺴﻞ‬as ‘to wash up after having sex or ejaculation’. In this
example, the ST verb must be rendered as a clause in English
due to a lack of equivalence.
2. Optional transposition. An example of optional transposition
is the English expression ‘As soon as he gets up’; if it were
translated back into French, it would be translated as: ‘Dès son
lever’, or ‘Dès qu’il se lève’. Hence, it is optional to employ either
transposition strategy (i.e. ‘Dès son lever’) or calque strategy
(i.e. ‘Dès qu’il se lève’). A further example is ‫اﻟﺬﯾﻦ ﯾﺆﺗﻮن اﻟﺰﻛﺎة‬, as it
can be translated as ‘zakat payers’ or ‘those who pay zakat’.
Transposition is similar to Catford’s categorical shifts. Other

Uploaded by S. M. Safi
p g

examples of optional transposition are translating ‘she


screamed when she saw the snake’ as ‫ﻟﻘﺪ ﺻﺮﺧﺖ ﻋﻨﺪ روﯾﺘﻬﺎ اﻟﺜﻌﺒﺎن‬
and ‫اﻟﻤﺆﻣﻨﯿﻦ‬, which may be translated into English as ‘those who
believed’.
E.
Modulation involves the changing of the semantics and point of
view expressed in the SL; this strategy is followed when literal
translation or transposition can result in unidiomatic or
unsuitable text in the TL. Similarly to transposition, there are
obligatory and optional modulations. An example of an obligatory
modulation is the phrase, ‘The time when’, which must be
translated into French as ‘Le moment où’ (literally: ‘the moment
when’). By contrast, optional modulation turns a negative SL
expression into a positive TL expression. In addition, a free
(optional) modulation can only be ixed (obligatory) when referred
to in dictionaries and grammar books. Other examples of
modulation are the French phrase ‘peu profond’, which may be
translated into English as ‘shallow’; ‘lend me your ears’, which can
be translated as ‫‘ ;اﻋﺮﻧﻲ اﺗﻨﺒﺎﻫﻚ‬a piece of cake’, which can be
translated as ‫ ;اﻣﺮ ﻫﯿﻦ او ﺳﻬﻞ‬the translation of ‘ups and downs’ as
‫ ;ﺗﻘﻠﺒﺎت اﻟﺤﯿﺎة‬and translating ‘you are going to have a child’ as ‫ﺳﺘﺼﺒﺤﯿﻦ‬
‫أﻣﺎ‬.
F. Equivalence is a strategy whereby different stylistic or structural
means are used by the SL and TL, respectively, as in idioms and
proverbs. In other words, the ST and TT can render the same
message using different styles or different structures. For example,
the much onomatopoeia of animal sounds, e.g. the sound of a
donkey in English would be transcribed as ‘heehaw’, while in
French it would be transcribed as ‘hi-han’. Most equivalence is of a
syntagmatic nature (i.e. interchangeable); hence, equivalence
mainly comprises a ixed phraseological repertoire of idioms,
clichés, proverbs, nominal or adjectival phrases and so on. For
example, the French proverb ‘Il pleut à seaux/des cordes’ is an
equivalent to the English proverb ‘It is raining cats and dogs’. Vinay
and Darbelnet (2004), however, warn against creating
equivalences or calques without having ready-made equivalences.

Uploaded by S. M. Safi
q q g y q

For example, a translator should not create an equivalent of the


previous proverbs in Arabic because they are not culturally
accepted. Other examples are ‫ﺷﺌﺖ أم أﺑﯿﺖ‬, which can be translated as
‘willy nilly’, and ‘let things slide’, which can be translated as ‫دع اﻷﻣﻮر‬
‫ﺗﺠﺮي ﻓﻲ أﻋﻨﺘﻬﺎ‬. Examples of proverbs are as follows:

Example

ST TT
All that glitters is not gold ً‫ﻟﯿﺲ ﻛﻞ ﻣﺎﯾﻠﻤﻊ ذﻫﺒﺎ‬

A friend in need is a friend indeed ‫اﻟﺼﺪﯾﻖ وﻗﺖ اﻟﻀﯿﻖ‬

G.
Adaptation is the last calque strategy proposed by Vinay and
Darbelnet (2004), and is the changing and/or explaining of cultural
differences between an SL and a TL. This strategy is employed to
create situational equivalence. For example, the English ‘hello’ can
be adapted to be assalamu alikum in Arabic, instead of its linguistic
equivalent ‫أﻫﻼ‬. This strategy is frequently used in translating
literary work. It is also used in translating movies. For example
translating swearwords (e.g. ‘fuck’, ‘damn’) as ‫اﻟﻠﻌﻨﺔ‬. Also, translating
‘boyfriend’ and ‘girlfriend’ as ‫ﺻﺪﯾﻖ و ﺻﺪﯾﻘﺔ‬.
After their discussion of these seven strategies of translation, Vinay
and Darbelnet (2004) conclude that these strategies frequently
overlap, as more than one strategy can be used within the same text or
even the same sentence. For example, as suggested by Vinay and
Darbelnet (2004), the translation of ‘private’ (as would be written on a
door) by ‘défense d’entrer’ can be considered as a simultaneous
transposition, modulation, and equivalence. It is a transposition
because the adjective ‘private’ is transformed into a nominal
expression; a modulation because a statement is converted into a
warning; and an equivalence since it is the situation that has been
translated, rather than the actual grammatical structure. A working
example of the use of the above-mentioned translation procedures can
be found in the following example.

Uploaded by S. M. Safi
Example
The principles guiding the development of information technology
and systems within the federal Government are contained in a
‘Federated Architecture Program’ run by the Treasury Board of
Canada Secretariat. The Committee is, however, concerned that the
child parliament project is run by civil society and therefore is not
provided with adequate support, especially inancial support, to
enable the programme to be sustainable.
‫ﺗﺮد اﻟﻤﺒﺎدئ اﻟﻤﻮﺟﻬﺔ ﻟﺘﻄﻮﯾﺮ ﺗﻜﻨﻮﻟﻮﺟﯿﺎ اﻟﻤﻌﻠﻮﻣﺎت وﻧﻈﻢ اﻟﻤﻌﻠﻮﻣﺎت داﺧﻞ اﻟﺤﻜﻮﻣﺔ‬
‫اﻻﺗﺤﺎدﯾﺔ ﻓﻲ ’اﻟﺒﺮﻧﺎﻣﺞ اﻻﺗﺤﺎدي ﻟﻠﺒﻨﯿﺔ اﻷﺳﺎﺳﯿﺔ‘اﻟﺬي ﯾﺪﯾﺮ ﺷﺆوﻧﻪ ﻣﺠﻠﺲ اﻟﺨﺰاﻧﺔ اﻟﺘﺎﺑﻊ ﻟﻤﺠﻠﺲ‬
‫اﻟﻮزراء اﻟﻜﻨﺪي‬.
‫ ﺗﺸﻌﺮ اﻟﻠﺠﻨﺔ ﺑﺎﻟﻘﻠﻖ ﻷن اﻟﻤﺠﺘﻤﻊ اﻟﻤﺪﻧﻲ ﻫﻮ اﻟﺬي ﯾﺪﯾﺮ ﻣﺸﺮوع ﺑﺮﻟﻤﺎن اﻟﻄﻔﻞ وﻻ‬،‫وﻣﻊ ذﻟﻚ‬
‫ ﻟﻀﻤﺎن اﺳﺘﻤﺮاره‬،‫ ﻻ ﺳﯿﻤﺎ اﻟﺪﻋﻢ اﻟﻤﺎﻟﻲ‬،‫ﯾﺤﻈﻰ ﻣﻦ ﺛﻢ ﺑﺎﻟﺪﻋﻢ اﻟﻜﺎﻓﻲ‬.

As seen in the Arabic translation in the example, some of the strategies


suggested by Vinay and Darbelnet (2004) have been used in the
translation. For example, the ST is reported in passive voice, while the
TT is reported in active voice, which is a modulation. Similarly, the ST
begins with a verb, which is common in Arabic, while the TT begins
with a noun, which is also a modulation procedure.
Also, translating ‘The principles guiding the development of
information technology and systems’ as ‘‫اﻟﻤﺒﺎدئ اﻟﻤﻮﺟﻬﺔ ﻟﺘﻄﻮﯾﺮ ﺗﻜﻨﻮﻟﻮﺟﯿﺎ‬
‫ ’اﻟﻤﻌﻠﻮﻣﺎت وﻧﻈﻢ اﻟﻤﻌﻠﻮﻣﺎت‬is a lexical claque where the ST words were
rendered into Arabic, preserving the syntactic norms of the TL.
Another example of calque is translating ‫ ﻣﺸﺮوع ﺑﺮﻟﻤﺎن اﻟﻄﻔﻞ‬as ‘the child
parliament project’—the ST adjectival word ‘concerned’ was translated
to a verbal phrase (i.e. ‫)ﺗﺸﻌﺮ ﺑﺎﻟﻘﻠﻖ‬, which is a transposition. A further
example that explicates the use of Vinay and Darbelnet’s procedures is
a segment of text from Gibran’s Arabic work The Broken Wings
(translated by Anthony Rizc Allah Ferris):

Example

Target text Source text

Uploaded by S. M. Safi
Target text Source text
…and my beloved, beautiful Selma is dead and ‫و ﺳﻠﻤﻰ – ﺳﻠﻤﻰ اﻟﺠﻤﯿﻠﺔ اﻟﻌﺬﺑﺔ ﻗﺪ ذﻫﺒﺖ إﻟﻰ ﻣﺎوراء‬
nothing is left to commemorate her except my ‫اﻟﺸﻔﻖ اﻷزرق وﻟﻢ ﯾﺒﻖ ﻣﻦ آﺛﺎرﻫﺎ ﻓﻲ ﻫﺬا اﻟﻌﺎﻟﻢ ﺳﻮى‬
broken heart and tomb surrounded by cypress ‫ﻏﺼﺎت أﻟﯿﻤﺔ ﻓﻲ ﻗﻠﺒﻲ وﻗﺒﺮ رﺧﺎﻣﻲ ﻣﻨﺘﺼﺐ ﻓﻲ ﻇﻼل‬
trees. That tomb and this heart are all that is left to ‫ ﻓﺬﻟﻚ اﻟﻘﺒﺮ وﻫﺬا اﻟﻘﻠﺐ ﻫﻤﺎ ﻛﻞ ﻣﺎﺑﻘﻲ ﻟﯿﺤﺪث‬.‫أﺷﺠﺎر اﻟﺴﺮو‬
bear witness of Selma. (p. ix)
‫(اﻟﻮﺟﻮد ﻋﻦ ﺳﻠﻤﻰ ﻛﺮاﻣﻪ‬p. 101)

In the translation above, there is a modulation in translating – ‫و ﺳﻠﻤﻰ‬


‫( ﺳﻠﻤﻰ اﻟﺠﻤﯿﻠﺔ اﻟﻌﺬﺑﺔ‬proper noun + proper noun + adjective + adjective) as
‘my beloved, beautiful Selma’. Ferris opted to delete the repeated
proper noun and to add the adjective ‘beloved’ to translate ‫اﻟﻌﺬﺑﺔ‬. The
translation also exempli ies transposition, as the ST expression ‫ذﻫﺒﺖ إﻟﻰ‬
‫( ﻣﺎوراء اﻟﺸﻔﻖ اﻷزرق‬literally: ‘went to the beyond of the red twilight’)
refers to death. Therefore, the translator rendered it as ‘dead’, which is
a transposition that made the translation lose the aesthetic feature
used in the ST. Similarly, ‫ ﻏﺼﺎت أﻟﯿﻤﺔ ﻓﻲ ﻗﻠﺒﻲ‬was rendered as ‘broken heart’,
which is a transposition.

Exercises
1.
Translate the following text, explaining which of Vinay and
Darbelnet’s strategies were used in translating the text.
New satellite imagery shows that construction on an
experimental reactor is making ‘expeditious’ progress—just
three months after the Kingdom announced plans to build it,
according to former director for nuclear inspections at the
International Atomic Energy Agency (IAEA), Robert Kelley.
Kelley estimated that the reactor could be completed in ‘nine
months to a year’.
The Kingdom has been open about its nuclear program with
the IAEA, which sent a team to Saudi Arabia last July to check on
building plans. It has repeatedly pledged that the program is
peaceful. But Crown Prince Mohammed bin Salman said last year
that ‘without a doubt if Iran developed a nuclear bomb, we will
follow suit as soon as possible’ (Source CNN).
2. Explain the differences between oblique and direct
translation procedures.

Uploaded by S. M. Safi
3.

As a translator, what procedures would you attempt irst in a


given translation task? Why?
4.
Does the employment of speci ic translation procedures
depend on the type of text? How?
5.
Translate the following texts, stating the procedure used in
your translation.
a.
There is a big living room in my house.
b.
We have two ranch hands, who do everything in the ranch.
c.
Amal is my true friend.

2.2.2 Van Leuven-Zwart’s Comparative-Descriptive Model


of Translation Shifts (1989)
Van Leuven-Zwart (1989) proposed a comparative model that aims to
carry out analysis above the level of a sentence. The model is primarily
based on Vinay and Darbelnet’s categorization of direct and oblique
translations, and consists of a comparative model and descriptive
model. The comparative model aims to analyse an ST and its TT at
micro levels, or based on microstructural shifts. Van Leuven-Zwart
divides texts into comprehensible units, which she called ‘transemes’.
For example, ‘I love my parents so much’ is a transeme because it is a
comprehensible unit. Its equivalent transeme in the TL is ‫أﻧﺎ أﺣﺐ واﻟﺪي‬
‫ﻛﺜﯿﺮا‬. The identi ied transeme is compared to what she calls an
‘architranseme’, the invariant principal meaning of the ST transeme,
but does not stand as a full equivalent for the ST transeme. In the
example ‘I love my parents so much’, ‘to love’ is the architranseme.
Then, each transeme is compared with its architranseme and the
relationship between the two transemes is recognized (Munday, 2001).
If the ST and TT transemes are found to be synonymous in relation to

Uploaded by S. M. Safi
the architranseme, then it can be deduced that no shift occurred.
However, if they are found not to be synonymous, then shifts are
assumed to have occurred. The main shifts are modulation,
modi ication and mutation. Within each main category, there are
subcategories. Table 2.1 explicates these three main categories. Let us
consider the following example and its translation for purposes of
clari ication (Table 2.2).
Table 2.1 Transeme and architranseme relationship (based on van Leuven-Zwart’s comparative-
descriptive model of translation shifts, 1989)

ST TT
Transeme Speak English luently ‫اﻧﺎ أﺗﺤﺪث اﻷﻧﺠﻠﺰﯾﺔ‬
Architranseme To speak ‫أﺗﺤﺪث‬

Table 2.2 Main categories of van Leuven-Zwart’s comparative model (from van Leuven-Zwart
1989, pp. 159–169)

Shift De inition
Modulation One of the transemes corresponds with the
architranseme; however, the other differs either semantically or stylistically. In
the previous example, there is a modulation because the ST has an extra word that
does not exist in the TT; that is, luently
Modi ication Both transemes show some form of disjunction (semantically, stylistically,
syntactically, pragmatically, or in some combination of these) compared to the
architranseme
Mutation It is impossible to establish an architranseme, either because of addition, deletion
or some radical change in meaning in the TT

Example
I speak English luently.
‫اﻧﺎ أﺗﺤﺪث اﻷﻧﺠﻠﺰﯾﺔ‬

With regard to the descriptive model, it is a macrostructural model


that aims to analyse the ST. It refers to the three metafunctions of
language: ideational, interpersonal and textual. The model, however,
has drawbacks, as in practice it is dif icult to apply to a long text. Also,
tracking shifts does not seem to be easy.

Uploaded by S. M. Safi
2.2.3 Overt and Covert Translations (House, 1997)
House (1997) views equivalence as a relation between an ST and its
translation. In House’s words, translation is doubly bound: on its ST on
one hand, and on its recipient’s communicative condition, on the other
hand. She adds that absolute equivalence is impossible, and that an
important term that should be discussed is ‘invariance‘, which refers to
dealing with equivalence according to each individual case. Based on
situational dimension and functional equivalence, House differentiates
between two types of translation: overt and covert. Overt translation
focuses on the universal meaning of a text, without addressing the
reader. This kind of translation is employed for translating STs of an
established value. She also considers that the intelligibility of a text
depends on the culture of a text. Hence, according to her, if a text is
indigenous, it needs overt translation, which can be provided through
annotations, insertions, or expansions (Venuti, 2004). This applies to
translating the Holy Quran, prophetic hadiths, president’s speeches
and so on. Overt and covert translations are examples of translation
approaches (or global strategies) that deal with the text at the macro
level. To achieve this, translation strategies (local strategies) are
always employed. In the case of overt translation, ST oriented
strategies are used, such as borrowing, literal translation and the like.
The following is an example of overt translation:

Example

ST TT
The Fatah-led Palestinian Authority in the occupied ،‫اﺗﻬﻤﺖ ﻣﻨﻈﻤﺔ ﻫﯿﻮﻣﺎن راﯾﺘﺲ ووﺗﺶ ﻗﻮات اﻷﻣﻦ‬
West Bank and Hamas authorities in the Gaza Strip ،‫اﻟﺘﺎﺑﻌﺔ ﻟﻜﻞ ﻣﻦ اﻟﺴﻠﻄﺔ اﻟﻔﻠﺴﻄﯿﻨﯿﺔ وﺣﺮﻛﺔ ﺣﻤﺎس‬
routinely arrest and torture peaceful critics and ‫ﺑﺎﺳﺘﺨﺪام اﻟﺘﻌﺬﯾﺐ اﻟﻤﻤﻨﻬﺞ واﻟﺘﻬﺪﯾﺪ واﻻﻋﺘﻘﺎﻻت‬
opponents, Human Rights Watch says. ‫ ﺿﺪ ﻣﻨﺘﻘﺪﯾﻬﺎ وﻣﻌﺎرﺿﯿﻬﺎ‬،‫اﻟﻌﺸﻮاﺋﯿﺔ‬

As seen in this example, the translation preserves the overtones and


undertones of the ST. To maintain the ST features in the translation,
borrowing was employed; for example, borrowing ‘Human Rights
Watch’ as ‘‫’ﻣﻨﻈﻤﺔ ﻫﯿﻮﻣﺎن راﯾﺘﺲ ووﺗﺶ‬. Literal translation was also employed
throughout the text.

Uploaded by S. M. Safi
Covert translation, by contrast, makes translation equal to a ST in
the target culture. In other words, a translated text will appear to be
original and not a mere translation. Thus, in covert translation, the ST
and its culture are not speci ically addressed. The most important
consideration is to convey the ST message in a functional manner. This
approach can be used to translate novels, drama and such texts. The
following is an example of covert translation low:

Example

ST TT

Source: Alice in Wonderland and its translation by Amira Kiwan

In the example above, the ST was adapted in the TL to sound natural


and idiomatic. For example, ‘get very tired’ was idiomatically
translated as ‘‫’ﺑﺪأت ﺗﻀﯿﻖ ذرﻋﺎ‬. Similarly, ‘she had peeped into the book’
was translated metaphorically as ‘‫’أﻟﻘﺖ ﻧﻈﺮة ﺧﺎﻃﻔﺔ‬. Adaptation was used in
the translation to make the TT sound idiomatic.
Based on the distinction between overt and covert translation),
House proposed a quality assessment model that offers criteria with
which to assess a translation. The model, which was revised in 2015, is
based on Halliday’s Systemic-Functional Theory (for details, see
House’s Translation Quality Assessment: Past and Present, 2015). The
model is based on the fact that texts have functions, and those
functions should be conveyed in the translation. Therefore, the ST and
the TT are compared to ind any mismatches between them. These
mismatches can be dimensional or non-dimensional. Dimensional
mismatches result from pragmatic errors that are pertinent to
language users and language use. In contrast, non-dimensional
mismatches are mismatches between the ST and TT at the denotative
level, and they may breach the TL linguistic system. Non-dimensional

Uploaded by S. M. Safi
mismatches are more serious than dimensional ones. The inal
qualitative judgement on the translation will then be based on the
matches and mismatches between the ST and the TT, as the functional
components of the two texts will be compared. In this regard, it is
important to draw the attention of readers to the fact that functional
equivalence is possible only in covert translation. In contrast, overt
translation is always dependent on the SL culture, which makes
functional equivalence dif icult to achieve. Overt and covert
translations are approaches of translation that encompass many
translation procedures or strategies.

Exercise: Examine the STs provided below and their


translations, and explain whether the approach used in the
translation is overt or covert

ST TT
1. The Philippines government has previously ‫ ﻧﻔﺖ اﻟﺤﻜﻮﻣﺔ اﻟﻔﻠﺒﯿﻨﯿﺔ اﺗﻬﺎﻣﺎت‬،‫وﻓﻲ وﻗﺖ ﺳﺎﺑﻖ‬
dismissed claims of human rights abuses, saying ‫ وﻗﺎﻟﺖ إن اﻟﺮﺋﯿﺲ‬،‫ﺑﺎﻧﺘﻬﺎك ﺣﻘﻮق اﻹﻧﺴﺎن‬
President Duterte had employed ‘lawful use of ‫دوﺗﯿﺮﺗﻲ اﺳﺘﻌﻤﻞ ‘اﻻﺳﺘﺨﺪام اﻟﻤﺸﺮوع ﻟﻠﻘﻮة’ ﺿﺪ‬
force’ against threats to the country. Eritrea has also ‫ ﻛﻤﺎ ﻧﻔﺖ‬،‫اﻟﺘﻬﺪﯾﺪات اﻟﺘﻲ ﺗﺘﻌﺮض ﻟﻬﺎ اﻟﻔﻠﺒﯿﻦ‬
strongly denied such allegations, and insists that it
‫ وﺗﺼﺮ ﻋﻠﻰ أﻧﻬﺎ‬،‫إرﯾﺘﺮﯾﺎ ﺑﺸﺪة ﺗﻠﻚ اﻻﺗﻬﺎﻣﺎت‬
treats its citizens well
‫ﺗﺤﺴﻦ ﻣﻌﺎﻣﻠﺔ ﻣﻮاﻃﻨﯿﻬﺎ‬

Exercise
What is the difference between overt and covert translation?
1.
In what types of text are we likely to adopt the overt translation
approach? Why?
2.
In what types of texts are we likely to adopt the covert
translation approach? Why?
3.
Do you believe that overt and covert translation approaches lie
at the extreme ends of a range or on a continuum? Why? Why
not?
4. House (1997, 2001, 2015) proposed a translation quality

Uploaded by S. M. Safi
assessment model: explain this model. Do you think that this
model can be used to assess any translation? Why? Why not?

2.2.4 Jakobson’s Equivalence (1959)


Jakobson (1959), a Russian linguist who studied linguistic meaning
and equivalence in meaning between different languages, observed
many differences among languages. He stated that meaning of any
linguistic sign (i.e. word) can be considered a further translation of this
sign. For example, the word ‘bachelor’ can be converted into a more
explicit sign, such as unmarried man. Jakobson differentiated between
three types of translation: intralingual, interlingual and intersemiotic.
1.
Intralingual translation refers to rewording, using signs of the
same language; in this type of translation, another less or more
synonymous word is used or circumlocution is employed in the
absence of a synonym. Jakobson further mentions that each code
unit (i.e. word) should be translated by an equivalent combination
of code units; for example, every ‘bachelor’ is an ‘unmarried man’,
and every ‘unmarried man’ is a ‘bachelor’. In intralingual
translation, a word is replaced by another that is a near synonym
or near equivalent. For example, celibate and bachelor can be near
synonyms, but they are not complete synonyms because every
celibate is a bachelor, but not every bachelor is a celibate. This
applies to Arabic as well. A case in point would be ‫ ﺿﯿﺎء‬and ‫ﻧﻮر‬,
which are near synonyms but they are not complete synonyms. The
irst word is always used to refer to the light accompanied by heat,
while the latter refers to light without heat. These nuances in
meaning, however, may not be realized by laymen language users.
In Arabic, teachers of translation attempt to develop the skill of
making their students able to choose the right word, which is not
an easy task because Arab students use their own local dialects in
everyday language use and, therefore, using standard Arabic is
thought-provoking. Students frequently know how to translate an
ST into their local dialect, but they ind it arduous to translate their
local dialect into standard Arabic.
2. Interlingual translation refers to replacing a verbal sign with
another sign but of a different language; on this level of translation

Uploaded by S. M. Safi
another sign but of a different language; on this level of translation,
there is no full equivalence between code units. Translation
substitutes only messages but not code units. Similarly, on the level
of interlingual translation, there is no full equivalence between
code units, while messages may serve as adequate interpretations
of alien code units or messages. For example, the word ‘cheese’ in
English does not have a complete equivalent in the Russian
language. Since languages converge in some linguistic features and
diverge in others, it is common to ind complete equivalents for
some lexemes, phrases and clauses in some cases but not in others.
Most universal lexis (e.g. love, hate, play, laugh) and expressions
(e.g. take upon one’s shoulders) have equivalents across many
languages. For example, ‘take upon one’s shoulders can be
translated into Arabic as ‫ﯾﺄﺧﺬ ﻋﻠﻰ ﻋﺎﺗﻘﻪ‬. However, some other
expressions and lexis are culture-bound and, therefore, equivalents
do not exist. Take for example the English expression ‘baby
shower’, which does not have an equivalent expression in Arabic.
However, some expressions that sound culture-bound may not be
and may have equivalents in other languages. For example, though
the word ‘baptism’ sounds culture-bound, it has an equivalent in
Arabic as ‫اﻟﺘﻌﻤﯿﺪ‬.
3.
Intersemiotic translation or transmutation refers to transmuting
verbal to non-verbal signs. In intersemiotic translation, the focus is
on the message more than wording (Jakobson, 1959/1966/2000).
To clarify, a text (verbal sign) may be translated as a picture, or
dancing, or any other type of performance (non-verbal sign). This
applies to particular types of text, such as the translation of
advertisements. So, intersemiotic translation implies a kind of
creativity on the part of the translator and, therefore, a single text
can be translated creatively and differently by different translators.
Jakobson stresses that full equivalence between any two linguistic
codes (i.e. words) is not achievable (Jakobson, 1959/2000). He does
not view translation as impossible; however, he argues that there are
linguistic limitations that make full equivalence impossible. Jakobson’s
views are similar to Vinay and Darbelnet, in that he considers
translation is possible in spite of cultural and linguistic limitations.

Uploaded by S. M. Safi
Exercise: How would you translate the following words
between English and Arabic? And into which type of
Jakobson’s types of translation do they fall?
1.
‫اﻟﺘﻘﻮي‬
2.
‫ﻋﯿﺪ اﻟﻔﻄﺮ‬
3.
‫اﻟﺤﺞ‬
4.
‫اﻟﺮؤﯾﺔ اﻟﺸﺮﻋﯿﺔ‬
5.
‫رؤﯾﺔ اﻟﻬﻼل‬
6.
Baby shower
7.
Halloween
8.
State of Union Speech

2.2.5 Dynamic Equivalence and Formal Equivalence


(Nida, 1964)
In 1964, Eugene Nida proposed his new notion of equivalence, which is
considered the irst attempt to differentiate between pragmatic
equivalence, on the one hand, and linguistic and cultural (i.e. formal)
equivalence, on the other hand. Nida presented two new types of
equivalence; dynamic (which he later ‘functional’) and formal
equivalence (Munday, 2008). Nida developed dynamic equivalence
Bible translation theory. He proposed his own scienti ic approach to
dealing with meaning, equivalence and translatability. His theory is
based on theoretical concepts and terminology from semantics and
pragmatics, and from Chomsky’s work on syntactic structure.
According to Nida, a word acquires its meaning through context and
can create varying responses according to culture (Munday, 2008).

Uploaded by S. M. Safi
Nida divides equivalence into two types: formal equivalence (or
formal correspondence) and dynamic equivalence.
1.
Formal equivalence or formal correspondence is inclined to be
more ST structure oriented. It is more concerned with the message
in the TL, but it should match as closely as possible the different
elements in the SL (Nida, 2000). This type of translation is called
‘gloss translation’, in which the translator seeks to produce the
closest approximation of the original. This translation requires a
reader to have moderate familiarity with the subject matter
(Shakernia, 2014). Yet, this type of translation has a drawback as it
requires several footnotes to make the text fully understandable.
This type of translation orients the target audience more towards
the SL culture (Munday, 2008; Panou, 2013).
2. Dynamic equivalence is more concerned with the effect of the
principle equivalent, where the relationship between the receptor
and the message should be signi icantly the same as that which
existed between the original receptors and the message (Nida, in
Venuti, 2004.). Dynamic translation is receptor oriented and,
therefore, aims at complete naturalness of expression, which
requires adaptations of grammar, lexicon and cultural references.
Moreover, in dynamic equivalence, the foreignness of the ST is
reduced to an extent that this method was criticized by culture
oriented theorists (Munday, 2008). Nida proposed certain
techniques that could be applied for the purposes of adjustment in
the translation of texts: additions, subtractions,
(1) Additions: Nida postulates that additions are legitimate in
translation, and he states that additions can be of many types
and for many purposes:
Filling out elliptical expressions: it may be necessary to
render some ST ellipted expressions into explicit
expressions to clarify the ST meaning and to avoid
ambiguity.
Obligatory speci ication: this may be needed to clarify
misleading references; for example, pronouns.
Additions required by grammatical restructuring: this may
occur when voice word class and reportedness are

Uploaded by S. M. Safi
occur when voice, word class, and reportedness are
changed or restructured in the TT. For example, changing
indirect speech to direct speech, or passive voice to active
voice, is likely to result in ampli ications and additions.
Ampli ication from implicit status to explicit status: some
implicit semantic elements may be needed to be rendered
explicitly, which results in additions
Answers to rhetorical questions: sometimes rhetorical
questions should be translated by providing answers to
them in the TT, due to differences between an SL and a TL.
Classi iers: sometimes classi iers are a part of language,
though they are redundant and their translation should
observe that. For example, translating Sohag into ‫ﻣﺤﺎﻓﻈﺔ‬
‫( ﺳﻮﻫﺎج‬Sohag governorate).
Connectives: repetition of segments of a preceding text is
sometimes a part of the nature of language, and they help
keep the text coherent.
Categories of the TL: when a TT has certain categories that
are obligatory or optional, it is necessary to add the
obligatory categories in the translation and decide on the
optional ones. For the optional categories, a translator may
need to make explicit what is implicit in a ST, and this can
be done by using doublets (using semantically
supplementary expression).
(2) Subtractions: they can be applied to many cases, such as
repetitions, speci ication of reference, conjunctions,
transitionals, categories, vocatives, and formulae.
a.
Repetitions: semantic supplementary expressions may be
tautological in some languages and thus deleting one of
the two expressions may be necessary; for example, in
translating ‘answered and said’ as ‘answered’—deleting
‘said’, which is redundant. However, repetitions that serve
a speci ic purpose, such as ‘emphasis’, should not be
deleted in the translation.
b. Speci ication of reference: some languages do not require
repetition of a speci ication of reference; therefore,

Uploaded by S. M. Safi
p p ; ,

c. translating them should observe that.


Conjunctions: conjunctions in some languages are used in
a way that is different from the language into which a
piece is being translated. For example, conjunctions in
Arabic may not need to be rendered into English.
d.
Transitionals: they mark the transition from one unit to
another. For example, in the Arabic language, ‫‘( ﻫﺬا‬this’)
marks a transition from one story to another and thus, in
translation, it may be deleted.
e.
Categories: some categories can be deleted in a
translation. It is not necessary to re lect all categories in a
translation. For example, English sometimes uses the
plural form with proper nouns to refer to the family of a
speci ic person, which may not be needed to be re lected
when translated into Arabic.
f.
Vocatives: Arabic, for example, makes use of the vocative
particle, which may be deleted when translated into
English.
g.
Formulae: some formulae may not be rendered into a TT;
for example, translating ‫ أﺻﺤﺎب اﻟﻔﺨﺎﻣﺔ و اﻟﺴﻤﻮ‬as ‘their
excellencies’.

3.
Alterations: Nida argues that any acceptable translation must
undergo alterations in sounds, categories, word classes, order,
clause and sentence structure, semantic problems involving single
words and semantic problems involving exocentric expressions.
4.
The use of footnotes: they are used to explain cultural and
linguistic differences between an SL and a TL, and to add
information that is necessary for historical and cultural
understanding of the ST concepts.
5. Adjustments of language to experience: a translator may need an

Uploaded by S. M. Safi
old use of language to a modern use of language based on
experiences.

After having discussed the techniques that may be need for


adjustment purposes, Nida expounded the translation procedures that
a translator needs to employ in his process of translation. These
procedures are broadly divided into technical and organizational
procedures. Technical procedures entail three phases: analysis of the
SL and the TL, careful study of the SL text, and determination of the
proper equivalents. Since the irst phase is clear, I will now discuss the
other two phases of technical procedure, which relate to:
1.
Analysis of the lexico-grammatical features of the immediate unit:
this includes analysis of all the semantic aspects of a ST: the
syntactic, referential and emotive.
2.
Discourse context: it is important to understand any stretch of
language in relation to its speci ic and general context.
3.
Communicative context: it is important to analyse the
communicative context of an ST. this may include the background
of the ST (e.g. the author), the way it is written, the factual
background of the ST and the circumstances that contributed to
the production of the ST.
4.
Cultural context of the SL.
5.
Cultural context of the TL.
Determination of equivalents is the second phase of technical
procedure and can be conducted by decomposing the ST into its
simplest semantic structure and recomposing such semantic
structures into their nearest equivalents in the TL.
Overall, Nida stresses that correspondence in meaning must have
priority over correspondence in style, if an equivalent effect can be
achieved (Panou, 2013). Nevertheless, Nida’s theory was criticized by a
number of translation theorists, such as Van den Broeck and Larose,

Uploaded by S. M. Safi
who questioned the measurability of the equivalent effect (Shakernia,
2014). Whang (2004) also criticized Nida’s theory by proposing the
same question, which is how a translator can ascertain that his
translation can induce the same effect as the SL text. Nida’s theory is
also rejected by religious groups, who maintain that the word of God is
sacred and unalterable; hence, making necessary changes to attain
dynamic equivalence is unacceptable (Gentzler, 2001).

Exercise
1. Translate the text below and explain into which of Nida’s
translation approaches your translation falls. Support your
answer with examples from your translation.
Saudi Arabia detained seven activists, including two US
citizens, on Thursday, sources tell CNN. It was the kingdom’s
irst sweep of arrests targeting dissidents since the killing of
journalist Jamal Khashoggi last year.
A State Department of icial con irmed to CNN on Friday that
two US citizens were arrested in Saudi Arabia, but declined to
provide names.
‘We can con irm that two US citizens were arrested in Saudi
Arabia’, the of icial said. ‘We have already engaged the Saudi
government in this regard. Due to privacy considerations, we
have no further comment.’
Salah al-Haidar, a dual Saudi-US citizen, who is the son of
prominent women’s rights defender Aziza al-Yousef, was one of
those arrested, according to two sources familiar with the
events. Yousef was temporarily freed from a prison in Riyadh
last month and is on trial along with 10 other women’s rights
defenders (Source CNN).
1.
Explain which of Nida’s types of equivalence is more
frequently applied by translators, including yourself. Why?
2.
One concept that was proposed by Nida is ‘principle
equivalent’. What does this concept mean? Was this concept
accepted by translation theorists? Do you think that this
concept is practical and achievable?

Uploaded by S. M. Safi
2.2.6 Communicative and Semantic Translation
(Newmark, 1981, 1988)
Newmark’s theory of translation is pertinent to the linguistic theory of
translation, as he follows Catford’s formal equivalence and dynamic
equivalence; however, he calls them ‘semantic translation’ and
‘communicative translation’ (As-Sa i, 2011). In translation, Newmark
(1988) underscores the importance of referring to the textual level,
referential level, cohesive level and the level of naturalness at the
process of translation. Textual level implies transposing the SL
grammar and lexis into TL equivalents. Referential level refers to
keeping in mind the referential level of a sentence during translation
(i.e. what each sentence or word means in context). Thus, referential
level and textual level should go hand-in-hand. Cohesive level denotes
the smooth moving from one idea to another without breaking the
textual cohesion, so that a translator could observe the structure and
moods of a text. Naturalness is how to make a TT appear to be an
original text. However, it is beyond doubt that Newmark’s levels cannot
be applied to any text. For example, a sacred text such as the Holy
Quran is not expected to be translated into a naturally equivalent TT.
Understanding Newmark’s four levels gives in-depth understanding as
to how the translation process should take place.
Newmark (1981), in his seminal work: Approaches to Translation,
differentiated between two main types of translation: communicative
translation and semantic translation. Communicative translation is
TT oriented, whereas, semantic translation is source text oriented.
Communicative translation attempts to produce a similar effect on its
readers to that of the original text, whereas semantic translation
attempts to render as closely as possible the semantic and syntactic
structures of the second language to allow the exact contextual
meaning of the ST. Thus, as Newmark argues, semantic translation is
more detailed and complex, and tends to over-translate to reach the
nuance of meaning in the ST.
Newmark states that, of the different methods of translation (which
will be discussed shortly) only semantic and communicative
translations, out ful il the two aims of translation: accuracy and

Uploaded by S. M. Safi
economy. In addition, a semantic translation is written at the author’s
linguistic level and is used for expressive texts. On the other hand,
communicative translation is written at the readership’s linguistic
level and is used for informative texts. Semantic translation is similar
to Nida’s formal equivalence, as it attempts to provide the semantic
and syntactic structure of the TL to achieve the exact contextual
meaning of the ST. Communicative translation, on the other hand,
agrees with Nida’s dynamic equivalence, as the effect on the TL
audience should be equivalent of that effect of the SL. However,
Newmark rejects the idea of producing an equivalent effect, which was
proposed by Nida, since it is impossible to render the same effect in
terms of space and time (Newmark, 1981/1982). Additionally,
Newmark believes that literal translation is not only the best, but is the
only valid method of translation (Newmark, 1981). Newmark adds that,
if the two forms of translation (communicative and semantic) are in
con lict, then communicative translation should win out.
Newmark (1981) mentions some problems that are faced by
translators. Among these problems is the intention of a translator. In
other words, the intention of a translator affects their translation,
whether they aim to convey the different aspects of an ST, or want to
convey the intended meaning alone. Another problem in translation is
the quality of the writing and the authority of the text. Newmark
(1998) mentions that a well-written text needs a translator to observe
the nuances between words, stating that lexis is the major problem in
translation, and not in grammar. Lexis includes words, collocations and
ixed phrases, neologisms and ‘un indable’ words. He adds that
problems may arise either from a lack of understanding of lexis, or
from inding them dif icult to translate. A lack of understanding of the
lexis of some languages results from a translator’s inadequate
knowledge of the different meanings of a word (i.e. physical, technical,
igurative, or colloquial meanings). The dif iculties in inding
equivalents or translating an ST vary from one text to another. These
variations between texts led Newmark to differentiate between
translation as a scholarship, research, or art. A translation, according to
Newmark, may be considered as scholarship when an SL text is
challenging and demanding, or requires interpretation or additional

Uploaded by S. M. Safi
explanations. Thus, translating the Holy Quran is a scholarship rather
than a profession.
Functions of Texts
Newmark, before discussing the different strategies of translation,
discussed the functions of sentences and the different types of text. He
correlates the functions of sentences to the types of text. Newmark
lists six types of function in sentences: the expressive function, the
informative function, the vocative function, the aesthetic function, the
phatic function, and the metalingual function. The expressive function
relates to the meaning intended by the speaker, writer, or author;
literary texts tend to be a good example of expressive texts. The
informative function relates to facts, reality and knowledge, such as
articles, newspapers, and scienti ic papers. The vocative function is
referred to sometimes as a pragmatic translation, as they are aimed at
the addressee or the readership. A typical example of vocative function
texts includes persuasion, propaganda or publicity writings. The
aesthetic function is concerned with pleasing senses through sounds,
images, or igures of speech; one example of this is translating poetry.
However, in literary texts such as poetry, there is always a con lict
between the aesthetic function and the expressive function.
Newmark (1988), unlike many other scholars, differentiates
between translation methods and translation procedures. According
to him, translation methods deal with the text as a whole, while
translation procedures deal with sentences and the smaller units of
language. He identi ied eight methods of translation: word-for-word,
literal, faithful, semantic, adaptive, free, idiomatic and communicative.
1.
Word-for-word translation: In this method of translation, the SL is
translated into a TL, keeping the same word order and the words
translated singly by their most common meanings, out of context.
This rarely happens between English and Arabic due to the
syntactic and semantic differences between the two languages. It is
mostly valid between close languages.
2.
Literal translation: This is similar to word-for-word translation,
the difference being that SL grammatical constructions are
converted to their nearest TL equivalents.
Faithful translation: This tries to render the closest and most

Uploaded by S. M. Safi
Faithful translation: This tries to render the closest and most
3.
precise contextual meaning to that of the original while, at the
same time, observing TL grammatical structures.
4.
Semantic translation: This is similar to faithful translation;
however, it attempts to keep the aesthetic value of the SL text.
5.
Adaptation: This is the freest form of translation, whereby the SL
is adapted to the TL culture. This method observes the TT culture;
hence, it is applicable in translating poetry and plays
6.
Free translation: This attempts to produce the content of the ST
text without its form. It is usually longer than the original ST
because it paraphrases the ST; that is why it is called interlingual
translation.
7.
Idiomatic translation: This reproduces the message of the original
but may distort nuances of meaning by preferring colloquialisms
and idioms where these do not exist in the original.
8.
Communicative translation: This attempts to render the exact
contextual meaning of the original in such a way that both content
and language are readily acceptable and comprehensible to the
readership.
As seen in the discussion above, some of these methods are source
oriented: word-for-word translation, literal translation, faithful
translation and semantic translation. Others are TT oriented: adaptive
translation, free translation, idiomatic translation and communicative
translation. However, as mentioned earlier, Newmark believes that the
only acceptable methods of translation are semantic translation (ST
oriented) and communicative translation (TT oriented).
In relation to translation procedures, Newmark (1988), in A
Textbook of Translation, proposed several general procedures to
translate from SL to TL. Procedures, unlike methods, deal with the
lowest levels of translation, such as the sentence, clause and word.
These procedures (or strategies) are: transference, naturalization,
cultural equivalent, functional equivalent, descriptive equivalent,
synonymy, thorough translation, shifts or transposition, modulation,

Uploaded by S. M. Safi
recognized translation, translation label, compensation, componential
analysis, reduction and expansion, and paraphrase.
1.
Transference: This is the process of transferring an SL word to a
TL text; this translation process is similar to Catford’s
transference, Vinay and Darbelnet’s borrowing, and Harvey’s
(2000) transcription. One of the methods used to apply
transference is transliteration. When necessary, a functional
equivalent should be added between brackets to clarify the
meaning of some semi-cultural words. For example, translating
‫ ﺟﻬﺎد‬as jihad, or jihad (striving). Other examples include words
such ‫ﻓﻼﻓﻞ ﻛﺸﺮي ﺑﺮﺳﯿﻢ ﻛﻨﺎﻓﺔ اﻟﻘﺎﻋﺪة ﺣﻼل‬, which can be translated as
falafel, koshri, berseem, al-Qaeda and halal. Examples from
English to Arabic include translating supermarket, radio and
mobile as ‫ﺳﻮﺑﺮ ﻣﺎرﻛﺖ رادﯾﻮ ﻣﻮﺑﺎﯾﻞ‬.
2.
Naturalization: This procedure succeeds transference and
adapts the SL word irst to the normal pronunciation, then to the
normal morphology (word forms) of the TL. For example,
translating ‫ ﺳﻌﻮدة‬as ‘Saudization’. Other examples include
translating televise, hallucination and hallucinate as ‫ﯾﺘﻠﻔﺰ و ﻫﻠﻮﺳﺔ‬
‫ﯾﻬﻠﻮس‬.
3.
Cultural equivalent: This involves translating an SL cultural
word by an approximate TL cultural equivalent. This is not an
accurate translation; it is merely an approximation. For example,
translating ‘ups and downs of life’ as ‫ ;ﺻﺮوف اﻟﺪﻫﺮ أو ﺗﻘﻠﺒﺎت اﻟﺰﻣﻦ‬and
‘he hit the nail on the head’ as ‫أﺻﺎب ﻛﺒﺪ اﻟﺤﻘﯿﻘﺔ‬.
4. Functional equivalent: This requires the use of a culture-free
word, with a new speci ic term; it neutralizes or generalizes the
SL word. It is a kind of cultural componential analysis. It is also the
most accurate way of translating (i.e. deculturalizing) a cultural
word. This method can be employed when there is no equivalent
for an SL word. It occupies a middle position between the SL
culture and the TL culture. It can result in under-translation when
the SL word is translated into a TL word (on a one-to-one basis),
or over-translation when translation occurs several times. For

Uploaded by S. M. Safi
example, translating ‘it is raining cats and dogs’ as ‘‫’إﻧﻬﺎ ﺗﻤﻄﺮ ﺑﻐﺰارة‬.
Further examples are as follows:
Example
a.
A piece of cake ‫ﺳﻬﻞ ﺟﺪا‬
b.
It is not my cup of tea ‫ﻻ ﯾﺮوق ﻟﻲ‬
c.
It is a pain in ass ‫ﻣﻮﻟﻢ و ﻣﺤﺮج‬

5.
Descriptive equivalent: This procedure is the explanation of the
meaning of a cultural term in several words. For example,
translating ‫ ﺑﻨﺖ ﻣﺨﺎض‬as ‘she-camel in 2nd year’. Another example
is translating the Arabic word ‫ ﺑﻌﺜﺔ اﻟﻨﺒﻲ‬as ‘The Prophet’s
appointment as messenger’. Other examples are as follows:
a.
‫ اﻟﺘﻘﻮى‬fear of Allah and abiding by His rules
b.
‫ ﻋﻘﯿﻘﺔ‬an occasion when an animal slaughtered in celebration of
the birth of a new-born
c.
‫ ﺟﻨﺎﺑﺔ‬a state when someone has ejaculated and has not yet
done ghusl.
d.
‫ ﻏﺴﻞ‬taking a bath, preferably in a ritual manner, after
ejaculation or having sex.
6. Synonymy: This is the use of a near TL equivalent in a context.
This method is employed when literal translation does not work
properly, or when a lexical item is not important for componential
analysis (i.e. economy precedes accuracy). This method should be
used when a compromise is needed because it can result in poor
translation. For example, translating ‫ ﺣﺞ‬as ‘pilgrimage’ for the
purposes of economy. Other examples include the following
words:
a.

Uploaded by S. M. Safi
a.
‫ ﺻﻼة‬prayer
b. ‫ ﺻﺪﻗﺔ‬optional alms

c.
‫ زﻛﺎة‬compulsory alms.
7.
Thorough translation: This translation procedure is calque or
loan translation. This kind of translation can be applied to
acronyms and international names. However, this translation
procedure should only be used when the term is common and can
be recognized. For example, translating FAO as ‫اﻟﻔﺎو‬. Other
examples include translating UNESCO as ‫اﻟﯿﻮﻧﺴﻜﻮ‬.
8.
Shifts or transposition: This procedure involves changing the
grammar from the SL to that of the TL; for example, changing a
singular word in SL into a plural word in the TL, due to the
syntactic constraints of each language. This procedure of
translation is similar to Catford’s shifts. For example, translating
the Quranic word ‫ أﺻﻮاف‬as ‘wool’. Other examples may include
translating ‘Paradise’ as ‫اﻟﺠﻨﺔ‬. There are many other examples that
exist between English and Arabic. Consider the following
examples:

Example
a.
Ahmed is smart ‫أﺣﻤﺪ ذﻛﻲ‬
b.
I have two blue cars ‫ﻟﺪي ﺳﯿﺎرﺗﺎن زرﻗﺎوﺗﺎن‬

9. Modulation: ‘a variation through a change of viewpoint, of


perspective and very often of category of thought’ (Newmark,
1988, p. 89), which is similar to Vinay and Darbelnet’s
modulation. For example, using positive for double negative, or
double negative for positive. For example, translating the negated
French sentence ‘I1 n’a pas hesite’ to the positive English sentence

Uploaded by S. M. Safi
French sentence I1 n a pas hesite to the positive English sentence
‘He acted at once’. This example indicates optional modulation.
However, sometimes, modulation is mandatory, as in the case
when there is a lexical gap. For example, shallow can be only
translated into ‘peu profond’ due to there being no lexical
equivalent in the French language. In translation between English
and Arabic, modulation is used extensively. For example,
translators tend to render the passive voice in English into active
voice in Arabic, because of the differences between English and
Arabic in terms of passivization. Arabic prefers the use of the
active voice, while English prefers to use the passive voice in
many situations. Consider the following example:
ST: It has been reported by informed sources that the
Egyptian president will run for presidency this year.
TT: ‫و ﻗﺪ أﻓﺎدت ﻣﺼﺎدر ﻣﻄﻠﻌﺔ أن اﻟﺮﺋﯿﺲ اﻟﻤﺼﺮي ﺳﯿﺘﺮﺷﺢ ﻷﻧﺘﺤﺎﺑﺎت اﻟﺮﺋﺎﺳﺔ ﻫﺬا‬
‫اﻟﻌﺎم‬
As seen in the example above, the passive voice in the English
ST was translated into the active voice in the Arabic TT as it
sounds more idiomatic in this form. Other modulation
procedures include: abstract for concrete, cause for effect, one
part for another, reversal of terms, active for passive, intervals
and limits, and change of symbols.
(a)
Abstract for concrete: for example, translating ‘sleep in the
open’ (which is abstract) as ‫( ﯾﻨﺎم ﻓﻲ ﻓﻨﺪق ﺟﻤﯿﻞ‬which is
concrete);
(b)
Cause for effect: for example, translating ‘You’re quite a
stranger’ (which is a cause) as ‘‫( ’اﻧﺎ ﻟﻢ أرك ﻣﻦ ﻗﺒﻞ‬which is an
effect);
(c)
One part for another: for example, translating ‘from cover to
cover’ as ‫;ﻣﻦ أول ﺻﻔﺤﺔ اﻟﻰ اﺧﺮ ﺻﻔﺤﺔ‬
(d)
Reversal of terms: for example, translating ‘health insurance’
as ‫;ﺗﺄﻣﯿﻦ ﻋﻠﻰ اﻟﻤﺮﺿﻰ‬
(e)
Active for passive: see the example given above regarding

Uploaded by S. M. Safi
Active for passive: see the example given above regarding
the Egyptian president;
(f) Intervals and limits (in terms of space and time): for
example, translating ‘I will come back in a minute’ as ‫ﺳﺎﻋﻮد ﻓﻲ‬
‫ﻏﻀﻮن ﻋﺪة دﻗﺎﺋﻖ‬. In this example, the time (‘minute’) was
translated into ‫( دﻗﺎﺋﻖ‬minutes);
(g)
Change of symbols: this can happen in the translation of
ixed expressions; for example, translating ‘you scratch my
back, I’ll scratch yours’ as ‫اﻷﺣﺪ ﻗﺪم اﻟﺴﺒﺖ ﺗﺠﺪ‬.
10.
Recognized translation: this is used for translating of icial and
institutional documents that are accepted of icially by
institutions; for example, translating ‘BBC’ as ‫ﻫﯿﺌﺔ اﻷذاﻋﺔ اﻟﺒﺮﯾﻄﺎﻧﯿﺔ‬, or
translating ‘student support fund’ as ‫ ﺻﻨﺪوق ﻣﻌﯿﻦ‬in the Omani
context.
11.
Translation label: this can be applied to translating new
institutional terms, as a translator attempts to create a new
equivalent term in the TL for a new emerging term in the SL or the
TL. For example, the ST word or acronym ‫ داﻋﺶ‬was irst translated
as ‘ISIS’ and subsequently other translators rendered it as ‘ISIL’.
12.
Compensation: this occurs when a loss of meaning, sound effect,
metaphor, or pragmatic effect in one part of a sentence is
compensated either in another part of that sentence, or in a
contiguous sentence. This procedure can be used in translating
poetry and drama. For example, ‘Parting is such sweet sorrow’, a
line of dialogue from Shakespeare, was translated as ،‫ﻫﺬا وداع اﻟﺤﺐ‬
‫ ﺣﺰن ﯾﻜﺘﺴﻲ إﺷﺮاﻗﺔ اﻷﻓﺮاح‬by Anani.
13.
Componential analysis: this implies the splitting up of a lexical
unit into its sense components, often anything from one up to as
many as four translations; for example, translating ‫ ذاﻛﺮ‬as
‘rememberer of Allah’. Another example is the word ‫ﻣﺘﻘﻲ‬, which can
be translated as ‘fearful of Allah’.
14. Reduction and expansion: this occurs when one lexical item is
translated into more than one item (expansion) such as when

Uploaded by S. M. Safi
translated into more than one item (expansion), such as when
translating ‘linguistics’ into the French ‘science linguistique’; or
two items are reduced to one item (reduction), as when
translating the French ‘science linguistique’ into ‘linguistics’.
Another example is translating ‫ ﯾﺘﻮﺿﺎ‬as ‘take ablution’.
15.
Paraphrase: this is explanation of the ST item(s); for example,
translating ‫ ﺗﯿﻤﻢ‬as ‘The Islamic act of dry ablution using sand or
dust, which may be performed in place of ritual washing if no
clean water is readily available, or if one is suffering from
moisture-induced skin in lammation or scaling’.

After his discussion of these 15 of translation, Newmark proposes


further procedures, which are inferred from the procedures already
presented. These procedures are: couplets, and notes, additions and
glosses.
1.
Couplets: this procedure implies combining two (i.e. couplets),
three (i.e. triplets), or four (i.e. quadruplets) of the previous
procedures to solve one translation problem, and can be used in
translating culturally bound terms.
2.
Notes, Additions and Glosses: these additions can be inserted
within the text between parentheses (brackets); they can be also
added at the bottom of the page, or at the end of the chapter, or
even at the end of the book.

Exercise: Based on Newmark’s procedures, what translation


procedures would you adopt to translate the following lexis
and expressions between English and Arabic
1.
‫ﻛﺄن ﻋﻠﻰ رؤﺳﻬﻢ اﻟﻄﯿﺮ‬
2.
‫ ﯾﻮم ﻟﻚ و ﯾﻮم ﻋﻠﯿﻚ‬.‫اﻟﺪﻧﯿﺎ دول‬
3.
Don’t judge a book by its cover.
4. I will do this when pigs ly.

Uploaded by S. M. Safi
5.
You can’t sit on the fence. You should decide whose side you are
on.

Exercise: Translate the text below, explaining the procedures


you have employed in translating it, based on Newmark’s
procedures
The government seeks to open up Oman’s skies to facilitate more
international airlines and offer more options for travellers.
Air travel between Oman and Turkey could become affordable
after the recent signing of a pact between the two countries at the
International Commission for Air Navigation (ICAN) 2018
Conference in Kenya on air services negotiations, said a statement
from the Public Authority from Civil Aviation (PACA).
The agreement opens up air space to operate any number of
lights between both countries as against the current limit of 28
weekly lights.

Exercises
1.
Explain the following translation dichotomies, including in
your answer examples from your own work.
A.
Overt vs. covert translation;
B.
Semantic vs. communicative translation;
C.
Direct vs. oblique translation;
D.
Formal correspondence vs. dynamic equivalence.
2. Jakobson differentiated between three types of translation.
Explain these types, highlighting the difference between them

Uploaded by S. M. Safi
and the other translation dichotomies proposed by other
scholars.

Exercise: Match Newmark’s communicative and semantic


translation methods, with the translations of the following
English source texts

English STs and their translations Newmark’s


translation
method
1. Ultra-processed foods linked to increased cancer risk
‫ﯾﻌﺘﻘﺪ ﺑﻌﺾ اﻟﻌﻠﻤﺎء أن اﻷﻃﻌﻤﺔ اﻟﻤﻌﺎﻟﺠﺔ ﻗﺪ ﺗﺰﯾﺪ ﻣﻦ ﺧﻄﺮ اﻹﺻﺎﺑﺔ ﺑﺎﻟﺴﺮﻃﺎن‬
2. Theresa May faces ‘meaningful vote’ on her deal.
‫ﺗﻮاﺟﻪ ﻣﺎي ﺗﺤﺪﯾﺎ ﺻﻌﺒﺎ ﻹﻗﻨﺎع أﻋﻀﺎء اﻟﺒﺮﻟﻤﺎن ﺑﺎﻟﺘﺼﻮﯾﺖ ﻷﻧﺴﺤﺎب ﺑﺮﯾﻄﺎﻧﯿﺎ ﻣﻦ اﻻﺗﺤﺎد اﻷورﺑﻲ‬
3. Trumps postpones his State of the Union speech to an unknown date
‫ﺗﺮاﻣﺐ ﯾﺆﺟﻞ ﺧﻄﺎﺑﻪ ﻋﻦ ﺣﺎﻟﺔ اﻻﺗﺤﺎد ﻷﺟﻞ ﻏﯿﺮ ﻣﻌﺮوف‬
4. Two heads are better than one
‫ﻣﺎ ﺧﺎب ﻣﻦ اﺳﺘﺸﺎر‬
5. Egyptians demand that Mubarak open the Rafah crossing-point into
Gaza, break off diplomatic relations with Israel, and even send weapons to
Hamas
‫ وأن ﯾﻘﻄﻊ اﻟﻌﻼﻗﺎت اﻟﺪﺑﻠﻮﻣﺎﺳﯿﺔ ﻣﻊ‬،‫ﻣﻌﺒﺮ رﻓﺢ اﻟﻤﺆدي إﻟﻰ ﻏﺰة‬ ُ
َ ‫ﻣﺒﺎرك‬ ‫وﯾﻄﺎﻟﺐ اﻟﻤﺼﺮﯾﻮن أن ﯾﻔﺘﺢ‬
‫ ﺑﻞ وأن ﯾﺮﺳﻞ ﺑﺎﻟﺴﻼح إﻟﻰ ﺣﻤﺎس‬،‫إﺳﺮاﺋﯿﻞ‬
6. Carl Ti lin and Billy Buck came back in the evening and they all had
supper. After supper, Jody sat by the ireplace and listened to his father
‫ ﺟﻠﺲ ﺟﻮدي ﻗﺮب اﻟﻤﻮﻗﺪ‬، ‫ وﺑﻌﺪ ذﻟﻚ‬.‫ وﺗﻨﺎول اﻟﺠﻤﯿﻊ اﻟﻌﺸﺎء‬،‫رﺟﻊ ﻛﺎرل ﺗﯿﻔﻠﻦ وﺑﯿﻠﻲ ﺑﻚ ﻓﻲ اﻟﻤﺴﺎء‬
‫واﺳﺘﻤﻊ ﻟﻮاﻟﺪه‬

2.2.7 The House, Nida, and Newmark’s Theories in a


Nutshell
As discussed, it seems that most of the theories presented share
certain features. For example, Nida’s functional or dynamic equivalence
is identical to Newmark’s communicative translation, and may sound
close to House’s covert translation. However, House’s covert

Uploaded by S. M. Safi
translation focuses more on the culture of the ST and the TT, rather
than the effect on a reader (Newmark, 2009). Similarly, Newmark’s
semantic translation and House’s overt translation are almost
identical, the only difference being that Newmark places greater
emphasis on the possibilities of literal translations (Newmark, 2009).
Newmark (1991) mentions that texts should be dealt with according to
their nature; for example, the more important and serious the text, the
closer to the ST should be the translation, and vice versa. Most of these
theories, in spite of using variant terms, focus on differentiating
between two main types of equivalence: pragmatic equivalence and
formal equivalence. Pragmatic equivalence aims to communicate the
message of the ST in the norms and culture of the TT, hence making
translation invisible. By contrast, formal equivalence aims to convey
the message of the ST with all of its linguistic and cultural values
(Venuti, 2004).
Overall, most translation approaches are two- or three-poled
theories (Munday, 2008). For example, Catford (1965) identi ies three
ranks of translation—word-for-word, literal and free translation, while
Newmark distinguishes between two major approaches to translation
—semantic and communicative translation (Newmark, 1981).

Exercises
1.
The translation theories mostly revolve around two or three
poles. Explain.
2.
How similar and different are the translation dichotomies
proposed by House, Nida, Jakobson, and Vinay and
Darbelnet?

2.2.8 Form-Based and Meaning-Based Translation


(Larson, 1998)
Larson (1998) identi ies two main kinds of translation: form-based
translation and meaning-based translation. Within these two basic
taxonomies, Larson makes another subdivision in the form of a
continuum that comprises seven kinds of translation ranging from the

Uploaded by S. M. Safi
‘very literal’ translation to the ‘unduly free’. He states that ‘unduly free’
translations are unacceptable translations for most purposes. One
reason for the unacceptability of unduly free translations is that they
add extra information that does not exist in the ST; hence, they change
the meaning presented in the SL text (Larson, 1998). Similarly, he
believes that literal translation is not acceptable because it does not
communicate the meaning; it is a mere string of words translated.
Additionally, Larson mentions idiomatic translation—which
reproduces the meaning of the SL in the natural form of the receptor
language—as the only acceptable translation; it reproduces the
message of the ST in the TT without retaining the form. Although
Catford, Newmark, and Larson use different theoretical terms, these
terms are almost the same in application. An example of literal
translation that is not accepted by Larson is translating ‘Heaven forbid
that he should leave because of me!’ as ‫اﻟﺴﻤﺎء ﺗﻤﻨﻊ أن ﯾﻐﺎدر ﺑﺴﺒﺒﻲ‬. It should,
however, be translated idiomatically as ‫ﻻ ﻗﺪر اﷲ أن ﯾﻐﺎدر ﺑﺴﺒﺒﻲ‬.

Exercises
1.
What is the difference between meaning-based and form-based
translation?
2.
According to Larson, which type of translation is accepted?
Why?
3.
What are the differences between literal translation, free
translation and idiomatic translation? Support your answer with
examples from your own work.

2.2.9 Halliday’s Typology of Equivalence


Halliday (2001) argues that translation equivalence is the central
organizing concept of translation. Halliday proposes his typology of
equivalence in terms of a systematic functional theory. This typology
centres on three vectors: strati ication, metafunction and rank, as
illustrated in Fig. 2.1.

Uploaded by S. M. Safi
Fig. 2.1 Halliday’s parameters of language

Uploaded by S. M. Safi
The irst vector—strati ication, according to Halliday—refers to the
organization of language in ordered strata. Such strata include the
phonetic/phonological, lexico-grammatical, semantic and contextual
levels of the multi-coding system of language. These strata do not carry
the same value in equivalence in translation. For example, semantic
equivalence is more important than lexico-grammatical equivalence.
Hence, each stratum should be valued according to the speci ic
translation task at hand. By way of illustration, let us look at two
examples:

Example
A.
Trump will deliver his State of the Union speech next Sunday.
‫ﺳﻮف ﯾﻘﻮم ﺗﺮاﻣﺐ ﺑﺎﻟﻘﺎء ﺧﻄﺎﺑﻪ ﻋﻦ ﺣﺎﻟﺔ اﻷﺗﺤﺎد ﯾﻮم اﻷﺣﺪ‬.

In example A, equivalence has been achieved at the lexico-grammatical


level as well as the semantic level because the TT retains the same lexis
and grammar of the ST. In the same way, the message and meaning of
the ST is conveyed in the TT.

Example
B.
It is raining cats and dogs.
‫إﻧﻬﺎ ﺗﻤﻄﺮ ﻗﻄﻄﺎ و ﻛﻼﺑﺎ‬.

In example B, equivalence was achieved at the lexico-grammatical level


but not at the semantic level. The lexis and grammar of the ST were
retained in the TT but the meaning was lost. Let us assume that the
translation of example B was ‘‫’إﻧﻬﺎ ﺗﻤﻄﺮ ﻛﺄﻓﻮاه اﻟﻘﺮب‬. In this case, semantic
equivalence was achieved; however, lexico-grammatical equivalence
was not achieved, and this explains why semantic equivalence is the
most important type of equivalence.
With regard to Halliday’s (2001) third vector—which is discussed
irst here due to it having certain similarities with the irst vector—
which is rank, it deals with how the formal strata (i.e. phonology and
lexico-grammar) are organized. In other words, it is concerned with
how clause complexes, clauses, phrases, groups, words and
morphemes are organized. However, rank deals with morphemes,

Uploaded by S. M. Safi
words, clauses and sentences. Similarly, to strata, equivalence in ranks
will differ in value. It is expected that the higher value will be assigned
to the highest formal level (i.e. the clause). Put differently, if clauses are
kept constant or equivalent, it does not matter a great deal if the words
vary. This, however, cannot be considered a rule that can be applied to
all texts. To return to example B: if it is translated as ‫ﻛﺄﻓﻮاه اﻟﻘﺮب إﻧﻬﺎ ﺗﻤﻄﺮ‬,
equivalence is achieved at clause level, but not at word level. Again,
equivalence at clause level is the most important, which, in turn, affects
equivalence at the semantic level.
As for the second vector (i.e. metafunction), it includes three
categories of function that all languages share: ideational,
interpersonal and textual. Ideational function is about the ‘content
function of language’ (Halliday, 2007, p. 183). Ideational function refers
to the use of language to express and talk about our experience of our
inner and outer worlds. In this sense, language is a cording system that
deals with the relation between man and nature. In sum, this function
serves to communicate new or unknown information to the audience.
The ideational function mainly consists of ‘transitivity’ and ‘voice’
(Wang, 2010). The transitivity system is composed of six processes:
material process, mental process, relational process, behavioural
process, verbal process and existential process. Let us consider a
variety of examples that explain these processes.
A. The material process
Material processes relate to doing or making things happen. This is
expressed by different types of verb, especially dynamic verbs (e.g.
play, kill, hit), and is used in the present progressive. So, if a verb does
not allow the progressive aspect, it means that the process cannot be
material. Material processes refer to those processes in which an
action is done. Thus, they include an action verb, actor and goal; for
example, ‘Ahmed is eating meat’ (Zhuanglin, 1988). In this example,
‘Ahmed’ is the actor, ‘is eating’ is the verb, and ‘meat’ is the goal. In this
process, there are six main participants:

Actor is the person who is performing the action;


Goal refers to what is affected by the action;
Scope refers to what remains unaffected by the action;

Uploaded by S. M. Safi
Attribute refers to a quality ascribed or attributed to an entity;
Client refers to whom/what the action occurs, and usually takes the
preposition ‘for’;
Recipient refers to the receiver of goods or services and usually takes
the preposition ‘to’.

Examples of the material process


1.
Ahmed was playing tennis yesterday.
2.
The doctor gave some medicines to Ali.
In example 1, the actor is ‘Ahmed’, while the goal is ‘tennis’. In
example 2, the actor is ‘the doctor’, while ‘Ali’ is the recipient.

B. The mental process


The mental process is more concerned with emotions, feelings,
affection, cognition, perception, or desire. It is realized through the use
of verbs such as ‘believe’, ‘love’ and ‘think’. The participants in the
mental process are ‘the senser’ and ‘the phenomenon’. The ‘senser’ is
the term used to refer to the person who experiences the feelings,
emotions, or experience. ‘Experience’, on the other hand, refers to the
mental process felt or experienced by the ‘senser’. Mental processes
express perceptions or mind-related activities; for example, ‘I love
Egypt’. In this example, the senser is ‘I’, while the experience is the
feeling towards Egypt.
C. The relational process
There are two types of relational process: attributive (e.g. ‘Ali is
clever’) and identifying (‘A horse is an animal’). In a relational process,
the progressive aspect is restricted. With regard to verbal processes,
these include exchanging information processes and they encompass
all the modes of indication or process; for example, ‘Ali told Ahmad that
Ali was absent today’. The participants in the verbal process include the
‘sayer/the addresser’, the ‘receiver/the addressee, the ‘verbiage’; for
example, ‘the charts show a growth in economy’.
D. The behavioural process

Uploaded by S. M. Safi
Behavioural processes refer to psychological or physiological
processes; for example, ‘I breathe’. In behavioural processes, the main
participant is the one behaving, and the second participant (if any) is
the behaviour. These are mostly intransitive situations, and can be
deemed to be a combination of mental and material processes; for
example, ‘the manager gave a black eye to the employee who came late’.
E. The existential process
Existential processes refer to the existence of something and are
usually connected with the verb ‘be’; for example, ‘there is a student in
the class’ (Zhuanglin, 1988).
The process in a nutshell
The existential and behavioural processes usually have only one
participant.
Mental processes are mostly used either in the simple present tense
or the past tense.
The relational process must include two participants. In the
relational-attributive clause, the participants are generally not
reversible, or at least the grammatical functions are ixed while they
are in the relational-identifying clause. Reversibility includes the
exchanging of positions, as well as passivization.
The verb ‘be’ is used as the main verb in relational or the existential
processes.
Sometimes, the language is used iguratively; therefore attention
needs to be paid to the intended meaning; for example, ‘the road
runs along the river’ shows a relational process and not a material
process.
Interpersonal function
Interpersonal function, on the other hand, refers to the use of
language to interact with others, and to establish and maintain
relations with them. It also implies the use of language to in luence
people, to please them, or to anger them. Language, in this sense, is a
medium between individuals (Halliday, 1971). Mood and modality are
typically used to express the interpersonal function. For example, if a
speaker uses an imperative mood, he is assuming that a listener will
obey the command; for example, ‘leave’. Modality embodies the
intermediate ranges between the extreme positive and the extreme

Uploaded by S. M. Safi
negative (Wang, 2010). Modality can express the speaker’s negative or
positive judgement of a topic. Put differently, modality is related in a
direct way to the social functions of language. It can express different
semantic implications, such as permission, request, obligation,
necessity, possibility and so on. Halliday views modality as a form of
participation by the speaker in the communicative act (Mishra, 2009).
Textual function
Textual function refers to how language functions as a system that
organizes messages in a common manner. In this sense, it explains how
the different messages it logically with those around them, and with
the wider context in which the talking or writing is takaing place. For
Halliday (1971, p. 334), ‘Language makes links between itself and the
situation; and discourse becomes possible because the speaker or
writer can produce a text and the listener or reader can recognize one’.
Unlike the previous two vectors, equivalence at the metafunctional
level is not hierarchical: there is no hierarchical relationship among the
three metafunctions. However, Halliday adds that ideational
metafunction has the highest value in translation, in the sense that
translation equivalence is usually de ined in ideational terms, and that
if a TT does not match the ST ideationally, it cannot be considered a
translation. Halliday concludes that a good translation is the text that is
equivalent in regard to the aforementioned linguistic features, which
are the most valued in the given translation context.

Exercises
1.
Halliday proposes a typology of equivalence that is based on
three vectors. What are these three vectors? Support your
answer with examples from your own work.
2. Explain the process involved in the sentences below. Then,
translate these sentences, explaining whether the ST processes
were maintained in the TT.
She’s moving tomorrow.
He was better after undergoing surgery on Saturday.
Record pro its were announced last week.
Record pro its were announced last week.
We’re getting married next year.

Uploaded by S. M. Safi
Next year, the museum is expecting even more visitors.
I was playing golf yesterday.
Yesterday the atmosphere at the factory was tense.
Ali donated ten thousand dollars to the Orphanage House in
Cairo.
My new wife is tall and blonde.
I hate hypocrites.
3.
Translate the following text highlighting the three vectors
proposed by Halliday, and the extent they are maintained in the
translation.
The death of a former president in most countries around the
world would normally make headline news domestically. But not the
case for Egypt, where ex-President Mohammed Morsi died at the age
of 67 on Monday after collapsing in a courtroom during his trial on
spying charges.
His sudden demise barely registered in Egyptian media—in fact,
papers there prioritised Egypt’s hosting of the forthcoming 2019
African Cup of Nations on its front pages, and instead relegated
Morsi’s death to the inside pages usually designated for criminal
affairs.
The state-run channels failed to even mention that Morsi—the
irst democratically-elected leader in Egypt—was a former
president, instead referring to him with his full name. (BBC: last
accessed 19 June 2019).

2.2.10 Catford’s Typology of Equivalence


Catford is a British linguist who based his theory of translation on
those of Firth and of Halliday (Manfredi, 2008). Catford’s book, entitled
A Linguistic Theory of Translation (1965/1978), is his most famous
book in translation. He, following Halliday, deemed language as
working functionally on a variety of levels (i.e. phonology, graphology,
grammar, lexis) and ranks (i.e. sentence, clause, group, word,
morpheme) (Manfredi, 2008). Catford (1965) argues that translation
between any two languages is possible, and that equivalences can exist

Uploaded by S. M. Safi
with any kind of spatial, temporal, social, or other relationship between
them. He states that relations between languages are bi-directional;
however, the translation process is unidirectional (i.e. from ST to TT).
Meaning, as seen by Catford, is the ‘property of language’, in the
sense that each language has its own distinctive meaning. Thus, values
of meaning are not carried over in translation. Catford (1965, p. 43)
states: ‘That is to say, the “values” of TL items are entirely those set up
by formal and contextual relations in the TL itself. There is no carry-
over into the TL of values set up by formal or contextual relations in
the SL’. Catford states that the only condition in which SL meanings can
be carried over into a TT is when using transference which, according
to Catford, is not a translation.
Catford argues (1965, p. 44) that transference can even occur at the
level of grammar, whereby ‘SL grammatical items are represented in
the TL text by quasi-TL grammatical items deriving their formal and
contextual meanings from the systems and structures of the SL, not the
TL’. Such transference implies the superimposition or creation of new
terms that basically belong to the SL. Catford suggests that this can be
done through the use of old English, numbers, or the creation of new
items. However, Catford mentioned that transference does not imply
that the total meaning of the ST will be transferred.
Catford states (1965, p. 50) that SL and TL items can never
linguistically have the same meaning. However, they can function in the
same situation and thus, in total translation, the SL and TL items are
interchangeable in a given situation. Catford states that ‘translation
equivalence occurs when an SL and a TL text or item are relatable to (at
least some of) the same features of substance’. Catford categorizes
translation in terms of extent, levels and ranks. According to Catford,
there are two types of translation in terms of extent (extent refers to
the syntagmatic sense of the SL text that is submitted to translation):
full translation and partial translation. In a full translation, every part of
the SL is translated to the TL; in partial translation some parts of the SL
text are left out in the translated text in the TL, perhaps because they
are untranslatable. Partial translation, as Catford states, is not that easy
as it may seem at irst sight because some parts will remain
untranslatable. This kind of translation applies to literary texts, and

Uploaded by S. M. Safi
surely applies to the translation of canonical and authoritative texts
such as the Holy Quran.
In relation to the levels of language involved in translation, Catford
(1965, p. 22) differentiated between total translation and restricted
translation. Total translation, to quote Catford, is ‘replacement of SL
grammar and lexis by equivalent TL grammar and lexis with
consequential replacement of SL phonology/graphology by (non-
equivalent) TL phonology/graphology’. Thus, according to this
de inition, replacement occurs only between grammar and lexis, while
phonology and graphology are not included. Restricted translation, on
the other hand, is ‘replacement of SL textual material by equivalent TL
textual material, at only one level’. Catford stresses the importance of
using ‘textual material’ in his de inition because not always the whole
ST is translated to TT; sometimes it is only a process of replacement, at
other times simply the transference of SL material into TL text. Thus, in
restricted translation, SL grammar may be translated by equivalent TL
grammar, without replacement of lexis, or SL lexis is translated by TL
lexis, without replacement of grammar.
In terms of rank, Catford classi ied translation according to the
grammatical hierarchy, at which level equivalence is established. For
example, in total translation, equivalence is assumed to be achieved at
every grammatical unit (word, clause, sentence). However, there could
be a rank-bound translation, in which equivalence can only be achieved
at one level. For instance, in word-rank-bound translation, we only
select equivalents at the same rank (i.e. word).
In relation to equivalence, Catford (1965) differentiated between
formal correspondence and textual equivalence. In formal
correspondence, any TL category occupies the same place in the
economy of the TL as the given SL category occupies in the SL. In
textual equivalence, any TL text (or portion of text is deemed to be
equivalent to a given SL text (or portion of text). The following is an
example of formal correspondence:

Example
Democrats are trying to oust Trump in 2020.
2020 ‫ﯾﺤﺎول اﻟﺪﯾﻤﻘﺮاﻃﯿﻮن اﻹﻃﺎﺣﺔ ﺑﺘﺮاﻣﺐ ﻓﻲ‬

Uploaded by S. M. Safi
In this example, the TT occupies the same place in the economy of the
TL as the given SL category occupies in the SL. The lexis used in the TT
is even less than the lexis used in the ST, conveying the same meaning
intended in the ST. The following is an example of textual equivalence
below:

Example
I am 20.
‫ ﻋﺎﻣﺎ‬20‫أﺑﻠﻎ ﻣﻦ اﻟﻌﻤﺮ‬

In this example, the TT underwent a shift, as some words were added


to clarify the meaning; however, the meaning of the TT was equivalent
to the meaning of the ST.
In general, when formal correspondence and textual equivalence
diverge, a ‘translation shift’ takes place. The term ‘translation shift’ was
irst introduced by Catford (1965) (Ni, 2009), who introduced the term
‘shift’ to replace the thorny term ‘equivalence’. Shifts are the process of
departing from the formal correspondence in the process of going from
the SL to the TL. Shifts—which can be in lexis, style, or grammar—are
able to provide translation that is pragmatic, functional and
communicative. Catford states that it is impossible for translation to
occur between the levels of phonology and graphology, or any of them,
on the one hand, and grammar and lexis, on the other hand. He states
that ‘relationship to the same substance [is] the necessary condition of
translation equivalence’ (Catford, 1965, p. 141). The only possible
shifts are from grammar to lexis and vice versa. Catford proposed two
kinds of shift: level and category. A level shift refers to the proposition
that something that is expressed by a linguistic level in one language
(e.g. grammar) can be equivalently expressed at a different linguistic
level (i.e. through vocabulary or different grammar) in another
language. For example, the imperfect verb in Arabic (e.g. ‫ )ﯾﺘﻨﺎﻫﻮن‬is
mostly translated into past simple or past continuous in English (e.g.
‘forbade each other’). Another example is translating the English
present progressive into lexis such as ‫اﻻن‬. Consider the following
example:

Example

Uploaded by S. M. Safi
What are you doing? ‫)ﻣﺎذا ﺗﻔﻌﻞ )اﻻن‬
I am watching TV. ‫أﺷﺎﻫﺪ اﻟﺘﻠﻔﺰﯾﻮن‬

In this example, the continuity aspect can be only translated by adding


the word ‫اﻻن‬, either in the question or its answer (for more details,
refer to the translation of tense in Chapter 4).
Category shifts are divided into four types: structural shifts, class
shifts, unit or rank shifts and intra-system shifts. Structural shifts
imply a change of grammatical structure; for example, in a translation
between English and Arabic, there is often a shift from AMH (article +
modi ier + head) to AHM (A + head + quali ier); for example, ‘The White
House’ (MH) is translated into ‫( اﻟﺒﯿﺖ اﻷﺑﯿﺾ‬MHQ). Due to the syntactic
differences between English and Arabic, there are always structural
shifts in the translation of most texts.
Class shifts include a change of a part of speech, which could occur
as a part of a structural shift. For example, an adjective in the ST may
have a noun as its equivalent in the TT; for instance, translating ‘a
medical student’ into Arabic as ‫ ﻃﺎﻟﺐ ﻃﺐ‬or ‫ﻃﺎﻟﺐ ﻓﻲ ﻛﻠﯿﺔ اﻟﻄﺐ‬. The class
shift occurred from the adjectival word ‘medical’ into the noun word
‫ﻃﺐ‬, or to the adverbial clause ‫ﻓﻲ ﻛﻠﯿﺔ اﻟﻄﺐ‬. Similarly, the noun ‫ اﻟﺤﻖ‬can be
translated to the adjective ‘the real’, and the verb ‫آﻣﻨُﻮا‬
َ can be rendered
as a noun; for example, ‘believers’.
Unit shifts or rank shifts include replacing units of different size,
such as a sentence, clause, group, word or morpheme. To clarify, a word
may be translated into a sentence or phrase in the TL. A case in point
would be translating the ST word ‫ أﻋﺘﻜﺎف‬into a string of words; for
example, ‘staying in the mosque for a speci ied period of time as an act
of worship’.
With regard to intra-system shifts, these occur when an SL and TL
have roughly the same systems, but the translation involves choosing a
non-corresponding item in the TL (Catford, 1965). For example,
English and French have the same system with regard to plurality
(singular vs. plural); however, in translation a singular English word
may be translated into a plural one or vice versa. A case in point is
translating the singular English word ‘advice’ into the plural French ‘des
conseils’, or the plural English word ‘trousers’ into the singular French
‘le pantalon’. Another case of the intra-system shift is the article system

Uploaded by S. M. Safi
in English and French. Although, the two languages share the same
system of articles, this is not the case in translation. Similarly, Arabic
and English share some features; however, in translation, a translator
may opt to translate the ST item into a non-equivalent item in the TL.
This can happen so as to maintain idiomaticity in the TL. For example,
the English sentence ‘He is a teacher’ is likely to be translated into
Arabic as ‫ﻫﻮ ﻣﺪرس‬, where the inde inite article is not translated. Catford
states that it is linguistically dif icult to give a TL and SL the same
meaning. Yet, we can consider two items in the SL and the TL as
equivalents when they are able to function in the same situation. In a
total translation, the items in the SL and the TL should be
interchangeable in a given situation. Another example is translating
‫ اﻟﻌﻠﻤﺎء ورﺛﺔ اﻷﻧﺒﯿﺎء‬into ‘scholars are inheritors of prophets’, whereby the
de inite article in Arabic was left out in the TT. A common example of
intra-system shifts is the passive case, whereby the passive voice in
English is often translated into the active voice in Arabic.

Exercises
1.
What is the difference between level shifts and category shifts?
2.
Can a translation of a single phrase or lexeme convey the use of
more than one type of shift?
3.
Translate the following text into Arabic, explaining the
translation shifts employed in your translation.
For decades, he was known as a godfather of excess. The
wealthiest man in the world for many years, the Sultan of Brunei
knew how to spend the vast riches that lowed from the oil deposits
bestowed upon the tiny Southeast Asian nation he controls with
absolute power.

2.2.11 Mona Baker’s Typology of Equivalence


Mona Baker’s typology of equivalence is outlined in her seminal work
In Other Words, in which she discussed the different problems of
equivalence in translation between any two languages. Equivalence has

Uploaded by S. M. Safi
always been identi ied as a central component of most of the
de initions of translation (e.g. Catford, 1965; Nida, 1959; Wilss, 1982).
However, other theorists avoided using the word ‘equivalence’ (e.g.
Frawley, 1984; Jakobson, 1959). According to Baker (2004), the notion
of equivalence can be de ined either normatively (i.e. the relation
between source elements and target elements that are assumed to be
achieved), or descriptively (i.e. discovering a relation of equivalence
between the source and target elements).
Baker (2004) argues that the notion of equivalence is problematic
due to its being circular—circular, here, in the sense that we de ine
translation in terms of equivalence, and we assess the quality of
translation in terms of equivalence. Baker, however, underscores the
importance of this notion due to its interrelatedness with other
theoretical notions in the ield of translation. For example, faithfulness
to the original is related to the desirability of equivalence. Also, the
notion of ‘shift’, which is an important notion in normative approaches,
is based on an assumption of equivalence that may or may not occur.
Shift, as a notion, postulates the existence of an ‘invariant’. Invariant
refers to the extent of closeness to achieving equivalence in
translation; invariants ‘are not or should not be affected by shifts in the
process of translation’ (Baker, 2004). A further notion that is related to
equivalence is the notion of the ‘translation unit’. Translation units are
usually discussed in relation to what units (words, clauses, phrases,
sentences and so on) are to be considered as equivalents, or what
translators in real life work with to produce an ‘equivalent’ version of
the ST (Baker, 2004). Baker concludes that the notion of equivalence is
so important because other theoretical notions of translation are
interlinked with it; and that this is why the notion of equivalence
should not be discarded or discredited. However, one question that
may surface is which perspective of equivalence should be considered
the most appropriate in translation. There are many perspectives of
equivalence. Equivalence can be regarded either as a semantic
category, in terms of the equivalence effect, or in terms of functional
equivalence.
Baker (2004) explains that the notion of equivalence as a semantic
category, which is drawn from the representational theory of meaning,
is static and close to the interlingual synonyms. It is dictated by the

Uploaded by S. M. Safi
content of the ST, rather than the communicative situation. This
semantic view of equivalence, as Baker states, is rejected in most
disciplines, and it is not applicable or tenable in translation. Another
understanding of equivalence can be in terms of the ‘equivalent effect’,
which postulates producing the same effect on target readers as the ST
produced on its readers. This approach originated with translators of
the Bible (Beekman and Callow, 1974; Larson, 1998; Nida, 1964; Nida
and Taber, 1969). This notion of ‘equivalent effect’ resulted in the
existence of other notions, such as ‘receptor’ as opposed to ‘target’
language, and dynamic equivalence as opposed to ‘formal equivalence’
(Baker, 2004). Although this notion of ‘equivalent effect’ sounds
interesting and easier than the semantic notion of equivalence, it was
also subject to much criticism. Baker (2004) questions the
measurability of achieving equivalent effect. In addition, the effect is
variable among different people and even a person may perceive the
same TT differently on a second reading. This notion of equivalent
effect seems to be imaginary: a translator cannot predict the effect of
his translation on its readers. Another problem with this notion, as
mentioned by Baker, is that a translator cannot identify with certainty
the intention of the author of the ST, especially in the case of a temporal
gap between the ST and the TT. Another point is that a translator’s job
is to interpret text, rather than understand it. Baker concludes that this
notion can be hardly veri ied.
Another notion of equivalence is ‘functional equivalence’, which
arose in the 1970s and 1980s (Baker, 2004). This notion postulates
that translation should produce an ‘equivalent message’ to that of the
ST in its TT. In the 1980s, a new notion of equivalence emerged,
especially in Germany: the functional equivalence of skopos. Skopos
was established by Vermeer and Reiss, according to which they regard
the target of the translation as what matters (see this chapter, for
details). Baker concludes that there has been a gradual shift away from
the notion of equivalence over the course of time. Baker (2004)
summarizes the debate on the notion of equivalence shifted away in
Table 2.3.
Table 2.3 The gradual erosion of the notion of equivalence in translation studies

Source text/target text Same meaning

Uploaded by S. M. Safi
Source text/target text Same effect on respective readers
Source text/target text Same function
Target text Independent function, speci ied by commission
Target text Independent function acquired in the situation in which it is received

Source Baker (2004)

Baker (1992) identi ied various types of equivalence: equivalence


at word level, equivalence above the word level, textual equivalence
and grammatical equivalence.
1. Equivalence at word level
The written word, as de ined by Baker (1992/2005), is any
sequence of letters with an orthographic space on either side. Baker
rejects the idea that the word is the smallest unit of meaning; she
argues that meaning can be carried by more or less than a word; for
example, the ‘-er’ in builder has a meaning (i.e. the person who does the
job of building). Baker states that there is no one to one
correspondence between orthographic words and their meanings,
either within the same language or across languages.
2. Non-equivalence as a problem
Vocabulary, as seen by Baker (1992/2005), is a set of words that
belong to semantic ields. These semantic ields are abstract concepts.
However, one problem with these semantic ields is that, in terms of
categorization, they are not that simple. For example, there are some
words (e.g. ‘just’, ‘only’) that can be iled under any semantic ield. Baker
states that semantic ields can only work well with words that have
propositional meanings. In relation to the importance of semantic
ields in studying translation, Baker states that understanding the
structures of semantic ields is important in translation for two
reasons: the irst reason is either to assess the value of a given item in
a lexical set, or to understand the differences between the structuring
of semantic ields in the ST and TT; the second reason is to understand
the hierarchical classi ication of words in terms of hypernyms and
hyponyms.
According to Baker (1992), it is important to distinguish between
lexical items and units of meaning to achieve good translation.
Meanings, furthermore, differ in the orthographic words that represent

Uploaded by S. M. Safi
them from one language to another. A meaning of one orthographic
word in one language may be represented by several orthographic
words in another language, and vice versa. For instance, ‫ ﻛﺴﻮف‬and
‫ ﺧﺴﻮف‬in Arabic have only one equivalent representation in English:
‘eclipse’. Another example is the English word ‘camel’, which is
represented by many words in Arabic (e.g. ‫ ﺑﻨﺖ ﻟﺒﻮن‬،‫ زاﻣﻠﺔ‬،‫ ﻧﺎﻗﺔ‬،‫ﺟﻤﻞ‬, among
others) (AL-Maani Online Dictionary, n.d.). Consequently, this means
that there is no one-to-one correspondence between orthographic
words and elements of meaning within or across languages. As
mentioned earlier, Baker discussed equivalence at a variety of levels;
these concepts are unpacked in the following sections.
3. Non-equivalence at the word level
Equivalence is a crucial notion in translation between any two
texts. However, there are many causes that contribute to the problem
of a lack of equivalence. Baker categorizes the most common non-
equivalences between languages at the word level into 11 types:
cultural speci ic concepts, SL concepts are not lexicalized in the TL,
semantically complex SL words, different distinctions in meaning in
the SL and the TL, the TL lacks a superordinate, the TL lacks a speci ic
term (hyponym), interpersonal or physical perspective differences,
differences in expressive meaning, differences in form, differences in
frequency and purpose of using speci ic form and the use of loan words
in the SL.
1.
Cultural speci ic concepts: concepts that are culturally bound.
There are many examples of this, such as the Arabic words ,‫وﺿﻮء‬
‫ ﺟﻨﺎﺑﺔ‬,‫ ﻏﺴﻞ‬,‫ وﻟﻲ اﻟﻌﺮوس‬,‫ ﻃﻬﺎرة‬,‫ أﻋﻨﻜﺎف‬,‫ ﺣﺞ‬,‫ ﺻﯿﺎم‬,‫ﺻﻼة‬. In English, Baker
(1992) gave examples of words that are culture-speci ic, such as
‘Speaker’ of the House of Commons.
2.
SL concepts are not lexicalized in the TL: Some concepts may be
well-known and perfectly well-understood in the TL; however,
they are not lexicalized in it. For example, the word ‘standard’ in
the sense of ‘ordinary’ is perfectly well-understood in Arabic.
However, it does not have an equivalent. Another example is
‘landslide’, which is understood in many languages, but not
lexicalized.

Uploaded by S. M. Safi
3. Semantically complex SL words: Sometimes one morpheme
expresses a set of meanings that may not be expressed by
sentences. For example, the Arabic word ‫ اﻟﺘﻘﻮى‬needs a sentence to
convey its meaning.
4.
Different distinctions in meaning in the SL and the TL: Languages
can vary in the number of distinctions in meaning they contain.
For example, Arabic makes a distinction in meaning between ‫ﺑﺨﯿﻞ‬
and ‫ﺷﺤﯿﺢ‬. The word ‫ ﺑﺨﯿﻞ‬is used to refer to a person who does not
like to spend money on others, which is equivalent to ‘stingy’.
However, the word ‫ ﺷﺤﯿﺢ‬refers to a person who does not like to
spend money on others or on himself. The distinction in meaning
between the two words does not exist in English. Another
example is that Arabic makes a distinction in meaning between
‫ ﺧﺴﻮف‬and ‫ﻛﺴﻮف‬. The word ‫ ﺧﺴﻮف‬is used to refer to a lunar (of the
moon) eclipse, while the word ‫ ﻛﺴﻮف‬is used to refer to a solar (of
the sun’ eclipse. English does not make this distinction by means
of a single word; ‘eclipse’, is used to refer to both lunar and solar
eclipses. Arabic is rich with such examples. Take, for instance,
how, when referring to camels, the Arabic language makes a
distinction in meaning between nouns that are based on age.
Arabic names for a camel that are based on its age are diverse and
many (e.g. ‫ ﻣﺨﻠﻮل‬،‫ ﻟﻜﻲ‬،‫ أﺑﻦ ﻟﺒﻮن‬،‫)اﺑﻦ ﻣﺨﺎض‬. However, all these words
can only be translated into English as a ‘camel’, as English does
not make a distinction in meaning between camels based on age.
5.
The TL lacks a superordinate: one language may have a
superordinate for an item, while another, instead, has many
hyponyms. For example, mounting a camel has two hyponyms in
Arabic that are not represented in English:
‫ﺣﺮذون‬: refers to mounting a camel with a saddle
‫ﺷﺬاد‬: refers to mounting a camel without a saddle.
6. The TL lacks a speci ic term (hyponym): One language may have a
hyponym or hyponyms for an item that does not exist in another
language. For example, English has many hyponyms for ‘house’:
‘bungalow’ ‘cottage’ ‘croft’ ‘chalet’ ‘lodge’ ‘hut’ ‘mansion’ ‘manor’

Uploaded by S. M. Safi
bungalow, cottage , croft , chalet , lodge , hut , mansion, manor,
‘villa’ or ‘hall’. However, Arabic does not have equivalents for these
hyponyms. Similarly, the verb ‘jump’ has many hyponyms: ‘leap’,

‘vault’, ‘spring’, ‘bounce’, ‘dive’, ‘clear’, ‘plunge’ and ‘plummet’. These


hyponyms do not exist in Arabic.
7.
Interpersonal or physical perspective differences: Physical
perspective refers to the relationship between things or people,
which may differ from one language to another. For example,
Arabic makes differences between maternal uncle and paternal
uncle. In Arabic, there are two words that describe these
relationships, ‫ ;ﻋﻢ و ﺧﺎل‬in English, there is only one word, ‘uncle’.
8.
Differences in expressive meaning: Words may share denotative
meaning in two languages; however, they may not share the
expressive meanings. For example, the word ‘homosexuality’ is an
inherently pejorative word in Arabic, whereas it is not so in
English. The same applies to words such as ‘lesbian’ and ‘gay’.
9.
Differences in form: Equivalent forms in an SL and TL are rarely
found. For example, in English, adjectives are derived from verbs
by adding certain suf ixes (e.g. work vs. workable); however, this
is not so in Arabic. Hence, translation from English to Arabic must
change the form to render the meaning, depending on the context.
Similarly, Arabic makes frequent use of pre ixes. A form is
changed in meaning by adding or changing a pre ix. For example,
the verb ‫‘( ﻓﺘﺢ‬opened’) can be changed to a different meaning by
adding the pre ix ‘‫ ’اﺳﺖ‬to become ‫‘ اﺳﺘﻔﺘﺢ‬asked someone to open’.
Also, in Arabic, the meaning can be changed by adding in ixes,
which is not common in English. For example, the verb ‫أﺷﺎر‬
(‘advised’) can be changed into different words and different
meanings by adding in ixes or af ixes. It can be changed to: ‫أﺳﺘﺸﺎر‬
(‘sought advice’), ‫‘( ﻣﺴﺘﺸﺎر‬advisor/consultant/judge’).
10. Differences in frequency and purpose of using speci ic form: This
occurs when one form occurs with greater frequency—say, it
occurs more frequently in the SL than in the TL. For example,
English makes use of ‘-ing’ more than any other language.

Uploaded by S. M. Safi
11.
The use of loan words in the SL: Loan words sometimes are used
in an SL to add an air of sophistication that may not be
transferrable to the TT. For example, the English loan word
‘dilettante’ does not have an equivalent in the Arabic language.
The use of loan words brings to attention the importance of
avoiding mistranslating the ‘false friends’. For example,
‘demander’ in French is not an equivalent of ‘demand’ in English.
Another example is the word ‫ﺟﻬﺎد‬, which can be either ‫ ﺟﻬﺎد دﻓﻊ‬or
‫ﺟﻬﺎد ﻃﻠﺐ‬. However, the English word ‘jihad’ refers partially to only
one type of jihad. Therefore, the English word ‘jihad’ and the
Arabic word ‫ ﺟﻬﺎد‬are not fully equivalent. Another example is the
word ‘harem’, which should not be confused with the word ‫ﺣﺮﯾﻢ‬.
The English word is used to refer to ‘a Muslim man, who has
several wives or sexual partners living in his house’. Of course,
this de inition creates a false image in the minds of non-Muslims,
as Muslims cannot have sexual partners without their being his
wives. Anyhow, the English use of the word ‘harem’ should not be
confused with the Arabic word, which means ‘women’.
Non-equivalence problems at word level sometimes overlap. A
culture-speci ic word is a particularly semantically complex word. Also,
some words can fall under more than one type of non-equivalence.
These problems of non-equivalence discussed require strategies that,
according to Baker, are followed by professional translators to deal
with non-equivalence at the word level. These strategies affect
translation as a inal product, and they are related to the problems of
non-equivalence at the word level.
4. Strategies followed by translators to overcome non-
equivalence
Baker describes eight strategies—though this list may not be
exhaustive—that are used by professional translators for dealing with
various types of non-equivalence at the word level. Strategies of
translation are particularly important as some losses in translation can
be attributed to employing improper strategies. The strategies
mentioned by Baker are: translation using a more general word
(superordinate), translation using a neutral/less expressive word,

Uploaded by S. M. Safi
cultural substitution, translation using a loan word or a loan word
accompanied by an explanation, paraphrasing using a related word,
paraphrasing using unrelated words, omission and illustration.
1.
Translation using a more general word (superordinate): This is
typically used to deal with non-equivalence at word level,
especially when dealing with propositional meaning. For example,
using the English word ‘money’ to render the Quranic word
‫ورق‬/wariq/, which literally means ‘silver coin’; however, this kind
of strategy does not appear to work with all types of text. Texts
such as the Holy Quran are so precise and accurate that the
outcome of using a general word instead of the speci ic word is not
an appropriate strategy.
2.
Translation using a neutral/less expressive word: An example of
this is translating the English word ‘standard’ into Arabic as
‫ﻗﯿﺎﺳﻲ‬/qiaasi/, which is less expressive than the SL word.
3.
Cultural substitution: This strategy depends on how much licence
has been afforded to the translator by the commissioner, and the
purpose of the translation. In this strategy, the SL speci ic item is
replaced by a TL speci ic item that is considered to create the same
effect; for example, translating the English item ‘Congress’ into the
Arabic item ‫اﻟﺸﻌﺐ ﻣﺠﻠﺲ‬/majlisu ashshaAAb/, to create the same
effect for TL readers.
4.
Translation using a loan word or a loan word accompanied by an
explanation: This strategy usually deals with culture-speci ic items,
modern concepts and buzz words. Thus, the translator sometimes
intentionally uses loan words to introduce the SL culture to the TL
culture, rather than merely providing a descriptive translation. For
example, lexical items such as ‫ﻣﺠﺎﻫﺪﯾﻦ‬/‘mujahedeen’/, ‘‫‘ اﻟﻘﺎﻋﺪة‬/al-
QaAAidatu/, and ‫ﻓﺪاﺋﯿﯿﻦ‬/‘Fedayeen’/were transferred into English
without translation.
5. Paraphrasing using a related word: This strategy is mostly adopted
when the SL word is lexicalized in the TL, but in a different form.
For example, the Arabic Islamic word ‫ﯾﺘﻮﺿﺄ‬/yatawdda’a/ is usually
d di ‘d bl i ’ ‘d d ’ ’

Uploaded by S. M. Safi
rendered into ‘do ablution’ or ‘do wudo’ua’
6.
Paraphrasing using unrelated words: This strategy is followed
when the SL word is not lexicalized in the TL. For example, the
Arabic word ‫ﻣﺮاﺑﻂ‬/murabet/ is not lexicalized in English; hence,
paraphrasing can be adopted as strategy to render it. It can be
rendered as ‘guarding the borders of a Muslim state’.
7.
Omission: This strategy is followed when the meaning can be
rendered without the omitted word. Hence, instead of creating
confusion for readers of the TT, omitting a word or phrase can be
an option.
8.
Illustration: This strategy is followed by translators when the ST
does not have a one-to-one equivalent; the SL word requires
considerable elaboration to be rendered, and hence a picture can
render the meaning more accurately. This strategy is employed in
translating advertisements. In fact, the strategies discussed above
overlap, and sometimes a translator may use two strategies
simultaneously. It is also the job of a translator to choose the best
strategy for translating an ST.
5. Grammatical equivalence
Baker de ines grammar as ‘the set of rules which determine the way
in which units such as words and phrases can be combined in a
language and the kind of information which has to be made regularly
explicit in utterances’ (p. 83). Baker adds that grammar is organized
according to two dimensions: morphology and syntax. Languages have
wide variations in the different aspects of grammar. These differences,
which pose the problem of a lack of grammatical equivalence, could be
in number, person, tense, or aspect, among others (Baker, 1992/2001).
For more details and examples, see Chapter 4.
6. Textual equivalence
Baker (1992) follows the model of cohesion in Halliday and Hasan
(1976). Halliday and Hassan identi ied ive cohesive devices in English,
reference, substitution, ellipsis, conjunction and lexical cohesion. Baker
adds that the level of cohesion differs from one language to another, or
even within the same language from one text to another. However,

Uploaded by S. M. Safi
explicit markers of cohesion contribute to raising redundancy in a text;
absence of these markers lowers it.

Exercises
1.
Baker (1992/2011) mentioned that non-equivalence at word
level occurs for many reasons. State these reasons, supporting
your answer with examples from your own work.
2.
Baker (1992/2011) mentioned that non-equivalence as a
translation problem could occur at the grammatical level.
Explain, with examples.
3.
Apart from the strategies mentioned by Baker (1992/2011), can
you think of any other translation strategies that you use to deal
with the problem of non-equivalence at the word level?
4.
Match the following translation strategies with the suggested
translations (there is one extra strategy):
Paraphrasing using a related word;
Paraphrasing using unrelated words;
Translation using a more general word (superordinate);
Translation using a loan word;
Cultural substitution;
Translation using a loan word or a loan;
Omission.

Source and target lexis The translation strategy


1. Knesset
‫ﻣﺠﻠﺲ اﻟﻨﻮاب‬
2. ‫ﯾﻌﺘﻤﺮ‬
do umrah
3. ‫ﻟﻤﻮﻗﻮذة‬
‘An animal that is hit fatally’
4. ‫ﻣﺠﺎﻫﺪﯾﻦ‬
Mujahedeen

Uploaded by S. M. Safi
Source and target lexis The translation strategy
5. Pigeon
‫ﻃﺎﺋﺮ‬
6. White supremacy
‫اﻷﻋﺘﻘﺎد ﺑﺴﯿﺎدة و أﺳﺘﻌﻼء اﻟﻠﻮن اﻷﺑﯿﺾ‬

2.2.12 Koller’s Notion of Equivalence


According to Koller (1979, 1989), equivalence can be ‘denotative’,
depending on an ‘invariance of content’; ‘connotative’, depending on
similarities of register, dialect, and style; ‘text-normative’, based on
‘usage norms’ for particular types of text; and ‘pragmatic’, ensuring
comprehensibility in the receiving culture (Koller, 1979, pp. 186–91;
1989, pp. 99–104). Munday, in his book Introduction to Translation
Studies (2001), mentioned that Koller (1979, pp. 186–191; 1976/1989,
pp. 99–104) differentiates between ive types of equivalence:
denotative, connotative, text-normative, pragmatic and formal.
1.
Denotative equivalence: This is related to equivalence of the
extralinguistic content of a text, or ‘content invariance’.
2.
Connotative equivalence: This is related to the lexical choices,
especially between near-synonyms, or ‘stylistic equivalence’.
3.
Text-normative equivalence: This is related to text types and the
fact that different types of texts behave in different ways; this is
close to the work of Katharina Reiss. According to this type of
equivalence, the SL and TL vocabulary leave the same effect on
their respective readers.
4.
Pragmatic equivalence: Also referred to as ‘communicative
equivalence’, this is oriented towards the receiver of the text or
message. This is similar to Nida’s ‘dynamic equivalence’. According
to this type of equivalence, the SL and the TL words have similar
orthographic or phonological features.
5. Formal equivalence: This is related to the form and aesthetics of
the text, and includes word play and the individual stylistic

Uploaded by S. M. Safi
the text, and includes word play and the individual stylistic
features of the ST. It is elsewhere referred to as ‘expressive
equivalence’ and is not to be confused with Nida’s term.

2.2.13 Popovič (1976) Types of Equivalence


Popovič (1976) in his dictionary distinguishes four types of
equivalence: linguistic, paradigmatic, stylistic (translational) and
textual (syntagmatic).
1.
Linguistic equivalence: This is found when the SL text and the TL
text are homogeneous at the linguistic level; that is, word-for-word
translation; for example, translating short texts such as ‘I live in
Cairo’ into ‫أﻧﺎ أﺳﻜﻦ ﻓﻲ اﻟﻘﺎﻫﺮة‬.
2.
Paradigmatic equivalence: This occurs when there is equivalence
of ‘the elements of a paradigmatic expressive axis’. Popovič
considers elements of grammar as being of a higher category than
lexical equivalence; for example, translating ‘Egypt defeated Israel
in 1973’ as 1973 ‫ﻣﺼﺮ ﻫﺰﻣﺖ أﺳﺮاﺋﯿﻞ ﻓﻲ‬. In this example, the syntactic
and lexical features of the ST were maintained in the TT. However,
it is dif icult to preserve this form in long texts due to the syntactic
disparities between English and Arabic.
3.
Stylistic (translational) equivalence: This occurs ‘when there is
functional equivalence of elements in both of the SL and TL aiming
at an expressive identity with an invariant of identical meaning’
(Popovič, 1976, p. 33). In other words, the ST meaning is conveyed
to the TT, maintaining the expressive meaning. For example,
translating Trump’s expression of ‘Iran’s downing of the American
drones is new wrinkles, a ly in the ointment’ as ‫أﺳﻘﺎط اﯾﺮان ﻟﻄﺎﺋﺮﺗﯿﻦ‬
‫أﻣﺮﯾﻜﺘﯿﻦ ﺑﺪون ﻃﯿﺎر ﻫﻮ زوﺑﻌﺔ ﻓﻲ ﻓﻨﺠﺎن‬. In this example, the functional
equivalence of the ST idiom was maintained, without preserving
the lexical items of it in the TT.
4. Textual (syntagmatic) equivalence: This occurs when ‘there is
equivalence of the syntagmatic structuring of a text, i.e.
equivalence of form and shape’ (Popovič, 1976, p. 33). This is quite
dif icult to achieve between English and Arabic due to the many
differences between the two languages.

Uploaded by S. M. Safi
Exercises
1.
Translate the following texts into Arabic, explaining which of
Popovič’s types of equivalence were applied in the translation.
His presidential election was just a two-horse race.
Election fever has started, there’s politics on every channel.
His scandal will be a political hot potato.
2.
Examine the ST and TT below, and explain which of Popovič’s
types of equivalence was applied (Source Reverso online).
ST: We have new wrinkles in the laws of war and accountability.
What do we do with things like unmanned slaughter?
TT: ‫ﻟﺪﯾﻨﺎ ﺗﺼﺎدﻣﺎت ﺟﺪﯾﺪة ﻓﻲ ﻗﻮاﻧﯿﻦ اﻟﺤﺮب و ﻣﺴﺆوﻟﯿﺎﺗﻬﺎﻣﺎ ﻓﺎﺋﺪة أﺷﯿﺎء ﻣﺜﻞ ﻣﺬاﺑﺢ ﺑﺪون‬
‫ﺗﺪﺧﻞ اﻹﻧﺴﺎن؟‬

2.2.14 The Cognitive Approach to Translation


Bell (1991) proposed a cognitive approach to translation, which is
based on Halliday’s SFL theory. In an attempt to understand how a
translator understands the meaning of a text, Bell posits that a text is a
product of three types of choice and, therefore, three types of meaning.
According to Bell, there are three types of meaning: cognitive,
interactional and discoursal. These three types of meaning are
organized by three metafunctions—ideational, interpersonal and
textual—and they are realized by three language systems—logical,
grammatical and rhetorical. To clarify, the cognitive type of meaning is
organized by the ideational function and realized by the logical system
of language; the interactional type of meaning is organized by the
interpersonal metafunction and realized by the grammatical language
system; and the discoursal type of meaning is organized by the textual
metafunction and realized by the rhetorical language system.
Cognitive meaning, which is what the text about, is expressed by
the ideational metafunction, which is represented through the

Uploaded by S. M. Safi
grammatical system of transitivity. Ideational, as discussed by Halliday,
refers to the ield (what the text about).
According to Bell, translation is a process of analysing and then
synthesizing. Both the analysing and synthesizing include three major
‘stages’: syntactic, semantic and pragmatic. The irst process—the ST
analysis—results in ‘language-free semantic representation’ (Bell,
1991, pp. 56–57), which will be then used as the starting point for its
conversion into the TT. Analysis is carried out through the functional
and pragmatic categories of clause structure, propositional content,
thematic structure, register features, illocutionary force and speech
acts. Synthesis, on the other hand, encompasses purpose, thematic
structure, style and illocutionary force before obtaining the syntactic
synthesis (Bell, 1991, pp. 58–60, cited in Manfredi, 2014, p. 17).

Exercise
One of the techniques used to examine the cognitive effect in
translation is ‘verbalization’. Based on this statement, translate the
following text using a ‘speak aloud’ mode of translation. Then,
explain the extent to which this technique was useful in enhancing
the translation process.
The murder of Jamal Khashoggi was ‘gruesome’ but a United
Nations report into the journalist’s killing is ‘ lawed,’ Saudi Arabia’s
Minister of State for Foreign Affairs Adel al-Jubeir said Thursday.
In an interview with CNN’s Christiane Amanpour al-Jubeir
denied that Riyadh should accept responsibility for Khashoggi’s
murder, and said he disagreed with the indings of a UN rapporteur
who laid the blame on the government.
‘This is a gruesome murder that took place without
authorization, for which the people who perpetrated (it) are being
punished now,’ al-Jubeir said.
In a much-anticipated report published Wednesday, UN
investigator Agnes Callamard said that there was ‘suf icient credible
evidence’ that Saudi Arabia’s Crown Prince Mohammed bin Salman
bears responsibility for Khashoggi’s killing, and that he should be
investigated for it.
Callamard said that Saudi Arabia was responsible under
international law for Khashoggi’s ‘deliberate, premeditated

Uploaded by S. M. Safi
execution,’ and that current sanctions on some senior of icials do
not go far enough (CNN, last accessed on 27 June 2019).

2.2.15 Functionalist Approach in Translation (Non-


equivalence Approach)
1. Katherina Reiss
The functionalist approach started in Germany in 1970s and 1980s.
Katherina Reiss looked at a text as the operating level of
communication. She borrowed Buhlerl’s of the classi ication of
language functions. Reiss relates language functions to their
corresponding language ‘dimensions’ and to the text types or
communicative situations in which they are used. The three types of
texts are:
A. Informative texts
Informative texts are texts that transfer information, knowledge
and opinions logically and referentially. The main focus of
communication is topic. Examples of such types of text are news and
scienti ic articles. The translation of these types of text should retain
the full message of the ST without redundancy. Explicitation may be
used if needed. The translation should be in terms of ‘plain prose’.
B. Expressive texts
This type of text uses aesthetic functions, such as is found in
literary works. The translation of this type of text should maintain the
aesthetic and artistic form of the ST. A translator needs to convey the
view of the ST’s author, adopting the identifying translation strategy.
C. Appellative or operative texts
One example of this type of text is found in advertisements. The
translation of such texts should be ‘adaptive’, in the sense that it should
create the same effect as that of the ST on it readers.
D. Audiomedial texts
These are texts that require non-printed media, such as movies and
songs. In this type of text supplementary methods are needed, such as
words to translate pictures, or vice versa.
According to Reiss (1971), the quality of a TT is assessed through
intralinguistic and extralinguistic criteria. Intralinguistic criteria
include semantic, lexical, grammatical and stylistic features;

Uploaded by S. M. Safi
extralinguistic criteria include situation, subject ield, time, place,
receiver, sender and affective implications. These intralinguistic and
extralinguistic criteria vary in terms of value depending on the text to
be translated. For example, in a news text, the semantic value is of
greater worth than any other value. This applies to all texts where the
content is of great importance. Although Reiss postulates that an ST
function should be translated to a similar TT function, she states that in
some cases the function of the TT may be different from the function of
the ST. She gives an example of the book Gulliver Travels, whose
function was operative because it was a satire. However, when
translated the translation takes on the form of an ordinary function,
and therefore, the function is expressive.
Relationship between an ST and a TT ( Nord, 2005)
After her discussion of idelity and equivalence between the ST and
the TT, Nord (2005) claims that the only valid way to offer an
acceptable notion of equivalence is to analyse the ST. Producing a
functionally equivalent TT, which is based on analysis of the ST, is one
of the purposes of translation. However, Nord considers that functional
equivalence is the exception, and not the normal skopos of translation.
Equivalence in the functional view of translation is not the be all
and end all but, rather, is subordinate to the translation (TT) skopoi
(functions). The skopos of translation is, thus, determined by the
function it is intended to ill in the TL. Fidelity is subordinate to the
skopos rule, which may change according to the needs of the TT. Two
concepts need to be clari ied here: the concept of idelity (intertextual
coherence) and the concept of adequacy. The intertextual coherence or
idelity in the functional perspective of translation is prioritized if it
achieves the functions intended in the TT; however, if it has not been
possible to achieve such idelity, adequacy and acceptability will be the
required standard. Nord accentuates that a TT that is not based on or
bound to a given ST cannot be considered a translation; rather, it
should be considered as cross-cultural consulting or a cross-cultural
technical writing. Hence, the relationship between an ST and a TT
cannot be ignored. Also, if possible, there should be compatibility
between the intention of the ST and the functions of the TT. This
compatibility implicates loyalty to the ST author. Therefore, the job of a

Uploaded by S. M. Safi
translator is double-bound, as he needs to be loyal to the sender of the
ST and, at the same time, should observe the needs of the TT receiver.
Translating as a form of translational interaction
Communication among humans occurs through intentional actions
through which they communicate interpersonally to convey their
messages; this justi ies the importance of action theory in explaining
translational communication (Nord, 2008). Nord posits that the
communication process occurs between a sender and a receiver (or
addressee) and that it is limited in time and space. In other words, the
translation process is conditioned by historical and cultural
dimensions. This implies that translation does not need to be literal to
be accurate. For example, a translator may diverge from the literal
meaning of the ST expressions to more functionally equivalent TT
expressions to convey the intended meaning accurately.
Nord differentiates between translation and translational action.
The former refers to what translators do when rendering a text; the
latter refers to what goes beyond translation. In Nord’s words:
‘Translating in the narrower sense always involves the use of some
kind of source text, whereas translational action may involve giving
advice and perhaps even warning against communicating in the
intended way’ (Nord, 2008, p. 17). Nord adds that translational action
is intentional and is voluntarily undertaken by the initiator of that
action. This intentionality may be different from that of the ST
originator. In this regard, the initiator or the person commissioning the
translation plays a role in the translational process and action. The
initiator is the person (possibly the client) who starts the translation
process and asks a translator to translate text based on a speci ic need
and a speci ic translation brief. The translator also has an important
role in the translational process, as they are the expert in the
translational action who acts based on the request of the initiator. The
translator’s role includes the evaluation of the translation brief
economically, ideologically and legally. They are also the person who
checks whether the translation is really necessary and may advise the
initiator not to translate the ST because it does not serve the intended
purpose. The translator may also decide to shorten the ST in the TL
(Vermeer, 1986).

Uploaded by S. M. Safi
Nord (2008) also states that the producer of an ST has a role in the
translation process, as he is the person who produced the text to be
translated. In this regard, Nord makes a distinction between a text
producer and a text sender. The text sender is the person or the
institution that uses the text to convey a speci ic message; the text
producer is the person who creates the ST using his stylistic and
linguistic skills. However, sometimes both the sender and text
producer are one and the same.
Nord (2008) also mentions that the TT receiver has a role, as they
are a part of the translation brief. Nord discriminates between the
receiver and addressee. The receiver is the person or the institution
that reads the translated text; the addressee is the expected receiver
from the standpoint of the text producer (i.e. the translator). Another
important factor that should be considered in the translation process
is the TT user, who may use the translation for training, teaching or fun.
Nord’s model seems to be in line with the model proposed by Holz-
Mänttäri (1984). Table 2.4 summarizes the main concepts related to
the translation action approach.
Table 2.4 Terms used in the translational action approach

Initiator The company or individual who needs the translation


Commissioner The individual who contacts the translator
ST producer Not necessarily always involved in the TT production
TT producer The translator
TT user Material or sales literature
TT receiver The inal recipient of the TT; for example, the students in a TT user’s class, or
clients reading the translated sales literature

Source Holz-Mänttäri (1984, as cited in Munday, 2001)

Documentary vs. instrumental translation (Nord, 1988/1991)


Nord (1991) differentiates between two types of translation:
documentary and instrumental. Documentary translation ‘serves as a
document of a source culture communication between the author and
the ST recipient’ (Nord, 1991, p. 72). It is ST oriented, in the sense that
the reader of the TT knows that what they read is a translation.
Strategies employed to apply this documentary approach include

Uploaded by S. M. Safi
word-for-word translation, literal translation and exoticized
translation.
On the other hand, instrumental translation ‘serves as an
independent message transmitting instrument in a new
communicative action in the target culture, and is intended to ful ill its
communicative purpose without the recipient being conscious of
reading or hearing a text which, in a different form, was used before in
a different communicative situation’ (Nord, 1991, p. 73). This type of
translation is TT oriented and, rather than sounding like a translation,
the TT sounds as though it is an original text.. The function of the ST
and the TT may, however, be the same.
Nord’s Text Analysis Model
Nord (1997), in her Translating as a Purposeful Activity, proposed a
model for text analysis for translation. She highlights certain important
concepts, such as the translation brief (or commission), the
importance of analysing the ST, and the functional hierarchy of
translation problems.
Translation brief
Nord postulates that a translator needs to compare the ST and the
TT pro iles based on the translation brief, so as to identify any
divergences between the two pro iles. The translation brief should
include the text functions, the addressees (sender and recipient), the
time and place of text delivery, the medium (speech and writing), and
the motive (why the ST was written and why it is being translated).
The role of the ST analysis
After the identi ication of the ST and TT pro iles, a translator needs
to analyse the ST in terms of the intertextual factors:
subject matter;
content: including connotation and cohesion;
presuppositions: real-world factors of the communicative situation
presumed to be known to the participants;
composition: including microstructure and macrostructure;
non-verbal elements: illustrations, italics, etc.;
lexic: including dialect, register and speci ic terminology;
sentence structure;
Suprasegmental features: including stress, rhythm and stylistic
punctuation (Nord, 1997, pp. 79–129).

Uploaded by S. M. Safi
The functional hierarchy of translation problems
Nord provides a functional hierarchy that a translator needs to
follow during a translation task:
1.
The translator needs to identify the intended function of the
translation, which can be either documentary or instrumental.
2.
The translator decides what elements need to be maintained in the
TT and what elements need to be adapted, based on the translation
brief provided by the commissioner.
3.
Based on the translation type, the translator decides whether the
translation is source culture oriented or target culture oriented.
4.
The translator handles the problems of the text at a lower linguistic
level.

Exercise
1. Following Nord’s model of analysis, analyse the following texts
suggesting the appropriate translation strategies. Then, translate
the texts.
1. ‫أﻋﻠﻨﺖ وزارة اﻟﺘﺮﺑﯿﺔ واﻟﺘﻌﻠﯿﻢ واﻟﺘﻌﻠﯿﻢ اﻟﻔﻨﻲ ﻋﻦ ﻓﺘﺢ ﺑﺎب اﻟﺘﻘﺪم ﻟﻘﺒﻮل دﻓﻌﺔ ﺟﺪﯾﺪة ﻣﻦ‬
‫اﻟﻄﻼب اﻟﺬﯾﻦ ﺣﺼﻠﻮا ﻋﻠﻰ اﻟﺸﻬﺎدة اﻹﻋﺪادﯾﺔ ﻟﻼﻟﺘﺤﺎق ﺑﺎﻟﻤﺪرﺳﺔ اﻟﻔﻨﯿﺔ اﻟﻤﺘﻘﺪﻣﺔ ﻟﺘﻜﻨﻮﻟﻮﺟﯿﺎ‬
‫ ﻣﻦ ﺟﻤﯿﻊ‬2019/2020 ‫اﻟﻄﺎﻗﺔ اﻟﻨﻮوﯾﺔ ‘ﺑﻤﺪﯾﻨﺔ اﻟﻀﺒﻌﺔ – ﻣﺤﺎﻓﻈﺔ ﻣﻄﺮوح’ ﻟﻠﻌﺎم اﻟﺪراﺳﻰ‬
‫ وذﻟﻚ ﻋﺒﺮ اﻟﻤﻮﻗﻊ‬،2019 ‫ ﯾﻮﻟﯿﻮ‬15 ‫ إﻟﻰ‬1 ‫ ﻓﻲ اﻟﻔﺘﺮة ﻣﻦ‬،‫ﻣﺤﺎﻓﻈﺎت ﺟﻤﻬﻮرﯾﺔ ﻣﺼﺮ اﻟﻌﺮﺑﯿﺔ‬
‫( اﻟﺮﺳﻤﻲ ﻟﻮزارة اﻟﺘﺮﺑﯿﺔ واﻟﺘﻌﻠﯿﻢ واﻟﺘﻌﻠﯿﻢ اﻟﻔﻨﻲ‬Youm 7).
2. President Donald Trump approved retaliatory military strikes
against Iran on Thursday before changing his mind, US media
report.
The New York Times, citing senior White House of icials, says
strikes were planned against a ‘handful’ of targets.
They say the operation was allegedly under way ‘in its early
stages’ when Mr Trump stood the US military down. The White
House has so far made no comment.
This comes after Iran shot down a US spy drone.
Tehran says the unmanned US aircraft entered Iranian airspace
early on Thursday morning. The US maintains it was shot down in

Uploaded by S. M. Safi
international airspace.
Tensions have been escalating between the two countries, with
the US recently blaming Iran for attacks on oil tankers operating in
the region. Iran has announced it will soon exceed international
agreed limits on its nuclear programme.
Last year, the US unilaterally pulled out of a 2015 nuclear deal
aimed at curbing Iran’s nuclear activities.
What do US media say?
The New York Times irst published details of the apparent
planned strikes late on Thursday night in Washington.
As late as 19:00 local time (23:00 GMT), it said, US military and
diplomatic of icials still expected the strikes on agreed targets,
including Iranian radar and missile batteries, to take place.

Exercise
Examine the ST and TT below, and then analyse the texts explaining
whether the translation is documentary or instrumental. Justify the
approach selected by the translator.

ST TT
‫ﻣﺜﻮل أﻣﯿﺮ ﺳﻌﻮدي أﻣﺎم إﺣﺪى ﻣﺤﺎﻛﻢ اﻟﻘﺎﻫﺮة‬ Cairo puts Saudi Prince on Trial
‫ﺻﺮﺣﺖ وﻛﺎﻟﺔ أﻧﺒﺎء اﻟﺸﺮق اﻷوﺳﻂ أﻧﻪ ﻣﻦ اﻟﻤﻘﺮر ﻣﺜﻮل‬ A SAUDI prince is to stand trial on March 12
‫اﻷﻣﯿﺮ اﻟﺴﻌﻮدي أﺣﻤﺪ ﺑﻦ ﺗﺮﻛﻲ ﻟﻠﻤﺤﺎﻛﻤﺔ ﯾﻮم اﻟﺜﺎﻧﻲ ﻋﺸﺮ‬ because his dangerous dogs mauled a ive-year-
‫ﻣﻦ ﻣﺎرس اﻟﺤﺎﻟﻲ وذﻟﻚ ﻋﻘﺐ ﻗﯿﺎم ﻛﻼﺑﻪ اﻟﺸﺮﺳﺔ ﺑﻤﻬﺎﺟﻤﺔ‬ old Egyptian girl while she was playing in the
garden of a Cairo hotel. The prosecutor holds
‫وﺗﺸﻮﯾﻪ وﺟﻪ ﻃﻔﻠﺔ ﻣﺼﺮﯾﺔ ﺗﺒﻠﻎ ﻣﻦ اﻟﻌﻤﺮ ﺧﻤﺲ ﺳﻨﻮات‬
Prince Ahmed bin Turki Al-Saud responsible for
‫ وﻗﺪ‬.‫ﺑﯿﻨﻤﺎ ﻛﺎﻧﺖ ﺗﻠﻌﺐ ﺑﺤﺪﯾﻘﺔ أﺣﺪ اﻟﻔﻨﺎدق اﻟﻜﺒﺮى ﺑﺎﻟﻘﺎﻫﺮة‬ causing the girl grievous bodily harm. The victim
‫ﺗﻘﺮر أن ﯾﻤﺜﻞ اﻷﻣﯿﺮ أﻣﺎم اﻟﻤﺪﻋﻲ اﻟﻌﺎم ﺣﯿﺚ ﯾﻌﺪ ﻣﺴﺌﻮﻻ‬ has undergone a series of operations done for
‫ﻋﻤﺎ ﻟﺤﻖ ﺑﺎﻟﻄﻔﻠﺔ ﻣﻦ ﺿﺮر ﺑﺎﻟﻎ ﺧﻀﻌﺖ ﻋﻠﻰ إﺛﺮه‬ facial injuries, the Middle East New Agency
‫ﻟﺴﻠﺴﻠﺔ ﻣﻦ اﻟﻌﻤﻠﯿﺎت اﻟﺠﺮاﺣﯿﺔ ﻟﻤﻌﺎﻟﺠﺔ إﺻﺎﺑﺎت وﺟﻬﻬﺎ‬. (MENA) said
(Source Translators Avenue)

The skopos theory


There is another approach to translation that is the polar opposite
to those mentioned hitherto; this approach focuses on the notion of
‘purposes’. Put simply, translation is designated to achieve a purpose.
Basically, this is the approach adopted by functionalists such as

Uploaded by S. M. Safi
Vermeer (1989). This approach adopts the notion of non-equivalence.
With a few exceptions (e.g. Reiss, 1971), most functionalists (e.g.
Vermeer, 1989) do not believe that equivalence between ST and TT is
achievable. According to this approach, the purpose (or
communicative skopos), in relation to the TT is the dominant factor in
the translation process and, hence, one ST can be translated into
different TTs to achieve different functions. Skopos theory centres on
the notion that translation is guided by its skopos (or purpose). In
skopos theory, the end justi ies means. There are three types of
purpose in translation: the purpose of the translator, which may be to
earn some money; the communicative purpose, which is the aim of the
TT and the target situation; and a particular translation strategy or
procedure. Vermeer (as cited in Nord, 2008) uses four concepts in
addition to the term ‘skopos’: aim, purpose, intention and function. Aim
refers to what an agent intends to achieve as a inal result; purpose
refers to the provisional stage to achieve the aim; function refers to
what the text is intended to mean from the receiver’s perspective; and
intention refers to an aim-oriented plan of action. Nord differentiates
between intention and function as follows:

Intention is de ined from the viewpoint of the sender, who


wants to achieve a certain purpose with the text. Yet the best of
intentions do not guarantee a perfect result, particularly in
cases in which the situations of the sender and the receiver
differ considerably. In accordance with the model of text-bound
interaction, the receivers use the text with a certain function,
depending on their own expectations, needs, previous
knowledge, and situational conditions. In an ideal situation, the
sender’s intention will ind its aim, in which case intention and
function would be analogous or even identical. (Nord, 2008, pp.
27–28)

Text in skopos theory is just an offer of information, whereby a


reader selects what they consider relevant. To clarify, an ST is an offer
of information and, similarly, a TT is offer of information made by a
translator. Hence, there is no point in talking about conveyance of the
meaning of the ST. In other words, the translation process is guided by

Uploaded by S. M. Safi
the translation brief, whereby a translator selects some parts of the
information offered in the ST to introduce them in the TT. The TT
readers then select what is relevant to them in speci ic situations. The
TT produced should be meaningful and communicative to the TL
readers, which is intratextual coherence. Another important type of
coherence is the intertextual coherence between the ST information
and the TT information. This intertextual coherence depends on the
translator’s interpretation of the ST and the skopos of the translation.
This, however, does not exclude cases where the TT is faithful to the ST,
which happens in the translation of certain literary texts. It may also
happen in the translation of the Holy Quran. Vermeer puts it as follows:

It might be said that the postulate of ‘ idelity’ to the source text


requires that e.g. a news item should be translated ‘as it was in
the original’. But this too is a goal in itself. Indeed, it is by
de inition probably the goal that most literary translators
traditionally set themselves. (Vermeer, 1989, p. 197)

According to Pym, ‘skopos’, which means ‘goal’, is the key to the


functionalist approach. In this kind of translation, the translator is
more concerned with the TT—in other words, how to create a
communicative translation of an ST, regardless of the lexis. Pym sees
that, according to skopos theory, a translator should work hard to
convey the intellectual and emotional intent of the ST. Reiss and
Vermeer (1984) aimed to establish a general translation theory for all
texts. The basic underlying ‘rules’ of the theory of Reiss and Vermeer
(1984), as cited in Munday (2008), are:
1.
A trunslatum (or TT) is determined by its skopos.
2.
A TT is an offer of information in a target culture and TL
concerning an offer of information in a source culture and SL.
3.
A TT does not initiate an offer of information in a clearly reversible
way.
4.
A TT must be internally coherent.

Uploaded by S. M. Safi
5. A TT must be coherent with the ST.

These ive rules stand in hierarchical order, the skopos rule being
predominant. Thus, translation is viewed as non-directional. In other
words, reversibility is not a prerequisite for good translation. Vermeer
and Reiss also underscore the importance of coherence and idelity for
a successful translation. The coherence rule means that the TT must be
interpretable as coherent with the TT receiver’s situation (Reiss &
Vermeer, 1984, as cited in Munday, 2008). The idelity rule merely
states that there must be coherence with the trunslatum. According to
the hierarchical order of the rules, intertextual coherence is of lesser
importance than intratextual coherence, which, in turn, is subordinate
to the skopos (rule 1) (Munday, 2008, p. 80). Thus, based on skopos
theory, the same text can be translated in different ways according to
the purpose of the TT and the commission given to the translator.
Therefore, if a text is ambiguous, according to skopos theory it can be
translated literally and then explained in a footnote (Munday, 2008).
However, this theory is criticized as it supports the position that any
translation can be justi ied if a translator has declared his intention at
the beginning of his translation process. Skopos theory is also is
criticized for locating coherence as the least important rule (Hodges,
2009).
Reiss argues that ‘text’ should be considered as level of equivalence,
rather than the word or the sentence. Reiss links the three functions to
their corresponding language ‘dimensions’ and to the text, types or
communicative situations in which they are used (Munday, 2008). The
main characteristics of each text type, according to Reiss, are: plain
communication of facts, creative composition, the inducing of
behavioural responses and audiomedial texts.
1.
Plain communication of facts: These facts may include information,
knowledge, opinions and so on; and the language dimension used
to transmit the information whether it is logical or referential. The
text type in this type of information is informative.
2. Creative composition: In this kind of text, the aesthetic dimension

Uploaded by S. M. Safi
of language is used. Thus, the text type is expressive.
3.
Inducing behavioural responses: This includes functions of appeal;
that is, to appeal to or persuade the reader or ‘receiver’ of the text
to act in a certain way. Reiss calls this text type ‘operative’.
4.
Audiomedial texts: This includes ilms, and visual and spoken
advertisements, that supplement the other three functions with
visual images, music and so forth. This is Reiss’s fourth main
characteristic of text type (Reiss as cited in Munday, 2008).
Pym believes that the notion of equivalence is a ‘social illusion’,
which people believe in even though it does not have linguistic
certainty; however, he states that we have to deal with such
‘equivalence beliefs’. Pym makes a distinction between two types of
equivalence: natural equivalence and directional equivalence. Natural
equivalence is basically based on the paradigm of equal value. In other
words, what is said in one language can be translated into another
language, with the same function or worth. As a result, the relation
between an ST and a TT is one of equal value at the level of form,
function, or anything in-between. For example, the English ‘Friday the
13th’ is a natural equivalent for the Spanish ‘Tuesday the 13th’ because
the two terms function in the same way, as each of these days refers to
bad luck in their respective cultures. Another example can be adopted
from Shakespeare: ‘she is as beautiful as a summer’s day’. This
expression can be translated functionally into Arabic as ‫إﻧﻬﺎ ﺟﻤﯿﻠﺔ ﻛﺎﻟﻘﻤﺮ‬
(she is as beautiful as a moon), as it does not sound natural to liken a
beautiful woman to any season in the Arabic culture. In addition,
‘summer’ is not a favourable season in the Arabic culture, as it is linked
to the scorching sun and extreme heat. Pym supports his paradigm by
quoting Nida and Taber’s natural equivalent in their theorizing of Bible
translation. Nida and Taber (1982, p. 12) state that ‘Translating
consists in reproducing in the receptor language the closest natural
equivalent of the source language message’. However, this claim,
according to Wendland (2012), is not accurate because Nida and Taber
prioritize meaning over style, which is not the case in Pym’s theory.
This approach is likely to employ many strategies of translation.
However, this approach was critiqued for its ‘presupposition’ of a ‘non-

Uploaded by S. M. Safi
existent symmetry’ between languages, its lack of a veri iable
‘psychological basis’, an underlying ‘imperialistic’ agenda, and the
promotion of ‘parochialism’ that prefers meaning over form
(Wendland, 2012, p. 95).
Directional equivalence is an asymmetric relation in which, if an
equivalent was created in one direction, it does not necessarily that
exist in the other. Contrary to natural equivalence, which assumes the
existence of equivalence between items bi-directionally, this approach
assumes that equivalence can exist mono-directionally. This approach
is likely to adopt two-opposed poles, such as literal translation vs. free
translation (Wendland, 2012, p. 95). Wendland observes that there is
no borderline between the two types of translation proposed by Pym,
since natural equivalence includes directional equivalence. In addition,
there is no full equivalence between any two languages unless they are
culturally close. It seems that Pym’s notion of equivalence is not clear.

Exercises
1.
Explain, with examples, the differences between Pym’s ‘natural
equivalence’ and ‘directional equivalence’.
2.
Translate the following sentences, indicating whether the
equivalence achieved in the translation is ‘natural equivalence’
or ‘directional equivalence’.
A.
The Democrats showed several clips of Trump’s public
comments about the Ukraine scandal.
B.
Governments should combat terrorism everywhere.
C.
Egypt is a Mecca for learners from all over the world.
D.
Australia has been ravaged by the worst wild ires seen in
decades, with large swaths of the country devastated since
the ire season began in late July.

Uploaded by S. M. Safi
2.2.16 Darwish’s Notion of Equivalence (2010)
In relation to the notion of equivalence, Darwish (2010) argues that
equivalence can be considered in terms of relative equivalence, rather
than absolute equivalence. He, further, proposes the terms
‘approximation’ and ‘alignment’ for the practical consideration of
equivalence. He argues that translation should aim at achieving
optimal approximation, rather than absolute equivalence. A translator,
then, should work on removing the language constraints to achieve
such approximation. Those constraints can be seen in terms of
transparency, opacity and translatability. Transparency is how far an ST
is lexically and structurally close to a TT at the macro and micro levels,
and, thus, that there is likely to be convergence. By contrast, if an ST is
lexically and structurally far from a TT at the macro and micro levels,
divergence occurs. Thus, transparency and opacity reveal how far a
text is translatable. Translatability, as de ined by Darwish, is the extent
to which a translation is possible without loss of meaning, and the
relative ease of such translation. Darwish adds that there are three
problems of translatability with regard to an ST: comprehensibility,
digestibility and cultural dependency. Thus, if a text is not
comprehensible to a translator, they will not be able to render its
meaning. Similarly, if a text has long clauses and sentences, and is
packed with a great deal of information, it will be dif icult for a
translator. Thus, digestibility relates to complexity. Cultural
dependency relates to the culturally bound terms.
As for untranslatability, it is de ined by Darwish as the inability to
render an ST into a TT due to constraints, at the syntactico-symantic,
pragmatic, or rhetorical levels. Bassnett (2005) mentions that Catford
identi ied two types of untranslatability: linguistic and cultural.
Linguistic untranslatability occurs when there is no lexical or
syntactical substitute in the TL for an SL item. However, cultural
untranslatability occurs when a relevant situational feature for the SL
text does not exist in the TL.

2.2.17 The Polysystem Theory


This theory was irst proposed by Even-Zohar in the 1970s; the English
version of the theory was published in his book entitled Papers in
Historical Poetics in 1978. It started as a literature theory, and later

Uploaded by S. M. Safi
developed into a translation theory. Even-Zohar (1979, 1997)
considered translation as a part of the polysystem of literature, and it
can occupy a primary position or peripheral position based on
different factors. Translated literature can occupy a primary position
when literature is young, or weak, or when literature is facing a crisis
(Venuti, 2000). In other words, translations that occupy a central
position in the literary polysystem will not follow the norms of the TL.
In contrast, those translations that occupy a peripheral position in the
literary polysystem will follow the TL norms. It views translation from
the TL literature perspective. It was developed basically for the
purpose of proposing a theory for translating Hebrew literature. Even-
Zohar (2000) postulated that all literary and non-literary works are
interrelated in a polysystem. One weakness of this theory is that it
ignored social factors and their in luence on the forming of literature.
The theory was then developed by Gideon Toury, who presented it as
the theory of norms in translation. Toury (1980) explored the reasons
behind choosing speci ic texts to be translated into Hebrew. He found
that the reasons are far from literary ones, as texts are mostly selected
based on personal reasons, a translator’s preferences, and the purpose
of translation. Toury’s approach to translation was TT oriented. Toury
argued that translation holds a middle position between the SL and the
TL; it can neither completely transfer the ST cultural norms, nor can it
be assimilated into the target culture. Toury rejected the notion of
complete equivalence and, at the same time, rejected the idea of
naturalness in the TL. As both are practically unachievable, he believes
that ‘equivalence’ cannot be disregarded because a translation is
regarded as a representative entity of the ST. However, he focused on
what he termed ‘factual replacement’. Toury called for consideration of
the historical facts of the target culture, which he called ‘translation
norms’; the term ‘norms’ is thus used by Toury to refer to a translator’s
preferences and the factors that in luence them. These factors are
mostly external ones, such as socio-cultural factors. Toury
differentiated between three types of norms: preliminary, initial and
operational. Preliminary norms are those that affect a translator’s
adoption of a speci ic strategy or their translation policy: which texts
to choose for translation. Preliminary norms are not a part of a
translator’s preferences. Initial norms refer to those that re lect a

Uploaded by S. M. Safi
translator’s preference for a speci ic translation approach or strategy
(e.g. being faithful to the ST, or adopting a TT oriented approach).
Operational norms are the norms that govern the actual act of
translation.

Exercises
1.
What are the drawbacks of Polysystem Theory?
2.
Do you agree with Toury’s concepts of norms in translation?
Why?

References
As-Sa i, A. B. (2011). Translation theories: Strategies and basic theoretical issues. Amman: Dar
Amwaj.

Baker, M. (1992/2011). In other words: A coursebook on translation (2nd ed.). London and New
York: Routledge.

Baker, M. (2004). The status of equivalence in translation studies: An appraisal. In Z. Yang (Ed.),
English-Chinese comparative study and translation (p. 1). Shanghai: Foreign Languages Education
Press.

Bassnett, S. (2005). Translation studies. Vasa (3rd ed.). New York: Routledge (Taylor & Francis e-
Library).

Beekman, J., & Callow, J. (1974). Translating the word of god: With scriptures and topical indexes.
Grand Rapids, MI: Zondervan Publishing House.

Bell, R. T. (1991). Translation and translating: Theory and practice. London and New York:
Longman.

Catford, J. C. (1965). A linguistic theory of translation: Language and language learning (1st ed.).
Oxford: Oxford University Press.

Chomsky, N. (1977). Recent contributions to the theory of innate ideas: Summary of oral
presentation. In J. Searle (Ed.), Philosophy of language. Oxford: Oxford University Press.

Darwish, A. (2010). Elements of translation. Melbourne: Writescope.

Even-Zohar, I. (1979). Polysystem theory. Poetic Today, 1(1978), 1–2.

Even-Zohar, I. (1997). Itamar Even-Zohar: Polysystem studies 1990. International Journal for

Uploaded by S. M. Safi
Theory and Analysis of Literature and Communication, 11(1), 88.

Even-Zohar, I. (2000). The position of translated literature within the literary polysystem. In L.
Venuti (Ed.), The translation studies reader (pp. 192–197). London: Routledge.

Frawley, W. (1984). Prolegomenon to a theory of translation. In W. Frawley (Ed.), Translation:


Literary, linguistic, and philosophical perspectives. London and Toronto: Associated University
Presses.

Gentzler, E. (2001). Contemporary translation theories. London and New York: Routledge.

Halliday, M. A. K. (1971). Linguistic function and literary style: An inquiry into the language of
William Golding’s The inheritors. In S. B. Chatman (Ed.), Literary style: A symposium (pp. 330–
365). London and New York: Oxford University Press.

Halliday, M. A. K. (2001). Towards a theory of good translation. In E. Steiner & C. Yallop (Eds.),
Exploring translation and multilingual text production: Beyond content (pp. 13–18). Berlin and
New York: Mouton de Gruyter.

Halliday, M. A. K. (2007). Language and education. London: Continuum.

Halliday, M. A. K., & Hasan, R. (1976). Cohesion in English. London: Longman Publishing House.

Harvey, M. (2000). A beginner’s course in legal translation: The case of culture-bound terms.
ASTTI/ETI, 2(24), 357–369.

Hodges, P. (2009). Compare and contrast two theoretical approaches to translation in


Zainurrahman. The theories of translation from history to procedures. Language and Education.
Retrieved from https://scholar.google.com/scholar?hl=enandq=De+Waard+NidaandbtnG=
Searchandas_ylo=andas_vis=0#8.

Holz-Mänttäri, J. (1984). Translatorisches Handeln. Theorie und Methode. Helsinki: Academiae


Scientarum Fennicae.

House, J. (1997). Translation quality assessment: A model revisited. Tübingen: Gunter Narr.

House, J. (2001, January). Translation quality assessment: Linguistic description versus social
evaluation. Meta: Journal Des Traducteurs, 46(2), 243–257. https://doi.org/10.7202/003141ar.

House, J. (2015). Translation quality assessment: Past and present. In Translation: A


multidisciplinary approach (pp. 241–264). London and Chicago: Palgrave Macmillan.

Jakobson, R. (1959/1966/2000). On linguistic aspects of translation. In R. A. Brower (Ed.), On


translation. Cambridge, MA: Harvard University Press.

Kenny, D. (1998). Equivalence. In M. Baker (Ed.), Routledge encyclopaedia of translation studies.


London and New York: Routledge.

Koller, W. (1976/1979). Einführung die Übersetzungswissenschaft. Heidelberg: Quelle and Meyer.

Koller, W. (1989). Equivalence in translation theory. In A. Chesterman (Ed. & Trans.), Readings in
translation theory (pp. 99–104). Helsinki: Oy Finn Lectura Ab.

Uploaded by S. M. Safi
Koller, W. (1995). The concept of equivalence and the object of translation studies. Target, 7(2),
191–222.

Larson, M. (1998). Meaning-based-translation. Oxford: University Press of American Inc.

Leuven-Zwart, K. V. (1989). Translation and originals: Similarities and dissimilarities I.


Amsterdam: John Benjamins.

Manfredi, M. (2008). Translating text and context: Translation studies and systemic functional
linguistics (Vol. 1) Translation theory (2nd ed., p. 97). Bologna: Centro di Studi Linguistico-
Culturali (CeSLiC). https://doi.org/10.6092/unibo/amsacta/2441. In: Quaderni del CeSLiC.
Functional Grammar Studies for Non-Native Speakers of English.

Manfredi, M. (2014). Translating text and context: Translation studies and systemic functional
linguistics. Volume 2: From theory to practice (p. 158). Bologna: Centro di Studi Linguistico-
Culturali (CeSLiC). https://doi.org/10.6092/unibo/amsacta/3219. In: Quaderni del CeSLiC.
Functional Grammar Studies for Non-Native Speakers of English.

Mishra, P. (2009, September). Strength for today and bright hope for tomorrow etymological
analysis of the English language words. Language in India, 12, 63–75.

Munday, J. (2001). Introducing translation studies. New York: Routledge.

Munday, J. (2008). Introducing translation studies: Theories and applications (2nd ed.). London
and New York: Routledge.

Munday, J. (2009). The Routledge companion to translation studies. London and New York:
Routledge.

Newmark, P. (1981). Approaches to translation. Oxford: Pergamon.

Newmark, P. (1988). A textbook of translation: Text. Hertfordshire: Pearson Education Limited.

Newmark, P. (1991). About translation. Clevedon: Multilingual Matters.

Newmark, P. (1998). Approaches to translation. Singapore: Prentice Hall.

Newmark, P. in Munday, J. (2009). The Routledge companion to translation studies. London and
New York: Routledge.

Ni, L. (2009). For translation and theories. English Language Teaching, 2(2), 78–83.

Nida, E., & Taber, C. (1982). The theory and practice of translation (2nd ed.). Leiden: E.J. Brill.

Nida, E. A. (1959/1975). Principles of translating as exempli ied by Bible translating. In A. S. Dil


(Ed.), Language structure and thought: Essays by Eugene A. Nida. Stanford: Stanford University
Press.

Nida, E. A. (1964). Toward a science of translating: With special reference to principles and
procedures involved in Bible translating. Leiden: E.J. Brill.

Nida, E. A. (2000). Principles of correspondence. In L. Venuti (Ed.), The translation studies readers

Uploaded by S. M. Safi
(pp. 126–140). London and New York: Routledge (First published in 1964).

Nida, E. A., & Taber, C. R. (1969). The theory and practice of translation. Leiden: E.J.Brill.

Nord, C. (1989). Loyalty instead of loyalty. Proposals for a functional translation typology. Living
Languages, 34(3), 100–105.

Nord, C. (1997). Translating as a purposeful activity. Manchester: St. Jerome.

Nord, C. (2005). Text analysis in translation (2nd ed.). Amsterdam: Editions Rodopi B.V.

Nord, C. (2008). Persuading by addressing: A functional approach to speech-act comparison.


Southern African Linguistics and Applied Language Studies, 26(2), 283–293.

Panou, D. (2013). Equivalence in translation theories: A critical evaluation. Theory and Practice in
Language Studies, 3(1), 1–6.

Popovič, A. (1976). Dictionary for the analysis of literary translation Edmonton. Alberta:
Department of Comparative Literature, University of Alberta.

Reiss, K. (1971). Möglichkeiten und Grenzen der Übersetzungskritik. Munich: M. Hueber [Trans. E.
F. Rhodes. (2000). Translation criticism: Potential and limitations]. Manchester: St. Jerome and
American Bible Society.

Shakernia, S. (2014). Study of Nida’s (formal and dynamic equivalence) and Newmark’s (semantic
and communicative translation) translating theories on two short stories. Merit Research Journal
of Education and Review, 2(1), 1–7.

Snell-Hornby, M. (1988/1995). Translation studies: An integrated approach. Amsterdam and


Philadelphia: John Benjamins.

Steiner, G. (1975/1998). After Babel: Aspects of language and translation. Oxford and New York:
Oxford University Press.

Toury, G. (1980). In search of a theory of translation (p. 159). Tel Aviv: The Porter Institute for
Poetics and Semiotics, Tel Aviv University.

Venuti, L. (2000). Translation, community, Utopia. In L. Venuti (Ed.), The translation studies
reader (pp. 468–488). London: Routldge.

Venuti, L. (2004). The translation studies reader. London and New York: Routledge.

Vermeer, H. (1986). Übersetzen als kultureller transfer. In M. Snell-Hornby (Ed.) (1990), Linguistic
transcoding or cultural transfer? A critique of translation theory in Germany (pp.79–86). In S.
Bassnett & A. Lefevere (Eds.).

Vinay, J. P., & Darbelnet, J. (1958/2004). A methodology for translation. In J. C. Sager & M.-J. Hamel
(Trans.) & L. Venuti (Ed.), The translation studies reader (pp. 128–137). London and New York:
Routledge.

Wang, J. (2010). A critical discourse analysis of Barack Obama’s speeches. Journal of Language
Teaching and Research, 1(3), 254–261. https://doi.org/10.4304/jltr.1.3.254-261.

Uploaded by S. M. Safi
Wendland, E. R. (2012). Framing the frames: A theoretical framework for the cognitive notion of
“Frames of Reference.” Journal of Translation, 6(1), 27–50.

Whang, Y. C. (2004). To whom is a translator responsible—Author or reader? In S. E. Porter & R. S.


Hess (Eds.), Translating the Bible: Problems and prospects (pp. 46–62). New York: Continuum
and T&T Clark International.

Wilss, W. (1982). The science of translation: Problems and methods. Tübingen: Gunter Narr.

Yinhua, X. (2011). Equivalence in translation: Features and necessity. International Journal of


Humanities and Social Science, 1(10), 1989–1991.

Zhuanglin, H. (1988). A course of linguistics. Peking: Peking University Press.

Uploaded by S. M. Safi
© The Author(s) 2020
N. Abdelaal, Translation between English and Arabic
https://doi.org/10.1007/978-3-030-34332-3_3

3. Grammatical Problems in Translation


Noureldin Abdelaal1
(1) University of Nizwa, Nizwa, Oman

Noureldin Abdelaal

Overview This chapter explicates the grammatical problems in


translation between Arabic and English, and provides examples of
such problems and how some translators have dealt with them. The
author also suggests some strategies for dealing with such
problems.
This chapter covers the following topics:
1.
Tense as a Problem in Translation
2.
Gender as a Problem in Translation
3.
Grammatical Category as a Problem in Translation
4.
Syntactic Order: Foregrounding and Backgrounding as a
Problem in Translation
5.
Shifting (iltifat) as a Problem in Translation
6.
Passivization.

Uploaded by S. M. Safi
Grammar was de ined by Baker (1992/2011, p. 83) as ‘the set of rules
which determine the way in which units such as words and phrases can
be combined in a language and the kind of information which has to be
made regularly explicit in utterances’. According to Ghazala (2008),
problems of translation can be at the grammatical, stylistic, lexical or
phonological levels. Problems of grammar mainly arise from the
complications of the SL grammar; differences between an SL and a TL
in grammatical or syntactic aspects, which may be identi ied as a
grammatical gap in the TL; and the syntactic word order. However,
lexical problems, as Ghazala argues, can mainly arise from literal
translation, synonymy, polysemy and monosemy, collocations, idioms,
proverbs, metaphors, technical translation and culture. Stylistic
problems, on the other hand, arise from the levels of formality and
informality in a language, fronting, parallelism, ambiguity, the degree of
complexity, short sentences vs. long sentences, passive vs. active text,
repetition and variation, redundancy, nominalization vs. verbalization,
irony and the translation of punctuation. However, as far as this study
is concerned, the major problems in translation are grammatical and
semantic problems. Each of these problems will be discussed in the
sections that follow.
To draw a distinction between grammar and meaning is a hard job
that is unlikely to be achieved because grammar basically serves
meaning (Cruse, 1997). Grammar is discussed here as an integral part
of meaning. In his book Approaches to Translation, Newmark
(1981/2001) mentioned that grammatical meaning is more important
than lexical meaning because it shows the tone of an SL. He also states
that grammatical meaning can be rendered by more or less standard
transpositions. He also argues that all lexical meanings have embedded
grammatical meanings. He explicates his idea by stating that a lexical
meaning starts when grammatical meaning inishes. Transferring such
grammatical meaning from an SL to a TL poses many problems in
translation.
Grammatical problems in translation are likely to be more
complicated when the translation process takes place between two
different language families, such as English and Arabic (Ghazala, 2008).
Arabic is a Semitic language, while English is a West Germanic

Uploaded by S. M. Safi
language. These variations between the two languages result in
differences in the grammatical and syntactic patterns. One main
grammatical problem is the differences between the types of sentence
in English and Arabic. The Arabic language has mainly three types of
sentences: nominal, verbal and non-functional (Ghazala, 2008). Each
type of sentence comprises many subtypes that have no equivalents in
English. In Arabic, nominal sentences in some of their subtypes do not
have verbs. By contrast, English sentences must have a verb. The
translation of conditional clauses is another grammatical problem in
translation, since Arabic only has two types of conditional clause, while
English has three. The change of the word class in translation is a
further problem in translation (Ghazala, 2008). All these grammatical
differences pose quandaries for a translator; he cannot render an
Arabic text into English without altering the textual pattern. Such a
prospective change in textual pattern inevitably affects the meaning
transferred to the TT. In addition, the syntactic and grammatical
variations between Arabic and English create a lexical gap in the
translation between an ST and a TT. For example, the cognate object is
not represented in the English language system, and thus rendering it
into English creates a lexical gap.

3.1 Arabic Tense as a Problem in Translation


Tense is one of the prominent grammatical problems in translation. It
is dif icult for a translator to preserve the tense of the ST, and thus the
meaning may be susceptible. Lack of tense equivalence between Arabic
and English poses many problems in translation. Arabic has only three
tenses: past, present (aorist) and future. These tenses are unpacked
below:
a. The past
The following example highlights the problem of translating the
past tense from Arabic to English.

Example ‫ذﻫﺐ أﺣﻤﺪ اﻟﻰ اﻟﻤﺪرﺳﺔ‬


The verb ‫ذﻫﺐ‬, which is in the past tense, can be rendered into
different aspects in English based on the context. It can be rendered
as:

Uploaded by S. M. Safi
I. ‘Ahmed went to school.’

II.
‘Ahmed had gone to school.’
III.
‘Ahmad had been to school.’
IV.
‘Ahmad has gone to school.’

Example I shows a past action that is inished, while example II


indicates a past action that happened prior to another past action. As
for example III, it highlights the action of going to school and coming
back. As for example IV, it refers to a recent action of going to school, or
to link the action to a current context of a situation. A translator should,
therefore, decide on the right translation based on the meaning
intended in the ST. However, rendering the English past tense into
Arabic does not cause any problems, as it has only one available option.
Another example is provided below.

Example ‫ﻛﻨﺖ ﺳﺎذﻫﺐ ﻟﻠﺴﻮق‬

The example above shows an intention in the past, while the action
was not completed in the past. In translation, this should be conveyed.
It can, then, be translated as: ‘I was about to go to the market’. However,
the same verb ‫ ﻛﻨﺖ‬sometimes implicates a different meaning. Consider
the following example:
‫ﻛﻨﺖ ﻗﺪ ﻗﺎﺑﻠﺖ أﺣﻤﺪ‬
This example indicates a past action, and not merely a past
intention. So, it should be rendered as:

Example
a.
‘I met Ahmed.’
b.
‘I have met Ahmed.’
c.
‘I had met Ahmed.’

Uploaded by S. M. Safi
The following is another example that shows how tense is
sophisticated in Arabic:

Example ‫ﻛﻨﺖ اﻟﻌﺐ ﻓﻲ اﻟﻄﺮﯾﻖ‬

This example indicates an action that was in process in the past, and
therefore the past continuous aspect or the past perfect continuous
aspect in English is equivalent to the Arabic text. It, then, should be
rendered as:

Example
a.
‘I was playing in the street/on the road’; or
b.
‘I had been playing in the street/on the road.’

b. The present
The present tense in Arabic needs lexical support to indicate the
aspect. Consider the following examples for the purposes of
clari ication:

Example
1.
‫ﯾﻠﻌﺐ اﻷﻃﻔﺎل داﺋﻤﺎ ﻓﻲ اﻟﺠﺪﯾﻘﺔ‬
2.
‫ﯾﻠﻌﺐ اﻷﻃﻔﺎل اﻻن ﻓﻲ اﻟﺠﺪﯾﻘﺔ‬

These two examples may not cause a serious problem in translation as


they can be rendered as:

Example
1a.
‘The children always play in the garden.’
2a.
‘The children are playing in the garden now.’

As seen in the translations above, example 1 was rendered into the


present simple tense, while example 2 was rendered into the present

Uploaded by S. M. Safi
continuous. One challenge may be translating the imperfect verb in
Arabic into English. Consider the following example:

Example ‫ﻻ ﯾﺘﻮﻗﻒ اﻷﻃﻔﺎل ﻋﻦ اﻟﻠﻌﺐ ﻓﻲ اﻟﺸﺎرع‬

This example can be rendered into the present simple. It can be


rendered as: ‘Children never stop playing in the street.’ Consider the
following example:

Example ‫ﻟﻢ ﯾﺘﻮﻗﻒ اﻷﻃﻔﺎل ﻋﻦ اﻟﻠﻌﺐ ﻓﻲ اﻟﺸﺎرع‬

This example can be translated as: ‘Children did not stop playing in the
street.’
The imperfect verb was rendered into the past simple in this case.
However, the imperfect verb ‫ ﯾﺘﻨﺎﻫﻮن‬is mostly translated into the past
simple or past continuous in English. So, a translator needs to translate
the Arabic imperfect verb based on the context of meaning.
c. The future
Future does not seem to be a problem in translation from Arabic to
English though, similar to the present tense, Arabic employs lexical
devices to express different aspects. Consider the following examples:

Example ‫ﺳﻮف أذﻫﺐ اﻟﻰ اﻟﺴﻮق ﻏﺪا‬


‫ﺳﻮف اﻛﻮن ﻗﺪ اﻧﺘﻬﯿﺖ ﻣﻦ ﻫﺬا اﻟﻤﺸﺮوع ﺑﺘﻬﺎﯾﺔ ﻫﺬا اﻟﺸﻬﺮ‬
‫ﻏﺬا ﻓﻲ ﻫﺬا اﻟﻮﻗﺖ ﺳﺎﻛﻮن اﺟﻠﺲ ﻷﻣﺘﺤﺎن اﻟﻜﯿﻤﯿﺎء‬

These examples can be rendered as:

Example ‘I will go to the market tomorrow.’


‘I will have inished this project by the end of next month.’
‘Tomorrow, at this time, I will be sitting the chemistry exam.’

The Holy Quran is a rich resource of examples of this kind of problem.


An example of problems in translating tense is provided by Ali et al.
(2012) from Surah al-Ahzab, ayah 10, which reads:

Example (33:10) ‫ﺎر َوﺑَﻠَ َﻐ ِﺖ‬ ُ‫ﺼ‬َ ‫اﻏ ِﺖ اﻷَ ْﺑ‬


َ ‫ِﻜ ْﻢ َو ِﻣ ْﻦ أَ ْﺳﻔَ َﻞ ِﻣ ْﻨ ُﻜ ْﻢ َوإِ ْذ َز‬
ُ ‫إِ ْذ َﺟﺎ ُءو ُﻛ ْﻢ ِﻣ ْﻦ ﻓَ ْﻮﻗ‬
‫ﺎﷲ ﱡ‬
‫اﻟﻈﻨُﻮﻧَﺎ‬ ِ َ‫ﻮب اﻟْ َﺤﻨ‬
َ ‫ﺎﺟ َﺮ َوﺗَ ُﻈﻨﱡﻮن ِﺑ‬ ُ ُ‫اﻟْﻘُﻠ‬

Uploaded by S. M. Safi
‘Behold! They came on you from above you and from below you,
and behold, the eyes became dim and the hearts gaped up to the
throats, and ye imagined various (vain) thoughts about Allah!’ (Ali,
2006, p. 287) (33:10)

As seen in this example, there is a shift in the verb tense from the past
َ ‫ َز‬/zaghati/ and ‫ َﺑﻠَ َﻐ ِﺖ‬/balaghati/) in the
ُ ‫ َﺟﺎ ُء‬/jaookum/, ‫اﻏ ِﺖ‬
tense (‫وﻛ ْﻢ‬
Quranic ST to the present tense (‫ َوﺗَ ُﻈﻨﱡﻮن‬/watathunnoona/). This shift
in the Holy Quran recurs to invoke an important action in the mind as if
it were happening at the moment of reading the ayah. However, the
translation could not convey this stylistic effect, as the translator
followed English grammar and could not convey the shift that exists in
the ST. Although the problem in translation of this example is
grammatical, it affects the style of the text. Another example of tense
loss was discussed by Al-Azab and Al-Misned (2012) in the translation
of the following ayah:
ْ
Example (18:100) ‫ﯾﻦ َﻋ ْﺮﺿﺎ‬ ِ ‫َو َﻋ َﺮ ْﺿ َﻨﺎ ﺟﻬﻨﻢ ﯾَ ْﻮ َﻣﺌﺬ ﻟِﻠ َﻜﺎﻓ‬
َ ‫ِﺮ‬
‘And on that day we shall present Hell to the disbeliever’s plain to
view’. (Khan & Al-Hilali, 1996)
‘On that day we shall present Hell to the disbelievers, plain to
view’. (Pickthall, 2001, p. 125)
‘And upon that day we shall present Gehenna to the unbelievers’.
(Arberry, 1982, p. 180)

In the Quranic ST text, the verb ‫ َﻋ َﺮ ْﺿﻨَﺎ‬/WaAAradana/ is in the past


form, though the ayah talks about the Day of Judgement, which has not
occurred yet; this recurs in the Holy Quran, and is used to re lect the
certainty and inevitability of the occurrence of things, even in the
future. However, the translators rendered the past verb into the future
(i.e. shall present); they could not follow the same tense as the Arabic
Quranic text, and thus created grammatical loss, which affected the
meaning conveyed. The translations could not convey the overtones
and undertones of the ST. The past form of verbs is used in Arabic to
talk about facts, whereas this achieved differently in English.
Another grammatical loss in the ayah is the translation of the
cognitive object. In Arabic, cognitive objects are derived from the same

Uploaded by S. M. Safi
root as the verb for the purposes of con irmation. By contrast, English
does not have this kind of grammatical style, which forces a translator
to ind other words, phrases, or expressions to compensate for the loss
in translation. This compensation strategy is suggested by Nida and
Taber (1982) and Hervey and Higgins (1992), whereby they suggested
that a translator would compensate for a loss by making an addition.
Put simply, to create the same effect as an ST in the TT, a translator may
add some words that do not exist in the ST.

3.2 Translating English Tenses and Aspects


into Arabic (Based on Collins COBUILD English
Grammar, 2005)
English has two tenses, the past and the present, but has has perfect
aspects and imperfect aspects. Tense and aspects are expressed by the
addition of in lections to the base form of the verb, or sometimes by
the use of auxiliary verbs: ‘In situations where you are discussing an
existing state of affairs, you use a verb that is in the present tense’
(Collins COBUILD English Grammar, 2005, p. 414). These tenses and
aspects are unpacked below:
a. The present simple tense
The present simple tense is used to speak about the thoughts and
feelings at the present time or immediate reactions. Consider the
following examples:

Example
1.
‘I’m awfully busy.’
‫إﻧﻲ ﻣﺸﻐﻮل ﻟﻠﻐﺎﯾﺔ‬
2.
‘They both taste the same.’
‫إﻧﻬﻤﺎ ﺑﻨﻔﺲ اﻟﻤﺬاق‬

In these examples, example 1 was translated into English without


maintaining the verb ‘be’ that exists in the SL, and which is used
explicitly in English. However, in Arabic, it may be translated implicitly,
as in the example above. The same applies to example 2, as, in Arabic,

Uploaded by S. M. Safi
we usually render verbs that are related to senses as nouns. In example
2, the verb ‘taste’ was rendered as ‫اﻟﻤﺬاق‬, which is a noun. The present
simple tense can be also used to express physical feelings. Consider the
following examples:

Example
1.
‘I feel tired.’
‫أﺷﻌﺮ ﺑﺎﻟﺘﻌﺐ‬
2.
‘She feels sleepy.’
‫أﻧﻬﺎ ﺗﺸﻌﺮ ﺑﺎﻟﻨﻌﺎس‬

In examples 1 and 2, the verb was rendered as a verb in the TL but the
adjective was rendered as a noun. Of course, the sentences above can
be translated differently, as clauses. They can be translated as:

Example ‫أﺷﻌﺮ ﺑﺄﻧﻲ ﻣﺘﻌﺐ‬


‫أﺷﻌﺮ ﺑﺄﻧﻲ ﻧﻌﺲ‬

The present simple tense can be also used to express facts and truths.
Consider the following examples:

Example ‘Laila eats meat.’


‫ﻟﯿﻠﻰ ﺗﺄﻛﻞ اﻟﻠﺤﻢ‬

This example shows a fact about Laila, and should be rendered


carefully. It can be simply rendered as: ‫ﻟﯿﻠﻰ ﺗﺄﻛﻞ اﻟﻠﺤﻢ‬. Although the
translation seems simple and clear, a problem may arise due to the
incorrect perception of the Arabic translation as happening at the
moment of speaking. Therefore, clari ication may be needed in some
cases, which can be achieved through adding some aspectual words
such as ‘always’ ‫ داءﻣﺎ‬or ‘usually’ ‫ﻋﺎدة‬.
The present simple tense can be also used to express habits.
Consider the following examples:

Example 1. ‘I start work at 11.’


11 ‫أﺑﺪأ اﻟﻌﻤﻞ اﻟﺴﺎﻋﺔ‬

Uploaded by S. M. Safi
2. ‘I have lunch at work every day.’
‫أﺗﻨﺎول اﻟﻐﺪاء ﻓﻲ اﻟﻌﻤﻞ ﻛﻞ ﯾﻮم‬

Exercise Translate the following sentences into Arabic.


1.
‘Gosh, he looks awful.’
2.
‘I want a breath of fresh air.’
3.
‘My stomach hurts.’
4.
‘I hear approaching feet.’
5.
‘My dad works in Saudi Arabia.’
6.
‘He lives in the French Alps near the Swiss border.’
7.
‘A chemical reaction occurs in the fuel cell.’
8.
‘Babies normally lose weight in the beginning.’
9.
‘The attitude is usually one of ridicule.’
10.
‘Traditionally, the Japanese prefer good quality clothes.’

b. Present progressive
Present progressive is used in English to express an action
happening at the moment of speaking, an action that is continuing for
some time, or even an action that will happen in the future. It can also
be used to express changes, developments and trends. Consider the
following examples:

Exercise
a.
‘Rabiaa is cooking dinner.’

Uploaded by S. M. Safi
b. ‘Rabiaa is leaving for Oman next week.’

c.
‘Rabiaa is doing her PhD.’

Example a indicates an action that was happening at the moment of


speaking, whereas example b expresses a near future action. Example c
expresses an action that will be taking place for several months, or
maybe years. They should, therefore, be rendered, respectively, as
follows:

Exercise
a.
‫رﺑﯿﻌﺔ ﺗﻄﺒﺦ اﻟﻌﺸﺎء اﻻن‬
b.
‫رﺑﯿﻌﺔ ﺳﺘﺴﺎﻓﺮ اﻟﻰ ﻋﻤﺎن اﻻﺳﺒﻮع اﻟﻘﺎدم‬
c.
‫رﺑﯿﻌﺔ ﺗﺪرس اﻟﺪﻛﺘﻮراة ﺣﺎﻟﯿﺎ‬

Notice that in a, the Arabic adverbial word ‫ اﻻن‬was added to clarify the
meaning. In example b, the pre ixed letter ‫ س‬was added to the TT verb
to express futurity. In example c, the adverbial word ‫ ﺣﺎﻟﯿﺎ‬was added to
the TT to clarify the meaning. A translator should not shy away from
adding words to the TT to clarify the meaning of the tense or the
aspect, if needed. However, in some cases, we may not need to add any
words, as the meaning is clear from the context. This happens mostly
when the present continuous is used in combination with another
tense or aspect. By way of illustration, see the following example:

Example
‘I do not like to go out now because it is raining.’
‫ﻻ أﺣﺐ أن أﺧﺮج اﻷن ﻷﻧﻬﺎ ﺗﻤﻄﺮ‬

In this example, we do not need to add any words that mark the
continuous aspect because continuity is evidenced by the word ‘now’
in the irst clause.

Uploaded by S. M. Safi
Exercise Translate the following sentences into Arabic.
1.
‘The village is changing but it is still undisturbed.’
2.
‘His handwriting is improving.’
3.
‘World energy demand is increasing at a rate of about 3% per
year.’
4.
‘I’m not planning on having children at present.’
5.
‘Nowadays itness is becoming a generally accepted principle of
life.’

c. Present perfect aspect


Present perfect expresses an action that started recently in the
past, or that started in the past and will continue into the future for
some time. Arabic does not have an equivalent aspect. Consider the
following example and how it could be translated:
1.
‘Rabeea has eaten her lunch.’
The example above can be translated as ‫ﺗﻨﺎوﻟﺖ رﺑﯿﻌﺔ ﻏﺪاءﻫﺎ‬. The
translation does not show that she has had lunch recently, but this can
be understood from the context. In Arabic, a past tense can refer to an
action or event that occurred a long time ago or recently but,
sometimes, we may need some lexical markers to indicate the point at
which the action or event occurred. In a natural Arabic context, the
translation above can be perceived as referring to a recent past action.
Note the following examples:

Example
2.
‘Sumaya has lived in this house since 2009.’

Uploaded by S. M. Safi
The above example can be translated as: 2009 ‫ﺳﻜﻨﺖ ﺳﻤﯿﺔ ﻫﺬا اﻟﺒﯿﺖ ﻣﻨﺬ ﻋﺎم‬.
The translation is also in the past tense, which sounds correct. Note
how the following example is translated differently:

Example
3.
‘Laila has not inished her homework yet.’

This example can be translated as ‫ﻟﻢ ﺗﻨﺘﻪ ﻟﯿﻠﻰ ﻣﻦ واﺟﺒﻬﺎ ﺑﻌﺪ‬. Here, it can be
noted that the present perfect was translated into present preceded by
a negating particle, which changes the meaning of the present verb into
a past tense.
In short, the present perfect aspect is mostly translated into the
past tense in Arabic but, in some cases, lexical markers may be needed
to explicate the meaning. In the negative case of the present perfect
aspect, it is translated to a present tense in Arabic.

Exercise Translate the following sentences into Arabic.


1.
‘They have raised £180 for a swimming pool.’
2.
‘I have noticed this trait in many photographers.’
3.
‘The settlers have left the bay forever.’
4.
‘I ate brown rice, which I have always hated, and vegetables from
my garden.’
5.
‘They have been back every year since then.’
6.
‘She has worked for him for ten years.’

d. Present perfect continuous


The present perfect continuous is used to emphasize the duration
of an action. It is mostly rendered to an imperfect verb in Arabic, as
shown in the following example:

Uploaded by S. M. Safi
‘I have been waiting you for an hour.’
This example can be translated as ‫أﻧﻲ أﻧﺘﻈﺮك ﻣﻨﺬ ﺳﺎﻋﺔ‬. It can also be
translated as ‫اﻧﺎ ﻓﻲ اﻧﺘﻄﺎرك ﻣﻨﺬ ﺳﺎﻋﺔ‬. As can be seen, the present perfect
tense was translated into the present tense, or into an adjectival word.

Exercise Translate the following sentences into Arabic.


1.
‘She’s been crying.’
2.
‘Some people will say that what I have been describing is not a
crisis of industry.’
3.
‘The Department of Aboriginal Affairs has recently been
conducting a survey of Australian Aborigines.’
4.
‘Until now the rumours that had been circulating were
exaggerated versions of the truth.’
5.
‘The doctor had been working alone.’
6.
‘He died in hospital where he had been receiving treatment for
cancer.’
7.
‘They had been hitting our trucks regularly.’

e. Past simple tense


The past simple tense is used in English to express an action that
happened at a speci ic point in the past. Consider the following
example and the suggested translation:

Example ‘I met the president yesterday.’


‫ﻗﺎﺑﻠﺖ اﻟﺮﺋﯿﺲ أﻣﺲ‬

As seen in this example, it seems that there is no problem in translating


the past simple tense.

Uploaded by S. M. Safi
Exercise Translate the following sentences into Arabic.
1.
‘The Israeli Prime Minister lew into New York yesterday to start
his visit to the USA.’
2.
‘Our regular window cleaner went off to Canada last year.’
3.
‘On 1 February 1968 he introduced the Industrial Expansion Bill.’
4.
‘They gave me medication to help me relax.’
5.
‘I bought a new car yesterday.’
6.
‘I got my PhD from a famous Malaysian university.’
7.
‘He lived in Paris during his last years.’
8.
‘Throughout his life he suffered from epilepsy.’

f. Past progressive
The past progressive is used to express an action that occurred for
a length of time in the past, highlighting the action. See the following
example, for further clari ication:

Example

‘I was playing football by this time yesterday.’

‫ﻛﻨﺖ اﻟﻌﺐ ﻛﺮة ﻗﺪم ﻓﻲ ﻣﺜﻞ ﻫﺬا اﻟﻮﻗﺖ أﻣﺲ‬

As seen in this example, the past continuous was translated as an


imperfect verb in Arabic, which seems to convey the meaning of the ST.

Exercise Translate the following sentences into Arabic.


1. ‘Her tooth was aching, her burnt inger was hurting.’

Uploaded by S. M. Safi
2.
‘He was looking ill.’
3.
‘Everyone was begging the captain to surrender.’
4.
‘I was meeting thousands of people and getting to know no one.’
5.
‘We were all sitting round there waiting for my brother to come
home.’
6.
‘He arrived at about six in the evening.’
7.
‘I was waiting angrily on Monday morning when I saw Mrs Miller.’

g. Past perfect
The past perfect is used to express an action that happened prior to
another action in the past. See the following example:

Example

‘By the time I arrived, my father had already gone.’

This example indicates that the arrival of the speaker happened after
the departure of their father. It can thus be translated as ‫ﻋﻨﺪﻣﺎ وﺻﻠﺖ ﻛﺎن‬
‫أﺑﻲ ﻗﺪ ﻣﻀﻰ‬, which is an overt copula + emphatic word + the past tense of
the main verb.
h. Past perfect continuous
The past perfect continuous is similar in use to the past continuous
and therefore its translation is the same. Consider the following
examples:

Example

‘Ahmed was studying all night.’

Uploaded by S. M. Safi
‘Ahmed had been studying all night.’

These two examples can be rendered as ‫ﻛﺎن أﺣﻤﺪ ﯾﺬاﻛﺮ ﻃﻮال اﻟﻠﯿﻞ‬. The
Arabic language does not discriminate between the past perfect
continuous and the past continuous.
i. Future simple
The future simple is simply rendered into the future in Arabic using
lexical devices such as ‫ ﺳﻮف‬followed by the present form (aorist) of the
verb. Another way is to use the letter ‫ س‬as a pre ix to the verb. Consider
the following example:

Example

‘I will open the door.’

‫ﺳﻮف أﻓﺘﺢ اﻟﺒﺎب‬

‫ﺳﺄﻓﺘﺢ اﻟﺒﺎب‬

Both these translations sound acceptable, though the second one


sounds more natural and idiomatic.

Exercise Translate the following sentences into Arabic.


1.
‘I shall do everything I can to help you.’
2.
‘You will stay at home and I shall go to your of ice.’
3.
‘We shall give him some tea.’

j. Future continuous
The future continuous expresses an action that is ensured to
happen in the future:

Example

Uploaded by S. M. Safi
‘I will be eating my lunch by 8.’

8 ‫ﺳﻮف أﻛﻮن اﺗﻨﺎول ﻏﺪاﺋﻲ ﻓﻲ اﻟﺴﺎﻋﺔ ال‬

As can be seen, the future continuous was translated similarly to the


future simple, with the exception of inserting the past form of the
copula between the word expressing the future (i.e. ‫ )ﺳﻮف‬and the aorist
tense.

Exercise Translate the following sentences into Arabic.


1.
‘I’ll be seeing them when I’ve inished with you.’
2.
‘She’ll be appearing tomorrow and Sunday at the Royal Festival
Hall.’
3.
‘I’ll be waiting for you outside.’
4.
‘I understand you’ll be moving into our area soon.’
5.
‘They’ll spoil our picnic. I’ll be wondering all the time what’s
happening.’
6.
‘Our people will be going to their country more.’

k. Future perfect
The future perfect is used in English to express an action that will
be completed at a speci ic point in the future:

Example

‘By next Monday, I will have inished writing this book.’

‫ﻓﻲ اﻷﺛﻨﯿﻦ اﻟﻘﺎدم ﺳﺄﻛﻮن ﻗﺪ اﻧﺘﻬﯿﺖ ﻣﻦ ﻛﺘﺎﺑﺔ ﻫﺬا اﻟﻜﺘﺎب‬

Uploaded by S. M. Safi
As seen in this example, the English future perfect aspect was
translated into the pre ix letter ‫س‬, which implies the future, in addition
to ‫أﻛﻮن‬, which is equivalent to the verb ‘be’ in English, followed by ‫ﻗﺪ‬,
which is used in Arabic for the purpose of emphasis, and after that the
verb ‫ اﻧﺘﻬﯿﺖ‬is used in the past to express an action that will be
completed at that speci ic point in the future. In the example, the pre ix
letter ‫ س‬can be replaced by the lexeme ‫ﺳﻮف‬, which functions in the
same way as the aforementioned pre ix letter.

Exercise Translate the following sentences into Arabic.


1.
‘By the time you get to the school, the concert will have inished.’
2.
‘Maybe by the time we get there he’ll already have started.’
3.
‘By then, maybe you’ll have heard from your sister.’

l. Future perfect progressive


Future perfect progressive is used to refer to the duration of an
event at a speci ic time in the future. Consider the following example:

Example

‘By the time the season ends, I will have been playing for ifteen
months without a break.’
‫ ﺷﻬﺮا دون ﺗﻮﻗﻒ‬15 ‫ﺑﻨﻬﺎﯾﺔ اﻟﻔﺼﻞ ﺳﺎﻛﻮن اﻟﻌﺐ ﻟﻤﺪة‬

As seen in this example, the ST English sentence was rendered in the


same way that the future perfect was rendered. This is because Arabic
does not differentiate between future perfect and future perfect
progressive.

Exercise Translate the following sentences into Arabic.


1.
‘The register will have been running for a year in May.’

Uploaded by S. M. Safi
2. ‘I will have been studying English for four years in next June.’

3.
‘She will have been cooking for ive hours continuously by 6 pm.’

m. Using ‘be going to’ to express the future


‘Be going to’ followed by the in initive form is used to express
events in the future that are likely to happen based on evidence.
Consider the following example:

Example

‘The sky is cloudy. It is going to rain.’

‫ ﯾﺒﺪوا أﻧﻬﺎ ﺳﺘﻤﻄﺮ‬.‫اﻟﺴﻤﺎء ﻣﻠﺒﺪة ﺑﺎﻟﻐﯿﻮم‬

As seen in this example, the future was translated into Arabic in the
same way that the future simple is rendered.

Exercise
Translate the following sentence into Arabic.
1.
‘I’m going to explore the neighbourhood.’
2.
‘Evans knows lots of people. He’s going to help me. He’s going to
take me there.’
3.
‘You’re going to have a heart attack if you’re not careful.’
4.
‘We’re going to see a change in the law next year.’

n. Using ‘due to’ to express futurity


‘Due to’ followed by an in initive is used to express events due to
take place in the near future. By way of illustration, see the following
example:

Uploaded by S. M. Safi
Example

‘He is due to start as a courier shortly.’

‫إﻧﻪ ﺳﻮف ﯾﺒﺪا ﻋﻤﻠﻪ ﻛﺴﺎﻋﻲ ﺑﺮﯾﺪ ﻗﺮﯾﺒﺎ‬

It can be noted that ‘due to’ was translated as ‫ﺳﻮف‬, similar to the
translation of ‘will’.

Exercise Translate the following sentences into Arabic.


1.
‘The work is due to be started in the summer.’
2.
‘Another 385 people are about to lose their jobs.’
3.
‘Are we about to be taken over by the machine?’

o. Present simple to express futurity


The present simple tense can be also used to express the future,
especially when it refers to scheduled events. Consider the following
example:

Example

‘The last train to Cairo leaves at 10:00 am.’

‫أﺧﺮ ﻗﻄﺎر ﻣﺘﺠﻪ ﻟﻠﻘﺎﻫﺮة ﯾﻐﺎدر اﻟﺴﺎﻋﺔ اﻟﻌﺎﺷﺮة ﺻﺒﺎﺣﺎ‬

In this example, we have not used any futurity markers (e.g. ‫)ﺳﻮف‬. The
verb was simply translated to an equivalent verb in the TL.

Exercise Translate the following sentences into Arabic.


1.
‘My last train leaves Euston at 11.30.’
2.
‘The UN General Assembly opens in New York later this month.’

Uploaded by S. M. Safi
3. ‘Tomorrow morning we meet up to exchange contracts.’

3.3 Gender as a Problem in Translation


Another type of grammatical problem in translation is the translating
of gender. Arabic, unlike English, discriminates nouns in terms of
gender across all grammatical cases. Consider the following example in
Surah Al-Hajj (as provided by Al-Azab & Al-Misned, 2012), ayah 2:

Example

َ ‫ﯾَ ْﻮ َم ﺗَ َﺮ ْوﻧَ َﻬﺎ ﺗَ ْﺬ َﻫ ُﻞ ﻛﻞ ﻣﺮﺿﻌﺔ ﻋﻤﺎ أَ ْر‬


(22:2) ‫ﺿ َﻌ ْﺖ‬

‘The day you shall see it, every nursing mother will forget her
nursling’. (Khan & Al-Hilali, 1996)

‘On the day when ye behold it, every nursing mother will forget
her nursling’. (Pickthall, 2001, p. 136)

‘On the day when you behold it, every suckling woman shall
neglect the child she has suckled’. (Arberry, 1982, p. 202)

As seen in the example provided by Al-Azab and Al-Misned (2012), the


Quranic word ‫ﻣﺮﺿﻌﺔ‬/murdiAAatin/ is in the feminine case in the ST,
which, according to Al-Zamakhshari (2000), serves to imply the
current status of breast-feeding, because the Quranic word is feminine
even without the marker of femininity. However, the translators sought
to translate it using two words (i.e. nursing mother, suckling woman)
to show femininity. In spite of that, there is a loss in translation
because the current status of nursing is not re lected in the translation.
They could have used a word such as ‘now’ to compensate the loss in
meaning (Al-Azab & Al-Misned, 2012). Further examples of gender
differences between English and Arabic Follow:

Example

Uploaded by S. M. Safi
1. ‫ﺗﺰوج ﻧﻮر ﻓﺘﺎة ﺟﻤﯿﻠﺔ‬
‘Nour married a beautiful girl.’
2.
‫اﻟﻘﻤﺮ ﺟﻤﯿﻞ اﻟﻠﯿﻠﺔ‬
‘The moon is beautiful tonight.’

In these examples, the Arabic adjectives ‫ ﺟﻤﯿﻞ‬and ‫ ﺟﻤﯿﻠﺔ‬refer to the same


sense with the only difference being the addition of the gender marker
in example 1. Example 1 indicates the use of the adjective in the
feminine form, while in example 2, the adjective is used in the
masculine form. However, the two adjectives will be translated to the
same lexeme in English: beautiful. Note the gender differences across
all the grammatical cases in the examples provided below.

Example

TT ST
‘Hazem bought a beautiful car.’ ‫اﺷﺘﺮى ﺣﺎزم ﺳﯿﺎرة ﺟﻤﯿﻠﺔ‬
‘Hazem bought two beautiful cars.’ ‫اﺷﺘﺮى ﺣﺎزم ﺳﯿﺎرﺗﺎن ﺟﻤﯿﻠﺘﺎن‬
‘Hazem bought three beautiful cars.’ ‫اﺷﺘﺮى ﺣﺎزم ﺛﻼث ﺳﯿﺎرات ﺟﻤﯿﻼت‬
‘Hazem bought a beautiful house.’ ‫اﺷﺘﺮى ﺣﺎزم ﺑﯿﺘﺎ ﺟﻤﯿﻼ‬
‘Hazem bought two beautiful houses.’ ‫اﺷﺘﺮى ﺣﺎزم ﺑﯿﺘﯿﻦ ﺟﻤﯿﻠﯿﻦ‬
‘Hazem bought three beautiful houses.’ ‫اﺷﺘﺮى ﺣﺎزم ﺛﻼث ﺑﯿﻮت ﺟﻤﯿﻠﯿﻦ‬

In these examples, the Arabic adjective ‫ ﺟﻤﯿﻞ‬has undergone different


lexical changes based on the gender and number. In contrast, the
English adjective has not undergone any changes across all the
grammatical categories.

Exercise Translate the following sentences into English.


1.
‫ﻓﻲ ﺑﯿﺘﻨﺎ ﺷﺠﺮة ﻣﺜﻤﺮة‬
2.
‫اﻟﻌﻤﻞ اﻟﺪؤوب داﯾﻤﺎ ﻣﺎ ﯾﻜﻮن ﻣﺜﻤﺮا‬
3. ‫ﻓﻲ اﻟﺠﻨﺔ ﺛﻤﺎر ﻟﺬﯾﺬة‬

Uploaded by S. M. Safi
4.
‫اﻟﺘﻔﺎح ﻟﺬﯾﺬ اﻟﻄﻌﻢ‬

3.4 Grammatical Category as a Problem in


Translation
Grammar presents problems in translation between English and
Arabic. Al-Azab and Al-Misned (2012) noted that, sometimes,
translators change plural forms into singular forms, or vice versa, to
follow the norms of the TL. The following example in Surah al-Nahl,
ayah 80, was provided by Al-Azab and Al-Misned (2012):

Example
َ َ َ
ِ ‫ﺎر َﻫﺎ َوأ ْﺷ َﻌ‬
(16: 80) ‫ﺎر َﻫﺎ‬ ِ َ‫َو ِﻣ ْﻦ أ ْﺻ َﻮاﻓِ َﻬﺎ َوأ ْوﺑ‬
‘And of their wool, fur, and hair’. (Khan & Al-Hilali, 1996)

‘And of their wool and their fur and their hair’. (Pickthall, 2001, p.
115)

‘And of their wool, and of their fur and of their hair’. (Arberry,
1982, p. 164)

As seen in these translations, the translators of the ayah had to change


the plural form of the three nouns in the ST (i.e. aswaaf, ashAAar,
awbaar) into the singular form in the TT (i.e. wool, fur, hair), to follow
the grammatical norms of English. However, this is a tolerable loss
because the meaning is not lost; it still carries the same meaning of
plural for a native speaker of English. Some other examples of the
differences in the grammatical category between Arabic and English
are illustrated by the following examples:

Example

Uploaded by S. M. Safi
1. ‫داﺋﻤﺎ ﻣﺎ ﯾﻘﺪم اﻷﺑﺎء اﻟﻨﺼﺎﺋﺢ ﻷﺑﻨﺎﺋﻬﻢ‬

‘Parents always give advice to their kids.’

2. ‫ﯾﺤﺘﻮي ﻫﺬا اﻟﻜﺘﺎب ﻋﻠﻰ اﻟﻌﺪﯾﺪ ﻣﻦ اﻟﻤﻌﻠﻮﻣﺎت اﻟﻤﻔﯿﺪة‬

‘This book contains a lot of useful information.’

In these examples, the ST words ‫ اﻟﻨﺼﺎﺋﺢ‬and ‫ اﻟﻤﻌﻠﻮﻣﺎت‬are in the plural


form, as in Arabic these two words have singular, dual and plural forms.
However, in English, they have only the uncountable form. Therefore, it
is of primary importance that a translator pays attention to such
linguistic differences. Another feature of the Arabic language that does
not exist in English is dualism. So, when translating from Arabic to
English, a translator needs to add the word ‘two’ to clarify the meaning
of duality that exists in Arabic. By way of illustration, see the following
examples:

Example

1. ‫أﺷﺘﺮى أﺣﻤﺪ ﺳﯿﺎرﺗﺎن‬

2. ‫ﺗﺰوج اﻟﺪﻛﺘﻮر ﻧﻮر زوﺟﺘﺎن‬

The examples above can be translated as:

1.
‘Ahmed bought two cars.’
2.
‘Dr Nour married two wives.’

As can be seen in these examples, the modi ier ‘two’ was added to the
TL to express the meaning of duality. In this regard, Arabic makes
extensive use of af ixations to change the category of any lemma. The
same lemma can have different meanings based on a minor change in
af ixation. See the following examples by way of illustration:

Example

Uploaded by S. M. Safi
Sold ‫ﺑﺎع‬
Bought ‫أﺑﺘﺎع‬
Selling ‫ﺑﯿﻮع‬
Accepting as a king or prince (homage) ‫ﺑﯿﻌﺔ‬
Bought (something) ‫ﻣﺒﺘﺎع‬
Paid homage ‫ﺑﺎﯾﻊ‬

As seen in these examples, the Arabic word has different meanings,


based on the same root or lemma. A translator, then, needs to pay
attention to such differences in meaning.

3.5 Syntactic Order: Foregrounding and


Backgrounding as a Problem in Translation
One prominent stylistic feature of the Arabic language is foregrounding
and backgrounding. However, this presents a basic challenge for a
translator that can be dif icult to resolve due to the style not being
universal. Put simply, each language has its own style that cannot be
followed in another language (Abdul-Raof, 2004). In Arabic, in general,
and in the Holy Quran in particular, style—including word ordering—
affects meaning (Abdelwali, 2007). The most important information is
typically foregrounded in Arabic, as the syntactic style in Arabic usually
allows this kind of foregrounding or backgrounding. Consider the
following example:

Example

‫ﻓﻲ ﺑﯿﺘﻨﺎ ﺣﺪﯾﻘﺔ ﺟﻤﯿﻠﺔ‬

‘We have a beautiful garden at our house.’

In this example, the ST expression ‫ ﻓﻲ ﺑﯿﺘﻨﺎ‬was foregrounded, it was


back-grounded in the TT due to the stylistic and syntactic differences
between English and Arabic. Consider the following example:

Example

Uploaded by S. M. Safi
‫ﺗﻼ وزﯾﺮ اﻟﺪﻓﺎع اﻟﺴﻮداﻧﻲ ﻋﻮض ﺑﻦ ﻋﻮف اﻟﺒﯿﺎن اﻷول ﻟﻠﺠﯿﺶ ﻣﻌﻠﻨﺎ اﻹﻃﺎﺣﺔ ﺑﺎﻟﻨﻈﺎم اﻟﺤﺎﻛﻢ‬
‫وﺗﻌﻄﯿﻞ اﻟﺪﺳﺘﻮر‬

‘The Sudanese minister of defence Awad bin Nouf announced in


the irst statement for the army the overthrowing of the ruling
regime in Sudan, and the suspension of the constitution.’

In this example, the ST foregrounds the most important information—


that is, the overthrowing of the ruling regime and the suspension of the
constitution; however, the translation did not follow the same syntactic
order. This can be attributed to the nature of the Arabic language, in
which the syntactic order is a part of its style; therefore, it affects its
meaning. The Holy Quran provides a useful source of examples of this
problem in translation. The following example, provided by Abdul-Raof
(2004), indicates how style affects meaning in the Holy Quran, and how
translation fails to convey the message of the Holy Quran:

Example

(20: 14) ‫ِﻢ اﻟﺼﻼ َة ﻟِ ِﺬ ْﻛ ِﺮي‬ َ ْ َ‫إﻧﻨِﻲ أَﻧَﺎ اﷲ ﻻ إِﻟَ َﻪ إِﻻ أَﻧَﺎ ﻓ‬
ِ ‫ﺎﻋ ُﺒ ْﺪ ﻧِﻲ َوأﻗ‬
‘Verily I am God; there is no god but I; therefore serve Me’.
(Arberry, 1982, p. 185)

This example shows how style was employed to convey the vividness
of the text of the Holy Quran, though it may seem to non-native
speakers of Arabic as redundant and replete with unnecessary
pronouns. For example, ‫ إﻧﻬﻨِﻲ‬/’innanii/, ‫’ أَﻧَﺎ‬/ana/, ‫ أَﻧَﺎ‬/’anaa/, all refer to
Allah Almighty. This Quranic style, as Abdul-Raof explained, serves two
propositions: the irst is related to Allah Almighty and His existence;
the second is about Allah Almighty’s Oneness. In addition, the use of ‫ف‬
‘fa’ indicates immediate action without hesitation. These entire
stylistic features in the Quranic ayah are not conveyed in translation.
Another example that indicates the failure of translation to keep the
same syntactic order as that of the ST is provided by Abdul-Raof
(2004):

Example
َ َ َ َ َ ْ ‫ُﱡ‬

Uploaded by S. M. Safi
(‫ﻟﻮ أن ﺑَ ْﯿﻨَ َﻬﺎ َوﺑَ ْﯿﻨَ ُﻪ أﻣﺪا‬
ْ ‫ِﻠﺖ ِﻣ ْﻦ ُﺳﻮ ء ﺗَ َﻮ ﱡد‬ ْ ‫ﯾَ ْﻮ َم ﺗَ ِﺠ ُﺪ ُﻛ ﱡﻞ ﻧَ ْﻔﺲ َﻣﺎ َﻋﻤ‬
ْ ‫ِﻠﺖ ِﻣ ْﻦ َﺧ ْﯿﺮ ُﻣ ْﺤﻀﺮا َو َﻣﺎ َﻋﻤ‬
3:30) ‫وف ﺑِﺎﻟْ ِﻌﺒَﺎ ِد‬ ‫ﺑَ ِﻌﯿﺪا َوﯾُ َﺤ ﱢﺬ ُر ُﻛ ُﻢ اﷲ ﻧَ ْﻔ َﺴ ُﻪ َو ﱠ‬
ٌ ‫اﷲِ َر ُء‬

‘On the Day when every soul will ind itself confronted with all
that it has done of good and all that it has done of evil, (every soul)
will long that there might be a mighty space of distance between it
and that (evil). God warns you of Himself. And God is full of pity for
(His) bondmen’. (Pickthall, 2001, p. 35)

In this example, the Quranic word ‫ ﻣﺤﻀﺮا‬goes after the word ‫;ﺧﯿﺮ‬
however, the Pickthall’s translation failed to preserve the same
syntactic order, resulting in semantic loss. The buffer word ‫ﻀﺮ‬ َ ‫ُﻣ ْﺤ‬
/muhdaran/ [be confronted with] serves to separate the two clauses ‫َﻣﺎ‬
‫ َﻋﻤِﻠَ ْﺖ ِﻣ ْﻦ َﺧ ْﯿﺮ‬/maa ‘amilat min khayrin /and ‫ ُﺳﻮء ِﻣ ْﻦ َﻋﻤِﻠَ ْﺖ َﻣﺎ‬/maa ‘amilat
min suu’in/ (Abdul Raof, 2004).
Sometimes, failure to preserve the syntactic order of the ST can
cause ambiguity. Sadiq (2008) gives an example of such ambiguity in
translating the following ayah from Surah al-Dukhan:

Example

(44:41) ‫ون‬ َ ‫ﯾَ ْﻮ َم َﻻ ﯾُ ْﻐﻨِﻲ َﻣ ْﻮل ى َﻋ ْﻦ َﻣ ْﻮل ى َﺷ ْﯿﺊ ا َو َﻻ ُﻫ ْﻢ ﯾُ ْﻨ‬


َ ‫ﺼ ُﺮ‬

‘The Day a patronizer will not avail any patronized thing’. (Ghali,
2005)

Sadiq (2008) argues that translating /shayan/‫ َﺷ ْﯿﺊ ا‬as ‘any patronized
thing’ is literal and syntactically vague; as the word ‘patronized’ in the
translation describes the ‘thing’, not the person. This translation is
unclear because translating ‫ ﺷﯿﺌﺎ‬strikingly literally as ‘thing’ created a
kind of ambiguity. Another example of loss in syntactic order (and thus
foregrounding and backgrounding) is given by Abdul-Raof in ayah 67 in
Surah Taha, which reads:

Example

َ ‫ﻓَﺄَ ْو َﺟﺲَﻓِﻲ ﻧَ ْﻔ ِﺴ ِﻪ ِﺧﯿﻔَﺔ ﱡﻣ‬


(20:6) ‫ﻮﺳﻰ‬

Uploaded by S. M. Safi
‘So Moses conceived is his mind a (sort of) fear’. (Ali, 1968, p.
209)

In the Quranic ST, the subject, which is prophet Musa (PBUH), is


backgrounded; however, in the TT it was foregrounded to follow the
English syntactic pattern. This surly created a kind of loss in meaning.
One more example of foregrounding and backgrounding, and how they
are lost in translation is provided by Abdul-Raof (2004) as follows:

Example

(6:100) ‫و ﺟﻌﻠﻮا ﷲَّ ﺷﺮﻛﺎء اﻟﺠﻦ‬

‘Yet they make the Jinns equals with God’. (Ali, 1968, p. 86)

In the Quranic text, the object (i.e. ‫ )اﻟﺠﻦ‬is backgrounded, and is taken
from its post-verbal position, whereas, ‫ﻟﻞ‬ ‫ ﱠ‬is foregrounded instead of in
its initial position. Thus, backgrounding and foregrounding serve
different communicative purposes, such as disapproving of what the
unbelievers say, bringing to the attention of the reader the notion of
calumny that the unbelievers attribute to God, condemning the
association of others with Allah Almighty, and keeping the supreme
status of Allah Almighty as Creator by foregrounding ‫ ﷲ‬/li-llahi/, and
showing the ordinary status of the Jinns who are themselves created by
Allah Almighty (Al Qurtubi, 2004, as cited in Abdul-Raof, 2004). None of
these purposes was communicated in the TT, as the translation could
not keep the same syntactic order as the authentic text due to the
linguistic limits of the English language, or perhaps because the
translator did not realize the communicative function of foregrounding
and backgrounding in the Quranic text.

3.6 Shifting (Iltifat) as a Problem in


Translation
Shifting or reference switching is a common feature of certain Arabic
genres, such as poetry, and the Holy Quran. One of the prevalent
stylistic features in the Holy Quran is the use of the grammatical shift

Uploaded by S. M. Safi
from one personal pronoun to another, and from one tense to another
(Abdul-Raof, 2004; Abdel Haleem, 2005). This is one of the peculiarities
of the rhetoric of the Arabic language; called ‘iltifat’, it serves various
purposes. Consider the following example:

Example

(2: 112) ‫ﺑَﻠَﻰ َﻣ ْﻦ أَ ْﺳﻠَ َﻢ َو ْﺟ َﻬ ُﻪ ِه ﱠﷲ َو ُﻫ َﻮ ُﻣ ْﺤ ِﺴ ٌﻦ ﻓَﻠَ ُﻪ أَ ْﺟ ُﺮ ُه ِﻋ ْﻨ َﺪ َرﺑﱢ ِﻪ َو َﻻ َﺧ ْﻮ ٌف َﻋﻠَْﯿ ِﻬ ْﻢ َو َﻻ ُﻫ ْﻢ ﯾَ ْﺤ َﺰ‬


َ ُ‫ﻧ‬
‫ﻮن‬

‘In fact, any who direct themselves wholly to God and do good
will have their reward with their Lord: no fear for them, nor will
they grieve’. (Abdel Haleem, 2004)

In this ayah, the implicit singular third person pronoun was used;
however, it shifted to the implicit plural third person pronoun by the
end of the same ayah, which the translator failed to convey. He
rendered both of the two ST Quranic pronouns as plural in the TT. The
translator could have rendered it faithfully as follows:

Example

In fact, he who directs themselves wholly to God and do good will


have their reward with their Lord: no fear for them, nor will they
grieve.

The success of such translations depends on the approach adopted by


the translator. Abdel Haleem seems to have adopted a communicative
approach, while in my own translation above I adopted a semantic
translation.
Another example of shift from the third person singular into the
plural adjective, which in Arabic follows the person or thing it
describes, is the following ayah from Surah al-Talaq (Abdul-Raof,
2004):

Example

َ ْ ْ ‫ﱠ‬

Uploaded by S. M. Safi
(65:11) ‫ِﯿﻬﺎ‬ َ ‫ﺎر َﺧﺎﻟِﺪ‬
َ ‫ِﯾﻦ ﻓ‬ َ ‫ﺻﺎﻟﺤﺎ ﯾُ ْﺪ ِﺧﻠْ ُﻪ َﺟﻨﻬﺎ ت ﺗَ ْﺠ ِﺮي ِﻣﻦ ﺗَ ْﺤﺘ‬
ُ ‫ِﻬﺎ اﻷ ْﻧ َﻬ‬ َ ‫ﺎﷲ َوﯾَ ْﻌ َﻤ ْﻞ‬
‫َو َﻣﻦ ﯾُ ْﺆ ِﻣﻦ ِﺑ ﱠ‬
‫ﺣﺴ َﻦ اﷲﱠُ ﻟَ ُﻪ ِر ْز ﻗﺎ‬
ْ َ‫ﻗﺪ أ‬
ْ ‫أَﺑﺪا‬

‘God will show anyone who believes in Him (God) and acts
honorably into gardens through which rivers low, to live there
forever. What a handsome provision God has granted him!’ (Irving,
1988)

In this ayah, there is a shift in the original that was not followed in the
translation; however, this did not affect the transmission of the ST
meaning. The following example is from the poetry of Amr bin
Kalthoum:

Example

‫ﻗﻔﻲ ﻗﺒﻞ اﻟﺘﻔﺮق ﯾﺎ ﻇﻐﯿﻨﺎ‬

‫ﻧﺨﺒﺮك اﻟﯿﻘﯿﻦ و ﺗﺨﺒﺮﯾﻨﺎ‬

‘Stop before separation you traveller

We tell you in certainty and you tell us’ (Author’s translation).

In this example, a shift occurred from the imperative mood (stop) to


the aorist (tell you) and (you tell us).
4.
Passivization
The frequency of use of the passive in Arabic and in English is
different. Arabic makes greater use of the active voice than the passive
voice (Al-Najjar, 1984). By contrast, English employs passivization in
many situations. Al-Najjar (1984) argues that English agentive
passives can be translated into Arabic as either agentive passive or
active voice. By way of illustration, consider the following example:

Example

Uploaded by S. M. Safi
‘“The subsidization on petrol will be lifted up next June”, the prime
minister said.’

1.
‫ ﻫﺬاﻣﺎ ﺻﺮح ﺑﻪ رﺋﯿﺲ اﻟﻮزراء‬،‫ﺳﯿﺮﻓﻊ اﻟﺪﻋﻢ ﻋﻦ اﻟﺒﺘﺮول ﯾﻮﻧﯿﻪ اﻟﻘﺎدم‬
2.
‫ﺻﺮح رﺋﯿﺲ اﻟﻮزراء ﺑﺄﻧﻪ ﺳﯿﺘﻢ رﻓﻊ اﻟﺪﻋﻢ ﯾﻮﻧﯿﻪ اﻟﻘﺎدم‬

As seen in these examples, the English ST could be translated into


either agentive passive (example 1) or active voice (example 2).
However, El-Yasin argues that it should be translated into Arabic topic-
comment structures, as shown in the following example:

Example

‘President as-Sadat was assassinated by El-Islambouly.’

‫اﻟﺮﺋﯿﺲ اﻟﺴﺎدات أﻏﺘﺎﻟﻪ اﻷﺳﻼﻣﺒﻮﻟﻲ‬

However, I do not think the TT sounds natural. I would translate it as:

Example

‫أﻏﺘﺎل اﻷﺳﻼﻣﺒﻮﻟﻲ اﻟﺮﺋﯿﺲ اﻟﺴﺎدات‬

Translating the English passive into Arabic should be based on the


intuition of the translator. They may decide to translate it into active
voice, which is mostly the case, or into passive voice. In contrast, when
translating from Arabic into English, the passive voice is mostly
maintained in the translation.

References
Abdel Haleem, M. A. S. A. (2004). The Qur’an. Oxford: Oxford University Press Inc.

Abdel Haleem, M. A. S. A. (2005). The Qur’an. Oxford: Oxford University Press.

Abdelwali, M. (2007). The loss in the translation of the Qur’an [Electronic version]. The
Translation Journal, 11(2), 120–125. Retrieved from http://www.accurapid.com/journal/

Uploaded by S. M. Safi
40quran.htm.

Abdul-Raof, H. (2004). The Quran: Limits of translatability. In S. Faiq (Ed.), Cultural encounters in
translation from Arabic. Frankfurt Lodge: Multilingual Matters.

Al-Azab, A., & Al-Misned, A. (2012). Pragmatic losses of Qur’an translation: A linguistic approach.
English Language and Literature Studies, 2(3), 42–49. https://doi.org/10.5539/ells.v2n3p42.

Ali, A., Brakhw, M. A., Bin Nordin, M. Z. F., & ShaikIsmail, S. F. (2012). Some linguistic dif iculties in
translating the Holy Quran from Arabic into English. International Journal of Social Science and
Humanity, 2(6), 588–590. https://doi.org/10.7763/IJSSH.2012.V2.178.

Ali, A. Y. (1968/2006). The Holy Qur’an, text, translation and commentary (Trans.). Beirut,
Lebanon: Dar Al Arabia.

Al-Najjar, M. F. (1984). Translation as a correlative of meaning (Unpublished PhD thesis).


Bloomington: Indiana University.

Al-Qurtubi, M. S. (2004). Al JamAA liahkam al Qur’an (Tafsir Al Qurtubi). Cairo, Egypt: Dar Al-Fikr.

Al-Zamakhshari, A. A.-Q. (2000). Al Kashshaf AAn haqaiq ghawamed attanzeel [The revealer of
facts of obscure revelations]. Cairo, Egypt: Arabic Publishing House.

Arberry, A. (1982). The Koran interpreted. Oxford: Oxford University Press.

Baker, M. (1992/2011). In other words: A coursebook on translation (2nd ed.). London and New
York: Routledge.

Collins COBUILD English Grammar. (2005). Glasgow: HarperCollins.

Cruse, D. A. (1997). Lexical semantics. Cambridge: Cambridge University Press.

Ghali, M. (2005). Towards understanding the ever–glorious Qur’an. Cairo: Dar An–Nashr Liljami.

Ghazala, H. (2008). Translation as problems and solution. Beirut: Dar El-Ilm Lilmalayin.

Hervey, S., & Higgins, I. (1992). Thinking translation: A course in translation method—French-
English. New York: Routledge (Taylor & Francis).

Irving, T. (1988). The Noble Qur’an. Cedar Rapids: Iowa Publisher.

Khan, M. M., & Al-Hilali, T. U. (1996). Interpretation of the meanings of the Noble Qur’an. Riyadh:
Maktaba Dar-us-Salam.

Newmark, P. (1981). Approaches to translation. Oxford: Pergamon.

Nida, E., & Taber, C. (1982). The theory and practice of translation (2nd ed.). Leiden: E.J. Brill.

Pickthall, M. (2001). The meaning of the glorious Qur’an. Kuala Lumpur: Islamic Book Trust.

Sadiq, S. (2008). Some semantic, stylistic and cultural problems of translation with special
reference to translating the glorious Qur’ân. Sayyab Translation Journal, 1(1), 37–59.

Uploaded by S. M. Safi
© The Author(s) 2020
N. Abdelaal, Translation between English and Arabic
https://doi.org/10.1007/978-3-030-34332-3_4

4. Lexical and Semantic Problems in


Translation
Noureldin Abdelaal1
(1) University of Nizwa, Nizwa, Oman

Noureldin Abdelaal

Overview This chapter presents the semantic and lexical


problems in translation between Arabic and English, and how to
deal with them. It gives examples of such problems in translation
and how to solve them.
The chapter covers the following topics:
1.
Lexical gaps at the semantic ield level (lack of equivalence
problem)
2.
Improper selection of vocabulary
3.
Lexical ambiguity: polysemy and homonymy
4.
Synonymy
5.
Problems in translation of rhetorical devices.

Newmark (1981) maintains that any lexical item can be viewed in


three different ways: dictionary items—types of senses (e.g. technical,

Uploaded by S. M. Safi
igurative, colloquial); the four degrees of frequency (e.g. primary,
collocational); and the core and the peripheral meanings. These
meanings may create problems in a translation if a translator cannot
differentiate between these meanings. There are various
manifestations of lexical and semantic problems in translation. These
problems are likely to create syntactic and semantic loss in translation
between any two linguistic codes in general, and between Arabic and
English in particular. Arabic is far richer than English and this poses
dif iculties in translations (Daryabadi, 2007). One of the major lexical
problems is translating metaphorical meaning as a non-metaphorical,
or vice versa. Another problem is translating synonyms, near-
synonyms, polysemous items, collocations and homonyms. Other
problems include problems of equivalence, lexical gaps, and denotative
and connotative meanings.
English and Arabic express reality in very different ways. Thus, in
translation between these two languages, losses occur and problems
arise (Abdul-Raof, 2005). One of these problems is the lexical gaps.
Lexical gaps are a kind of discrepancy that may occur due to syntactic
divergences, cultural differences between the SL and the TL,
lexicalization differences, divergences in connotations, or differences
in the denotative meanings (Bentivogli & Pianta, 2000). Lexical gaps
can also occur at different levels, including the semantic and the
morphological (Abdul-Raof, 2005). These problems are unpacked in
the following sections.

4.1 Lexical Gaps at the Semantic Field Level


(Lack of Equivalent Problem)
The term ‘semantic ield’ refers to lexical concepts that share semantic
properties, or simply to the relatedness of meaning among the word
class (Lobner, 2002). Following this de inition, lexical gaps occur when
one lexical item is missing in the semantic ield structure (Lyons,
1977), which occurs due to differences between languages (Darwish,
2010; Lyons, 1977). For example, the Arabic verb ‫ ﯾﺤﺞ‬/ya’hujj/ does not
have an equivalent in English; in other words, it is not lexicalized in
English. However, a translator needs to ind the appropriate strategy to

Uploaded by S. M. Safi
translate it. It may be resolved by inding near-equivalents or
undertaking componential analysis of the ST word. Let us consider the
word ‫ﯾﺤﺞ‬, which can be translated as:

Example
1.
Do/perform/act pilgrimage;
2.
Do/perform/act haj;
3.
Do/perform/act haj (pilgrimage).

These three options can provide proper strategies for translating the
ST word; it is the translator’s decision to select the proper strategy
based on their translation beliefs and ideologies, and based on the
commission given to them.
Baker (1992) discussed the lack of lexicalization as one of the major
problems in translation between Arabic and English. An example that
highlights this problem was given by Conner (1983), who introduced
the example of the semantic ield of temperature, which is represented
in English by four words: cold, cool, hot, and warm. By contrast, in
Arabic the same semantic ield of temperature is represented by three
words: ‫ داﻓﺊ‬،‫ ﺑﺎرد‬،‫ﺣﺎر‬. There is no lexical item that matches the English
item of ‘cool’. As a result of this lexical gap, the two words ‘cool’ and ‫ﺑﺎرد‬
may be translated as synonyms, though they are antonyms (Abdul-
Raof, 2005). However, this may be a problem for a novice translator; an
expert translator would not fall into this trap.
Another example of the lack of lexicalization is the Arabic word ‫ﺟﻬﺎد‬
/jihad/, which is not represented in the English language. However, it
can be rendered as:
1.
Jihad;
2.
Striving;
3.
Striving (holy war).

Uploaded by S. M. Safi
These three translations can be sought by a translator. He may also
opt to transliterate the ST word (example 1), accompanied by a
paraphrase.
Lack of lexicalization occurs due to the lack of lexical equivalence,
which occurs due to differences between languages (Benfoughal,
2010). Each language has its own peculiarities in terms of vocabulary,
grammar, or style. Some words are lexicalized in one language, but not
in the other. This applies to all languages. Take, for example, the
adjectival word ‘standard’; although this is a very common word in
English, it does not have an equivalent item in Arabic (Baker,
1992/2011), though translators tend to translate it as ‫ﻣﻌﯿﺎر او ﻣﻘﯿﺎس‬.
Although Baker considers the word ‘standard’ to be a word that does
not have an equivalent in Arabic, adopting Vinay and Darbelnet’s
notion of equivalence as being any lexical terms that are regarded as
equivalents in a bilingual dictionary, we should then think that ‫ﻣﻌﯿﺎر او‬
‫ ﻣﻘﯿﺎس‬are proper translations of the English word. Moreover, the
purpose of the ST word is conveyed in Arabic if it is translated as ‫ﻣﻌﯿﺎر او‬
‫ﻣﻘﯿﺎس‬.
Darwish (2010) posits that the difference in a denotative meaning
between an SL and a TL is another cause for lexical gaps in translation;
for example, the Arabic word ‫ﺻﻮم‬, which is always rendered into
English as ‘fasting’, has different denotative meanings according to
culture. ‘Fasting’ in Christianity is completely different from ‘fasting’ in
Islam. However, I think that rendering the Arabic word ‫ ﺻﻮم‬as ‘fasting’
is a good translation strategy because the skopos of the translation
conveys the primary meaning, which is conveyed. The other shades of
meaning are of only concern to people such as specialists, researchers,
or newly-converted Muslims. To whomsoever it may be of concern,
they may read books on the topic to further their education on the
culturally bound word or expression. Thus, the Arabic word ‫ ﺻﻮم‬can be
rendered as ‘fasting’ with or without glossing. It is the translator’s
decision whether to provide an explanation of the ST word, or to leave
it as it is in the TL.
In a similar vein, Nugroho (1999) underscored the importance of
understanding the components of meaning in the SL so as to be able to
render them accurately to a TL. A denotative meaning may undergo a

Uploaded by S. M. Safi
syntactic marking (e.g. the difference between ‘he saw a cloud’ and ‘the
quarrel will cloud the issue’), or a semiotic marking (i.e. the
interrelationships among words). For example, the difference between
‘he runs a company’ and ‘his nose is running’ is a difference that occurs
due to the different subjects and their meanings (Nugroho, 1999).
Similarly, Newmark (1988) mentioned that componential analysis is a
useful tool in understanding the differences between synonyms.
Analysing an ST word into its components can be a useful tool that
helps resolving the problem of deciding on the most appropriate
equivalent. Consider the following examples (extracted from reverso):

Example 1.‫وﺗﻘﺘﺮح اﻟﺴﯿﺎﺳﺔ آﺟﺎﻻً وأﺣﻜﺎﻣﺎً ﻟﺘﻨﻔﯿﺬ ﺗﺸﺮﯾﻊ ﻣﻜﺎﻓﺤﺔ اﻟﺘﺤﺮش اﻟﺠﻨﺴﻲ‬.
‘The Policy intends to propose terms and provisions for the
enactment of sexual harassment legislation.’
2.‫ ﺑﻤﺎ ﯾﺸﻤﻞ اﻏﺘﺼﺎب‬،‫وﺗﻐﻄﻲ اﻟﻤﻮاد اﻟﻤﺬﻛﻮرة ﺣﺎﻻت اﻟﺘﺤﺮش اﻟﺠﻨﺴﻲ واﻻﻏﺘﺼﺎب‬
‫اﻟﻘﺼﺮ‬.
ّ
‘The above articles cover cases of molestation and rape, including
statutory rape.’

As seen in these two examples, the ST word ‫ اﻟﺘﺤﺮش اﻟﺠﻨﺴﻲ‬was translated,


irst, as ‘sexual harassment’, but then as ‘molestation’ in the second
instance. Let us analyse the ST words and TT words to see the reason
behind providing two translations (or more) for the same ST word.
Molestation
+ harassing
+ abuse
+ against children
+ against women
Sexual harassment
+ Unwelcome remarks
(+−) abuse
(+−) against children
(+−) against women
As can be seen, the two TT words are similar in meaning; however,
sexual harassment is more neutral and less abusive than molestation.
However, the ST word ‫ اﻟﺘﺤﺮش اﻟﺠﻨﺴﻲ‬can imply both molestation and
sexual harassment, depending on the context of its use. This justi ies

Uploaded by S. M. Safi
why example 1 was translated as ‘sexual harassment’; it is about
legislation that should ban the less abusive, dangerous and serious
action lest it leads to the more abuse. In example 2, the word
‘molestation’ was used because it was referring to rape and to a more
advanced act of sexual harassment beyond simply unwelcome
remarks.
Another facet of the problem of the lack of lexicalization arises
when a speci ic concept is expressed by one lexical unit in an SL, while
the same concept is expressed by a free combination of words in the
TL (Bentivogli & Pianta, 2000; Darwish, 2010). For example, the
English kinship relation of ‘cousin’ is represented by eight words in
Arabic: ‫ ﺑﻨﺖ اﻟﻌﻤﺔ‬،‫ أﺑﻦ اﻟﻌﻤﺔ‬،‫ ﺑﻨﺖ اﻟﺨﺎﻟﺔ‬،‫ أﺑﻦ اﻟﺨﺎﻟﺔ‬،‫ ﺑﻨﺖ اﻟﺨﺎل‬،‫ أﺑﻦ اﻟﺨﺎل‬،‫ ﺑﻨﺖ اﻟﻌﻢ‬،‫أﺑﻦ اﻟﻌﻢ‬.
Thus, one lexical item in English is represented by eight counterparts
in Arabic, creating a lexical gap that results in problems for translators.
A translator, thus, needs to know the actual and contextual usage of the
ST word, which cannot be divorced from its meaning, so as to avoid
making errors in translation. In this case, one English lexical item will
be translated into a two-word expression to convey the accurate
meaning of the ST word. Darwish (2010) refereed to the same problem,
viz., when the SL and TL languages taxonomies are different; for
example, eclipse in English has two Arabic counterparts in relation to
the sun and the moon. One last cause of lexical gaps, as mentioned by
Darwish, is cultural gaps which cause lack of lexicalization. All these
causes, concisely, can be described as lexicalization differences
between an SL and a TL. Examples of words that do not have lexical
equivalents in English may be the following words:

Example
‫( دم ﺣﻔﯿﻒ‬lit. light-blooded): This is used to describe a person who is
liked by others because of his nice nature.
‫( دم ﺛﻘﯿﻞ‬lit. heavy-blooded): This describes a person whom people do
not ind likable due to his nature.
‫ﻧﻌﯿﻤﺎ‬: This is an expression said to a person after he has taken a
shower or had his beard shaved.
‫ﻋﻘﺒﺎﻟﻚ‬: This means ‘I wish you the same’. It is mostly said to as a
reply to a person who congratulates you on a happy occasion, such

Uploaded by S. M. Safi
as marriage, having a new baby and so on.
‫( اﻟﺒﻘﺎء ﷲ‬lit. Only Allah ever exists): This is used as a condolence to
someone over the death of a family member or relative.
‫( أﺣﺴﻦ اﷲ ﻋﺰاءﻛﻢ‬lit. May Allah reward you): This is also used as a
condolence to someone over the death of a family member or
relative.
‫واﺳﻄﺔ‬: This refers to someone who can help you expedite a process
or do something unlawful, or even get a job. It is comprehensive, as
it can refer to any way that someone can help another person
achieve anything, even buying tickets.
‫ﺳﻬﺮة‬: This means staying up late for the purposes of pleasure,
perhaps at a café.
‫إن ﺷﺎء اﷲ‬: This means ‘if Allah wills’. It is used in the Arabic context as
a inal con irmation that you will do something in the future. It is an
Islamic word.
‫ﻋﻘﯿﻘﺔ‬: This refers to the Islamic habit of slaughtering a sheep if you
have a newborn daughter, or two sheep if you have a newborn son.
‫ﻋﺸﺎء اﻟﻌﺮﯾﺲ‬: This refers to an excess of food cooked by the family of
the bride and sent to the groom’s house on the wedding day.
‫اﻟﺨﻠﻮة‬: This is a religious reference to the condition of the groom
being alone with the bride to have sex for the irst time as husband
and wife.
‫ﻟﯿﻠﺔ اﻟﺤﻨﺎء‬: This is the night that precedes the wedding day. The bride
and groom, and some of their family members, celebrate with
singing, dancing and the putting on of henna.
‫ﻣﻠﻚ ﯾﻤﯿﻦ‬: This refers to a woman who can act as a wife if she is taken
as a captive in a holy war between Muslims and non-Muslims.
‫اﻟﻌﯿﺪﯾﺔ‬: This is money given to children on Eid al Fitr and Eid Al Adha
days. It is usually given by parents, uncles and relatives.

As can be seen, there are lexical gaps between English and Arabic.
However, the fact that these words do not have equivalents does not
mean that they cannot be translated. They can be translated using
different strategies, such as borrowing, paraphrasing, paraphrasing
with glossing, transliteration or periphrastic translation.

Uploaded by S. M. Safi
The Holy Quran is rich with examples that show lexical gaps in the
Quranic translation. A case in point is the difference between
ّ
‫ﻧﺰل‬/nazzala/ (ayah 3) and ‫أﻧﺰل‬/anzala/ (ayah 4) in Surah al-Imran that
cannot be conveyed in translation due to the lexical or morphological
gap. The irst word /nazzala/ re lects a piecemeal revelation, while the
latter word, /anzala/, re lects a single event of complete revelation.
However, Yusuf Ali rendered the two words as one word in his
translation (Abdul-Raof, 2004). The Quranic word ‫ وﯾﻞ‬in Surah al-
Humaza is another example of a lexical gap (Al-Ghazali, 2010).
Consider the following ayah and its translation:

Example 104:1) ‫)وﯾﻞ ﻟﻜﻞ ﻫﻤﺰ ة ﻟﻤﺰة‬


‘Woe to every (kind of) scandal-monger and backbiter’. (Ali, 2006,
p. 462)
‘Woe to every slanderer, defamer’. (Shakir, 1999)

As can be seen, the Quranic ST word is in the nominative case, while


the translations by Ali and Shakir rendered it in the subjunctive.
Although this affects the meaning due to the nominal case in Arabic
generally indicating continuity, and, in this context in particular, refers
to ongoing torture and punishment (Al-Ghazali, 2010), it is still an
acceptable translation, as it conveys the denotative meaning of the ST.
Another example that indicates the lexical gap between any two
languages is two Quranic words in Surah al-Kahf: ‫ أﺳﻄﺎع‬/istaAAa/ and
‫ أﺳﺘﻄﺎع‬/istataAAa/. Both words were translated interchangeably as
‘could’ or ‘was able to’ by Sale, Muhammad Ali, Pickthall, Rodwell and
others. The two words are not identical in meaning; there is delicate
difference in meaning between them. The Quranic word ‫أﺳﻄﺎع‬
/istaAAa/ is only used for relatively easy actions, such as climbing a
hill, whereas ‫ أﺳﺘﻄﺎع‬/istataAAa/ is used for a more dif icult task, such as
boring a tunnel through a hill (Khalifa, 1989), A further example of
lexical gaps is the two Arabic words /ridwanun/‫ رﺿﻮان‬and /rida/‫;رﺿﺎ‬
these two words are not complete synonyms, as ‫ رﺿﻮان‬is more
pregnant with meaning than ‫رﺿﺎ‬, as it means being completely pleased
with believers. However, English lacks the ability to show such nuances
between these near-synonyms. The Quranic words /hayawan/‫ﺣﯿﻮان‬

Uploaded by S. M. Safi
and /hayatun/‫ ﺣﯿﺎة‬provide another example of lexical gaps in
translation. The Quranic word ‫ ﺣﯿﻮان‬was mentioned once in the Holy
Quran in the context of the virtue of the Hereafter (i.e. Jannah) over the
earthly life. ‫ ﺣﯿﻮان‬is the real and complete life, which belongs only to the
everlasting life in Jannah (Al-Qurtubi, 2004). This may be due to the
fact that adding certain letters in the Arabic language, such as ‫ﺣﯿﻮان‬,
adds extra meaning to a word. Translating these words can be achieved
through the use of modi iers or intensi iers. For example, the word
‫ ﺣﯿﻮان‬can be translated as ‘real life’, ‘true life’, or ‘complete life’. A
translator should not look for a one-to-one equivalent when translating
culturally bound terms: they should seek other translation strategies
that address the issue raised in the translation process.
Another example of lexical gaps is the Quranic verb ‫أﺳﺮى‬/asraa/,
which cannot be rendered into an equivalent lexeme in English (Abdul-
Raof, 2004) because it is a semantically complex verb. Semantic
complexity, as identi ied by Baker (1992), is one of the non-
equivalence problems in translation between Arabic and English. Other
examples may include words such as /tayammamoo/‫‘( ﺗﯿﻤﻤﻮا‬take some
clean sand and wipe your face and hands with it’),/yastarikhoona/
‫‘( ﯾﺼﻄﺮﺧﻮن‬cry out loud’), and /yatatahhar/‫‘( ﯾﺘﻄ ّﻬﺮ‬to stay chaste’).
These words are used in the exaggerated form in Arabic Quranic
language. However, English does not have such a feature. A working
example could be the Quranic verb ‫ ﯾﺼﻄﺮﺧﻮن‬/yastarikhoona/, which
was mentioned in the Holy Quran in the context of telling of the
torment disbelievers will suffer on the Day of Judgement. They do not
simply ‘cry’. They howl with sorrow from the depth of their hearts, to
ask Almighty Allah, as they think, to give them another chance to go
back to earthly life to do good deeds (Ibn Ashour, 1984).
Another example that indicates lexical gaps in translation between
an ST and a TT is the following example from Surah al-Baqarah (Abdul-
Raof, 2004):

Example (2:2) ‫ِﯿﻦ‬ َ ‫ﺎب ﻻَ َر ْﯾ َﺐ ﻓِﯿ ِﻪ ﻫﺪى ﻟﱢﻠْ ُﻤﺘﻘ‬


ُ َ‫ِﻚ اﻟْ ِﻜﺘ‬
َ ‫َذﻟ‬
‘This is the Book; in it is guidance sure, without doubt, for those
who fear God’. (2:2) (Ali, 1968, p. 8)

ْ‫ﱢ‬

Uploaded by S. M. Safi
َ ‫ ﻟﱢﻠْ ُﻤﺘﻘ‬was translated as ‘for those who fear God’, which
In this example, ‫ِﯿﻦ‬
is redundant and inaccurate because the Quranic word has sensitive
overtones that encompass performing all kinds of good deeds ordained
by Allah Almighty and avoiding everything Allah Almighty forbade
(Abdul-Raof, 2004). Abdul-Raof commends the solution Khan and Hilali
adopted, as they gave a periphrastic translation after providing the
transliteration.
Similarly, in his study of the lexical gap in the translation of the
Quranic verb, ‫ ﻛﺎد‬/kada/, Al-Utbi (2011) signposted how the translation
failed to ind equivalents of the verb ‫ﻛﺎد‬, due to the lexical gap between
the SL and the TL. The translations investigated revealed several
changes in the word class of the Quranic verb to different word classes:
adverbs, verbal constructions and adjectives. Largely, the translations
of the Holy Quran are abundant with examples that show such lexical
gaps in translation.
In relation to the problems such lexical gaps cause in translation,
Abdul-Raof (2004) postulates that the only way to translate such
Quranic lexemes (i.e. those that suffer from problems of lexical gaps) is
to seek periphrastic translation. Abdul-Raof gives examples of Quranic
words, such as /‫ اﻟﻤﻮﻗﻮذة‬/mawqoothatu/, that can only be rendered
periphrastically as ‘any animal that receives a violent blow, is left to
die, and then eaten without being slaughtered according to Islamic law’.
Other examples provided by Abdul-Raof (2004) include /
‫ﻛﻈﯿﻢ‬/katheem/, /‫اﻟﺼﻤﺪ‬/aS-Samad/, ‫ﺗﯿﻤﻤﻮا‬/tayammamoo//, and many
other lexical items.

4.2 Improper Selection of Vocabulary


One of the problems that may occur in translation is the improper
selection of what seem to be equivalents of the ST words when they
may not be true equivalents. This may result from ambiguities in the
ST lexemes or syntax. In this regard, Newmark (1988) de ines
ambiguity as a word or a syntactic structure that has more than one
meaning, even in its context. ‘In its context’, here, indicates that a word
cannot be considered ambiguous without referring to context, as each
word may be ambiguous out of context. Newmark divides ambiguity

Uploaded by S. M. Safi
into seven types: grammatical, lexical, pragmatic, cultural, idiolectal,
referential and metaphorical.
1.
Grammatical ambiguity: This may occur when sentences have
been poorly written, or due to the effect of the use of grammatical
and functional words. Prepositions, phrasal verbs and pronouns
can also cause ambiguity.
2.
Lexical ambiguity: This is more common and more dif icult to
resolve than grammatical ambiguity. Many words have more than
one sense (i.e. polysemous), or have literal and metaphorical
meanings. Homonymy is also one of the causes of lexical ambiguity.
3.
Pragmatic ambiguity: This arises when the tone or emphasis in an
SL sentence is not clear.
4.
Cultural ambiguity: This arises if ‘the function or the substance of
a cultural feature changes at a point of time and the term remains
whilst the period background is not clear in the SL text’ (Newmark,
1988, p. 220). In addition, there are certain concepts that are ‘near-
internationalisms’; these concepts, however, may have different
uses in different languages. For example, the word ‘king’ in Saudi
Arabia does not have the same sense as that of a ‘king’ in the United
Kingdom.
5.
Idiolectal ambiguity: This arises from people perceiving words
differently from one another.
6.
Referential ambiguity: All types of ambiguity can be referential;
however, Newmark states that he means the ambiguous use of
proper names in an SL text.
7.
Metaphorical ambiguity: Most sentences can have metaphorical
meaning and literal meaning. However, typically only one speci ic
meaning is intended. For example, ‘kick the bucket’ can refer to the
literal meaning of the expression, or to its metaphorical meaning,
which is ‘to die’.

Uploaded by S. M. Safi
All these types of ambiguities can be referred to as linguistic
ambiguity. Linguistic ambiguity in translation can arise from
ambiguity in the ST. If a translator fails to identify and resolve such
ambiguity, or due to a lack of full understanding of the ST and the
context of its situation, a translator may create an ambiguous
translation. Among the various linguistic ambiguities mentioned by
Newmark, lexical ambiguity is the most problematic. Lexical ambiguity
can be divided into two main categories; one holds that words have
lexical ambiguity prior to their semantic occurrence in a text; the other
considers that lexical ambiguity is context dependent, which means
that it occurs due to the effect of the text (Simpson, 1981). Lexical
ambiguity can result from either homonymy or polysemy (Newmark,
1988). Examples of problems in translating polysemous and
homonymous words are discussed in the following sections.

4.3 Lexical Ambiguity: Polysemy and


Homonymy
Newmark (1988) argued that lexical ambiguity is one of the
ambiguities that cannot be easily resolved. Such ambiguity mainly
arises from polysemy, homonymy, or metaphorical vs. literal meanings.
Lexical ambiguity is very common in language, as a single string of
words may lead to more than one interpretation due to one of the
words having more than one meaning: polysemy (Klepousniotou,
2002; Simpson, 1981). However, polysemy can be confused with
homonymy, in which two words with the same spelling or
pronunciation have two different meanings. According to the
generative lexicon approach, homonymy ensues when discrete senses
are stored separately, whereas, in polysemy, only the elementary
meaning is stored in the lexical repertoire of the language user
(Klepousniotou, 2002; Klepousniotou & Baum, 2005). In other words,
polysemy refers to a multiplicity of meanings such as when one word is
used in different ields with different meanings (Geeraerts, 2010). A
case in point is the word ‫ ;ﻋﯿﻦ‬it has several meanings in Arabic, such as:
‫ اﻟﺼﻮاب ﻋﯿﻦ‬and ‫اﻟﺤﻘﯿﻘﺔ ﻋﯿﻦ‬, which mean ‘absolutely right’, and ‫اﻻﺑﺮة ﻋﯿﻦ‬,
which means the ‘eye of a needle’ or a ‘spy’ (Sadiq, 2008). Hence, these

Uploaded by S. M. Safi
words are polysemes because they have the same etymological root
(Sadiq, 2008). Such polysemy may create ambiguity for a reader.
However, I argue that these polysemous words can create problems for
a novice translator, or a translator who does not have suf icient
knowledge regarding the SL. Polysemous words are unlikely to cause
problems for an experienced translator. English and Arabic are replete
with examples of polysemy. Consider the following examples in English
and how they may be translated into Arabic:

Example

English Arabic
‘Ali has his mouth full of food.’ ‫ﻋﻠﻲ ﻓﻤﻪ ﻣﻤﺘﻠﺊ ﺑﺎﻟﻄﻌﺎم‬
‘Mary kissed John on his mouth.’ ‫ﻣﺎري ﻗﺒﻠﺖ ﺟﻮن ﻋﻠﻰ ﺷﻔﺘﯿﻪ‬
‘My mouth is sore.’ ‫ﻓﻤﻲ ﻣﻠﺘﻬﺐ‬
‘Watch your mouth.’ ‫اﻧﺘﺒﻪ ﻟﻜﻼﻣﻚ‬
‘I have three mouths to feed.’ ‫ﻟﺪي ﺛﻼث اﻓﻮاه أﻃﻌﻤﻬﺎ‬

As can be seen in these examples, the ST word ‘mouth’ is polysemous,


as it has multiple meanings. However, it seems that the ST word is
polysemous in the TL as well. In many cases, English and Arabic share
the polysemous nature of some words and phrases.
Consider the following examples:

Example

English Arabic
‘Ahmed is healthy.’ ‫أﺣﻤﺪ ﯾﺘﻤﺘﻊ ﺑﺼﺤﺔ ﺟﯿﺪة‬
‘Ahmed’s exercise regimen is healthy.’ ‫اﻟﺘﻄﺎم اﻟﺮﯾﺎﺿﻲ اﻟﺬي ﯾﺘﺒﻌﻪ أﺣﻤﺪ ﺟﯿﺪ‬
‘Ahmed’s complexion is healthy.’ ‫ﺑﺸﺮة أﺣﻤﺪ ﺗﺒﺪو ﻋﻠﯿﻬﺎ اﻟﻨﻀﺎرة‬

As can be seen in these examples, the word ‘healthy’ has several


meanings; therefore, it was translated differently in Arabic. In this case,
the word ‘healthy’ is not polysemous in Arabic, and that is why it was
translated into different lexemes. The co-text usually clari ies the
meaning for a translator.

Uploaded by S. M. Safi
The Holy Quran is rich with examples of polysemy, as it is one of its
linguistic features. For example, the word ‘‫ ’أﻣﺔ‬/ummah/ has nine
polysemic meanings in the Holy Quran. It can mean ‘a period of time’, as
in Surah Yusuf; or a ‘leader’ in a religious sense, ‘a person who leads
people to the right path’, as in Surah al-Nahl. A situation where
polysemy presents a problem when translating the Holy Quran is the
following example from Surah al-Baqra, ayah 187, which reads:

Example (2:187) ‫ﻫﻦ ﻟﺒﺎس ﻟﻜﻢ و أﻧﺘﻢ ﻟﺒﺎس ﻟﻬﻦ‬


‘They are your garments and ye are their garments.’ (2:178)
(Ali, 1968, p. 24)

As seen in Ali’s translation, the Quranic word ‫ ﻟﺒﺎس‬was translated


literally as ‘garment’ which according, to Ali et al. (2014), is inaccurate
because the Quranic word means that ‘A man can repose with his wife,
so they become as a cover to keep all the relations between them
secret in the home’. In fact, polysemy causes problems for any
translator of the Holy Quran. One way of resolving this could be by
reference to trusted exegetic books to decide which meaning would be
the most appropriate in this context. Alternatively, as suggested by
Ilyas (2013), intertextuality can provide a solution. However, in the
event that the meaning provided by exegeses is changeable,
transliteration associated with periphrastic translation could be the
solution. This may give depth to the translation. In addition, if a
translator opted for one meaning rather over another, he should make
it clear, in a footnote, that his translation is based on a speci ic exegetic
book.
Homonymy is another main cause of lexical ambiguity
(Klepousniotou, 2002). Homonymy refers to the sense relationship
that arises when two words have the same spelling but different
meanings. A common example of homonymy is the word ‘bank’, which
can refer to a bank as a inancial institution, or a bank of a river
(AlQinai, 2012; Geeraerts, 2010). The Holy Quran is rich with
homonyms that cause lexical ambiguity, thereby requiring a translator
to decode the homonymous meanings to render an accurate
translation. In their study of the Quranic word ‫( ﻓﺴﺎد‬fasad: ‘corruption’),
Rasekh et al. (2012) found that the homonymous nature of the Quranic

Uploaded by S. M. Safi
language causes ambiguity in translation. Rasekh et al. (2012)
identi ied two types of ambiguity in translation; conscious ambiguity
and unconscious ambiguity. Conscious ambiguity occurs when the ST
is ambiguous in itself. In this situation, a translator should retain the
intended ambiguity in the TT. Unconscious ambiguity can be
disambiguated, according to its situational or linguistic context, by
sacri icing the less important meaning. An example of
misunderstanding homonymous words can be found in relation to
ayah 40 from Surah Ash-Shura, as provided in Abedelrazq (2014):

Example (42:40) ‫ﻧﻪ َﻻ ﯾُ ِﺤ ﱡﺐ‬ُ ِ‫َو َﺟ َﺰا ُء َﺳﯿﱢﺌَﺔ َﺳﯿﱢﺌَ ٌﺔ ِﻣ ْﺜﻠُ َﻬﺎ ﻓَ َﻤ ْﻦ َﻋﻔَﺎ َوأَ ْﺻﻠَ َﺢ ﻓَﺄَ ْﺟ ُﺮ ُه َﻋﻠَﻰ اﷲ إ‬
‫اﻟﻈﺎﻟ ِِﻤﯿﻦ‬
‘The guerdon of an ill-deed is an ill the like thereof. But whosoever
pardoneth and amendeth, his wage is the affair of Allah. Lo! He loveth
not wrong-doers.’ (Pickthall, 2001, p. 198)
‘The recompense for an injury is an injury equal thereto (in
degree): but if a person forgives and makes reconciliation, his reward
is due from God: for (God) loveth not those who do wrong.’ (Ali, 1968,
p. 343)

In this example, the homonymous Quranic word ‫ َﺳﯿﱢﺌَﺔ‬was translated as


though it were the same word, even though the irst use of the word
refers to a ‘bad act’, and the second to ‘hostility and aggression’ Al
Tabari (Abedelrazq, 2014).
In his study on intra-textuality as a solution for translating
ambiguous terms in the Holy Quran, Ilyas (2013) noted that intra-
textuality could be a solution for translating problematic items in an ST
that have more than one parallel occurrence in a text. He mentioned,
for example, how translators rendered the Quranic word ‫ ﺳﺠﺮت‬in the
ayah to different lexemes that fall under the meaning of ‘boiling’,
though commentators of the Holy Quran gave different meanings of the
word. For example, Asyuti interpreted it as ‘burnt’; Al Tabari rendered it
as ‘ looded’; while Al-Razi used the two senses in one term (i.e. ‘burn
and lood’). Consider the following ayah and its translation:

Example (81:6) ‘‫ﺳﺠﺮت ﺳﻮرة اﻟﺘﻜﻮﯾﺮ‬ّ ‫’و إذا اﻟﺒﺤﺎر‬


‘when the seas shall be set boiling’. (Arberry, 1982, p. 357)

Uploaded by S. M. Safi
However, Ilyas (2013) argued that a solution could be provided by
referring to other ayahs to understand the meaning of the Quranic
word. For example, understanding of the previous ayah can be
achieved by referring to the following ayah from Surah Ghafer:
ْ
Example (72:40) ‫ون‬ ِ ‫ﯿﻢ ﺛُﻢ ﻓِﻲ‬
َ ‫اﻟﻨﺎر ﯾُ ْﺴ َﺠ ُﺮ‬ ِ ‫ﻓِﻲ اﻟ َﺤ ِﻤ‬
‘They are dragged through boiling waters; then they are thrust
into the Fire’. (40:72) (Pickthall, 2001, p. 193)

Employing intra-textuality by linking or referring to words in different


contexts in the Holy Quran helps in understanding and clarifying the
meaning. For example, ayah 72 of Surah Ghafer made it easier to
understand the verb in ayah 6 of Surah At-Takwir: Ayah 72 of Surah
Ghafer describes in idels who are fated to be burnt in ire. Thus,
Arberry’s translation is incorrect, as ‘boiling’ is not the accurate
meaning of the verb (Ilyas, 2013). This brings into consideration the
problem of avoidable losses that are created by translators due their
lack of reference to various exegetic books. Another lexical problem in
translation is synonymy.

4.4 Synonymy
Synonymy, a lexical relationship term that is used to refer to a
sameness of meaning (Lobner, 2002; Palmer, 1981), has been
recognized as one of the challenges in translating from Arabic to
English. This may be due to the fact that Arabic as a language is rich
with synonyms. Stanojević (2009) states that absolute synonyms are
unlikely to exist in a language and that it is predictable that translating
synonyms between two culturally and linguistically different languages
is problematic, such as is the case between Arabic and English.
Arabic is rich with synonyms. For example, Asyuti (2008) states
that there are 41 synonyms for the word ‫اﻟﺴﯿﻒ‬/al ssayf/ ‘(‘the sword’),
and 80 synonyms for the word ‫اﻟﻌﺴﻞ‬/al AAasal/ (‘honey’). The wealth of
synonyms in Arabic vocabularies in general sets up pitfalls for a
translator. A translator may use one synonym in lieu of another that is
more accurate. A further problematic type of synonymy is cognitive

Uploaded by S. M. Safi
synonymy. Cognitive synonymy refers to incomplete synonyms; for
example, ‘liberty’ and ‘freedom’ (Stanojević, 2009).
Shunnaq (1992) states that translating cognitive synonyms can be
misleading, due to the slight variations between synonyms. Cognitive
synonyms share the same referent, but they are connotatively
different. Hence, one of the proposed criteria is the intuition of a native
Arabic speaker who is better able to judge such nuances. For example,
the slight difference in meaning between ‫ ﯾﻐﺒﻂ‬/yaghbit/ and ‫ﯾﺤﺴﺪ‬
/yahsid/ is dif icult to identify without having intuitive and profound
knowledge of the differences between synonyms in Arabic. Lack of such
knowledge may cause some translators to render the two cognitive
synonyms (i.e. ‫ ﯾﺤﺴﺪ‬and ‫ )ﯾﻐﺒﻂ‬as ‘envy’, which is a complete divergence
from the true meaning (Shehab, 2009). Other examples provided by
Shehab include the nuances between /matar/‫ ﻣﻄﺮ‬and /ghayth/‫ﻏﯿﺚ‬,
which are rendered as ‘rain’. The two Quranic words are connotatively
different. Even though both words refer to ‘rain’, ‫ ﻣﻄﺮ‬has a negative
meaning that implies punishment and destruction, whereas, ‫ ﻏﯿﺚ‬has a
positive meaning that shows mercy. Hence, as Shehab highlighted,
when the two words are rendered as ‘rain’, the connotative meaning is
lost.
Synonyms in a religious context, such as that of the Holy Quran, are
a more complicated issue. Translators sometimes render some words
as synonyms when they are not. Arberry, for example, rendered ‫رﯾﺢ‬
/reeh/ and ‫ رﯾﺎح‬/riaah/ as synonyms. He translated them as ‘wind’ and
‘winds’, respectively. Even though it is true that the irst word is
singular and the second is plural, they do not have this implication in
the Quranic language. The singular form is utilized in the Holy Quran to
refer to punishment, while the plural form is utilized to refer to
blessings and bounties. However, the strategy followed by Arberry is
partially correct; he could have added a footnote explaining the
differences between ‘wind’ and ‘winds’ in Arabic. Translating what
seem to be synonymous verbs is also problematic. For example,
‫ﯾﺤﻠﻒ‬/yahlef/ and ‫ﯾﻘﺴﻢ‬/yuqsem/ were considered to be synonyms by
Arberry, as they were translated as ‘swear’. In Arabic, the two verbs
have different implications and associations. The verb ‫ ﯾﺤﻠﻒ‬is used in
the Holy Quran to refer to hypocrites and disbelievers, and refers to

Uploaded by S. M. Safi
breaking an oath, while the verb ‫ ﯾﻘﺴﻢ‬is utilized in the Holy Quran to
refer to believers who ful il their promises and oaths (Shehab, 2009).
Similar to the situations discussed above, Abdul-Raof (2004)
highlighted some of the problems that translators face and sometimes
fail to overcome. Abdul-Raof considers that failing to differentiate
between the meanings of cognitive synonyms and realize the nuances
between them will result in semantic voids. Consider the following
example from Surah al-Imran, ayah 3:
َ ‫ﻧﺠ‬ َ َ َ‫ﺼ ﱢﺪﻗًﺎ ﻟﱢ َﻤﺎ ﺑَ ْﯿ َﻦ ﯾَ َﺪ ْﯾ ِﻪ َوأ‬
َ ‫ﺎب ﺑِﺎﻟْ َﺤ ﱢﻖ ُﻣ‬
َ َ‫َز َل َﻋﻠَْﯿ َﻚ اﻟْ ِﻜﺘ‬
Example (3) ‫ﯿﻞ‬ ِ ْ ‫ﻧﺰل اﻟﺘﱠ ْﻮ َرا َة َو‬
ِ ‫اﻹ‬ ‫نﱠ‬
‘He has sent down upon thee the Book with the truth, con irming
what was before it, and He sent down the Torah and the Gospel’.
(Arberry, 1982, p. 43)

In this ayah, Arberry rendered the two verbs ‫ﻧﺰل‬/nazzala/ and ‫أﻧﺰل‬
/anzala/ as ‘send down’, as though the two verbs were synonyms.
However, in the SL they have different senses; the irst verb, ‫ﻧﺰل‬,
suggests the piecemeal revelation of the Holy Quran over 23 years. In
contrast, the second verb, ‫أﻧﺰل‬, suggests the revelation (of the gospel
and Torah) as a single event. These differences were not conveyed in
the TT. However, translating the ST word ‘sent down’ conveys the
primary meaning. It is not assumed that a TT will be a carbon copy of
its ST, but a translator attempts to reduce variations and maximize
sameness. Once again, Abdul-Raof’s (2004) argument, that the nuances
of meaning were not conveyed, is not of practical importance. It is
common for losses to occur in translation, even when the work is
carried out by the most professional of translators.
Another challenge regarding the use of synonyms relates to
collocated cognitive synonyms, which refers to the use of synonymous
words that come together for both emphasis and stylistic or aesthetic
purposes. The second synonym is commonly used to add beauty to the
text, or to create rhythm. Rendering such synonyms is dif icult to
achieve accurately in a TT, and when rendered they sound redundant
(Shehab, 2009). For example, ‫ اﻟﺮﺣﯿﻢ اﻟﺮﺣﻤﻦ‬are two names for Allah
Almighty; translating them usually creates a semantic void as the
nuances between the two Holy names cannot be conveyed in the
translation. However, I disagree with Shehab (2009) in the sense that,

Uploaded by S. M. Safi
in the Holy Quran, each synonym or lexeme serves a purpose that goes
beyond the mere aesthetic goal. The repetition in the Holy Quran,
though aesthetic, serves to strengthen or deepen the meaning
(Khorami, 2014). Translating the two names of Allah the Almighty can
be carried out by using intensifying words. For example, ‫ اﻟﺮﺣﻤﻦ‬can be
translated as ‘the most Merciful’, while ‫ اﻟﺮﺣﯿﻢ‬can be translated as
‘merciful’. Another option is to transliterate the word and provide
explanation to the word between brackets, or preferably in a footnote.
Another purpose for repeating synonyms is af irmation, as
mentioned by Az-Zarkashi (2006). For example, Surah al-Anam, ayah
125, reads:

Example ِ ُ‫ِﻺ ْﺳ َﻼ ِم ۖ َو َﻣﻦ ﯾُ ِﺮ ْد أَن ﯾ‬


(125) ‫ﻀﻠﱠ ُﻪ ﯾَ ْﺠ َﻌ ْﻞ‬ َ ‫ﻓَ َﻤﻦ ﯾُ ِﺮ ِد اﷲﱠُ أَن ﯾَ ْﻬ ِﺪﯾَ ُﻪ ﯾَ ْﺸ َﺮ ْح‬
ِ ْ ‫ﺻ ْﺪ َر ُه ﻟ‬
‫ﻮن‬ َ ‫ﺲ َﻋﻠَﻰ اﻟﱠﺬ‬
َ ُ‫ِﯾﻦ َﻻ ُﯾ ْﺆ ِﻣﻨ‬ َ ‫اﻟﺮ ْﺟ‬ َ ‫ﺎء ۚ َﻛ َﺬﻟ‬
‫ِﻚ َﯾ ْﺠ َﻌ ُﻞ اﷲﱠُ ﱢ‬ ‫ﺿ ﱢﯿﻘًﺎ َﺣ َﺮ ًﺟﺎ َﻛﺄَﻧﱠ َﻤﺎ َﯾ ﱠ‬
‫ﺼ ﱠﻌ ُﺪ ﻓِﻲ ﱠ‬
ِ ‫اﻟﺴ َﻤ‬ َ ‫ﺻ ْﺪ َر ُه‬
َ
‘And whomsoever it is God’s will to guide, He expands his bosom
to Islam (surrender), and whoever He wills to send astray, He makes
his bosom close and narrow as if he were engaged in sheer ascent to
the sky. Thus God lays humiliation upon those who disbelieve’.
(Pickthall, 2001, p. 67)

The two Quranic words ‫ﺿﯿﱢﻖ ا‬ َ and ‫ َﺣ َﺮ ًﺟﺎ‬are synonymous; however, they
serve the purpose of af irmation in addition to their role in adding
aesthetic lavour to the Quranic text (Al-Munajjid, 1997). The translator
rendered the two ST words as ‘close and narrow’, which is appropriate
and conveys the primary meaning of the ST words.
In a similar vein, Abdelwali (2007) gave an example of translating
the following ayah to show how loss in meaning is generated in
translating what look to be synonyms:

Example (10:5) ‫ﻫﻮ اﻟﺬي ﺟﻌﻞ اﻟﺸﻤﺲ ﺿﯿﺎء واﻟﻘﻤﺮ ﻧﻮرا‬


‘It is He who made the sun to be shining glory and the moon to be
a light’. (Ali, 1968, p. 127)

Although ‫ ﺿﯿﺎء‬/diaa’n/ and ‫ ﻧﻮرا‬/nuran/ may super icially look


synonyms (i.e. light), they have two distinct meanings. The word ‫ﺿﯿﺎء‬
re lects light accompanied by ‘heat’, while the second word, ‫ﻧﻮرا‬,

Uploaded by S. M. Safi
conveys light only. The translation could not render the two words
equivalently. Analysing the ST words, it can be noted that sun has the
features of glowing at daytime and that the moon has the features of
glowing at night. I suggest that better translations of these ST words
could have been ‫ ﺿﯿﺎء‬as ‘radiant’ and ‫ ﻧﻮرا‬as ‘luminous’. Again, analysing
the ST words componentially can help to provide the nearest
equivalents. A translator should translate based on the situation he is
experiencing. Another example is provided by Abdelwali and shows a
problem in translation:

Example (34:22) ‫اﷲ‬ ‫ﻗﻞ ادﻋﻮ اﻟﺬﯾﻦ زﻋﻤﺘﻢ ﻣﻦ دون ﱠ‬


‘Say: Appeal to those whom you claim to instead of God’. (Irving,
1985)

The Quranic word ‫ دون‬was rendered as ‘instead’, which does not convey
the different connotations and denotations of the word. The Quranic
word refers the inferiority of those taken as gods in comparison to
Allah’s Almighty power, Who Only deserves worship. However,
translating the ST word as ‘instead’ seems to be suf icient as it shows
the primary meaning, though it may not convey the other shades of
meaning.
In her study of the extraordinary vocabulary in the Holy Quran,
Emara (2013) underscored the problems of translating some
‘extraordinary’ vocabularies in the Holy Quran. Consider the
translation of the word ‫ اﻟْ َﻌ َى َﺚ‬in the following context:

Example (4:25) ‫…ذﻟﻚ ﻟﻤﻦ ﺧﺸﻰ اﻟﻌﻨﺖ ﻣﻨﻜﻢ‬


‘…This (permission) is for those among you who fear sin’. (Ali,
1968, p. 53)

As can be seen in the translation, the word ‫ اﻟﻌﻨﺖ‬was translated as ‘sin’,


which is not accurate because the Quranic word in this context refers
to ‘adultery’ or ‘fornication’ (Emara, 2014). This is prevalent in the
translation of many ayahs. The ST word, however, could have been
translated as ‘adultery’ or ‘fornication’ because this is the meaning
intended in this context, as mentioned in commentary books. In short,
translation is a cognitive process that is based on situation and,

Uploaded by S. M. Safi
therefore, a translator should adopt the appropriate translation
strategies to handle any problems faced in the translation process.

4.5 Problems in Translation of Rhetorical


Devices
Style is not only an important linguistic component in translation, but
is also a component of meaning (Ghazala, 2008). It does not serve only
aesthetic purposes but is a vital component that, without translation,
creates an incomplete translation (Zaky, 2001). Many theorists in
translation underscored the importance of style in translation. For
example, Nida and Taber (1982), in their de inition of translation,
identi ied natural equivalence as an element that should occur, irst, in
meaning and, second, in style. Rhetorical devices—for example,
metaphor, ellipsis and metonymy—are one of the prominent stylistic
features of a language likely to cause problems in translation. Quinn
(1993) postulates that rhetorical igures are intended deviation from
the normal use of language. Based on Corbett and Huhmann, as cited in
Cui and Zhao (2014, p. 59), igures of speech include alliteration,
anadiplosis, anaphora, antithesis, antimetabole, ellipsis, epanalepsis,
epanorthosis, epistrophe, hyperbole, irony, metaphor, metonymy,
paradox, parallelism, pun, rhetorical questions, rhyme and simile.
1.
Alliteration: This refers to repetition of the same initial consonant
sound in series of words or sentences.
2.
Anadiplosis: This is the repetition of the last word of one clause at
the beginning of the following clause.
3.
Anaphora: This is the repetition of a word or group of words at
the beginning of successive clauses.
4.
Antithesis: This is the juxtaposition of contrasting ideas, often in
parallel structures.
5.
Antimetabole: This is the repetition of words in successive
clauses, in reverse grammatical order.

Uploaded by S. M. Safi
6. Ellipsis: This is the deliberate omission of a word or words
readily implied by the context.
7.
Epanalepsis: This is the repetition at the end of a clause of the
word or phrase that has occurred at the beginning of the clause.
8.
Epanorthosis: This is making a claim that calls that claim into
doubt.
9.
Epistrophe: This is the repetition of the same word or group of
words at the ends of successive clauses.
10.
Hyperbole: This is the use of exaggerated terms for emphasis or
heightened effect.
11.
Irony: This is the use of a word in such a way as to convey a
meaning opposite to the literal meaning of that word.
12.
Metaphor: This implies comparison between two things of
dissimilar natures.
13.
Metonymy: This is the substitution of an attributive or suggestive
word for what is actually meant.
14.
Paradox: This is an apparently contradictory statement that
nevertheless contains a measure of truth.
15.
Parallelism: This is refers to a similarity of structure in a series of
two or more related words, phrases, or clauses.
16.
Pun: This is the use of a word that has different meanings,
repeating a word but in such a way that it has a different meaning
each time, changing a word’s meaning, or using words that sound
alike but differ in meaning.
17.
Rhetorical questions: This is the asking of a question for a speci ic
purpose other than to obtain information.
18.
Rhyme: This is the repetition of sounds at the end of words or
phrases.

Uploaded by S. M. Safi
19. Simile: This is explicit comparison made by the use of ‘like’ or ‘as’
to attribute connotations and meanings of one object to another.
Rhetorical devices are frequently used in poetry, advertisements
and even in everyday language use. In this regard, Xu (2008) believes
that literal translation, free translation and modulation could be
useful strategies in translating rhetoric devices.
Another genre that is rich in its use of rhetorical devices is the
Quranic language, which is characterized by its use of alliteration,
antithesis, metaphor and oxymoron. Unfortunately, these devices are
often lost in translation, which presents a challenge to the maintaining
of accuracy in translation. Metaphor is one of the most prominent
rhetorical devices in the language of the Holy Quran; it is the
phenomenon we utilize to talk and think about something in terms of
something else (Semino, 2008). The cognitive view of metaphor
considers metaphor not only as a rhetorical by-product of objective
thinking, but also as the basis of the human conceptual system. One of
the most important types of metaphor is the conceptual metaphor,
which is widely used in the Holy Quran. Conceptual metaphor refers to
understanding one conceptual domain in terms of another conceptual
domain; in other words, a source domain that can be de ined as the
domain from which we draw metaphorical expressions to understand
another conceptual domain: the domain that is understood in this way
is the target domain (Kövecses, 2010). Metaphor, for Newmark, has a
broader meaning than the traditional one. Newmark (1988) states that
metaphor, for him, refers to any igurative expression; this include
phrasal verbs, collocations, idioms, proverbs, allegory and
personi ication among others. Metaphors, he adds, have two main
functions: referential and aesthetic. Metaphor makes concepts and
thoughts clearer (Said Ghazala, 2012). They are a basic scheme by
which human experience and the outside world are conceptualized
(Gibbs, 1994). Al Salem (2014) gives the following examples of the
translation of metaphors in Mahmoud Darwish’s poems:

Example

ST TT

Uploaded by S. M. Safi
ST TT

1.‫ﻻ ﻣﻮﺳﻮﻋﺔ اﻷزﮬﺎر‬ ‘No encyclopaedia of lowers is any help to Me’


‫ﺗﺴﻌﻔﻨﻲ‬
2.‫واﻟﺼﻼة ﺗﻜﻠﺴﺖ‬ ‘The prayer calci ied’

3.‫أ ﯾﮭﺎ اﻟﺤﺎﺿﺮ! ﺗﺤﻤﻠﻨﺎ‬ ‘Oh present! Be a little patient with us, for we are only passers-by with
heavy shadows’
‫ﻗﻠﯿﻼ ﻓﻠﺴﻨﺎ‬
‫ﺳﺒﯿﻞ ﺛﻘﻼء‬
ٍ ‫ﺳﻮى ﻋﺎﺑﺮي‬
ّ
‫اﻟﻈﻞ‬

4.‫ ﺣﺒّﻲ ﻧﺰﮬ ٌﺔ‬:‫أﺟﺎب‬ ‘He answered: ‘My love is a short outing, a glass of wine, an affair/a love
affair.’
‫ﻗﺼﯿﺮة‬
‫ أو ﻣﻐﺎﻣﺮة‬.. ‫أو ﻛﺄس ﺧﻤﺮ‬

In example 1, Al Salem notes that the metaphor was translated into a


sense, instead of a functionally equivalent metaphor. However, in
example 2 the translator rendered the ST metaphor literally. In example
3, the ST metaphor was translated as ‘heavy shadows’, which is
inaccurate literal translation. I argue that the ST metaphor should have
been rendered functionally as ‘bleeder’ or ‘boring’. This would cause the
TT to lose the effect of the igurative use of the expression in the ST. In
example 4, the metaphor ‫ﻧﺰﮬﺔ ﻗﺼﯿﺮة‬ٌ ‫ ﺣﺒّﻲ‬was translated literally as ‘My
love is a short outing’.
In relation to the Holy Quran, metaphors are not only utilized for
aesthetic purposes, but also serve other purposes, such as conveying
abstract meanings, clarifying the intended meaning, and impressing its
readers (Kabali, 2006). However, metaphors are dif icult to render into
English accurately and with the same effect that exists in the authentic
Quranic text; that is why, sometimes, translators render them into
literal language that does not convey the true meaning. Consider the
following example from Surah Yusuf, ayah 9, as provided by Ali et al.
(2012):

Example (12:9) ‫ﺿﺎ ﯾَ ْﺨ ُﻞ ﻟَ ُﻜ ْﻢ َو ْﺟ ُﻪ أَِﺑﯿ ُﻜ ْﻢ َوﺗَ ُﻜﻮﻧُﻮا ِﻣﻦ ﺑَ ْﻌ ِﺪ ِه َﻗ ْﻮ ًﻣﺎ‬


ً ‫اﻃ َﺮ ُﺣﻮ ُه أَ ْر‬
ْ ‫ﻒ أَ ِو‬ ُ ُ‫ْاﻗﺘُﻠُﻮا ﯾ‬
َ ‫ﻮﺳ‬
‫ﯿﻦ‬
َ ‫ِﺤ‬
ِ ‫ﺻﺎﻟ‬َ
‘Kill you Joseph, or cast him forth into some land, that your
father’s face may be free for you, and thereafter you may be a

Uploaded by S. M. Safi
righteous people’. (Arberry, 1982, p. 141)

Considering Arberry’s translation, it is clear that it did not convey the


intended meaning; it shifted away from the authentic meaning. Arberry
translated ‫ ﯾﺨﻞ ﻟﻜﻢ وﺟﻪ أﺑﯿﻜﻢ‬as ‘your father’s face may be free for you’. The
metaphorical meaning of the ayah is that after killing Yusuf (PBUH), the
attention and care that the prophet Yaqoub gave to Yusuf (PBUH(will
drive his other sons apart from Yusuf (PBUH). The translation could
not convey this metaphorical meaning (Ali et al., 2012). However, this
seems to be the only possible way of translating the Quranic metaphor.
Literal translation may be the only way to translate the Quranic
igures of speeches and images that do not have equivalents in the TL.
The Holy Quran is an authoritative canonical text that cannot be
rendered using a functional equivalent.
Another example is provided by Abdul-Raof (2004) and indicates
how translation failed to convey the meaning expressed in the Holy
ayah. Consider the following ayah from Surah Maryam (ayah, 4):

Example (19:4) ‫ِﻚ َر ﱢب‬ َ ‫س َﺷ ْﯿﺒﺎ َوﻟَ ْﻢ أَ ُﻛﻦ ِﺑ ُﺪ َﻋﺎﺋ‬


ُ ْ‫اﺷﺘَ َﻌ َﻞ اﻟﺮأ‬
ْ ‫ﺎل َر ﱢب إِﻧﱢﻲ َو َﻫ َﻦ اﻟْ َﻌ ْﻈ ُﻢ ِﻣﻨﱢﻲ َو‬
َ َ‫ﻗ‬
‫َﺷﻘِﯿﺎ‬
‘Praying: O my Lord! in irm indeed are my bones, and the hair of
my head doth glisten with grey: but never am I unblest, O my Lord, in
my prayer to Thee!’ (Ali, 1968, p. 200)

In the ayah, the word ‫اﺷﺘﻌﻞ‬/ishtaAAala/is used iguratively, as it literally


means ‘burn’, but the translation lost the metaphoric use of the word.
The English term ‘glisten with grey’ does not convey the same
authentic meaning of the Quranic word (i.e. ‫)اﺷﺘﻌﻞ‬. The Quranic word is
used metaphorically to convey the spread of greyness in the hair of the
Prophet Zakkariah, which, in turn, re lects growing old (Abdul-Raof,
2004).
Likewise, metonymy—which refers to the substitution of contained
for the container, or the effect for the cause (Quinn, 1993)—is another
of the rhetorical devices employed in the Holy Quran. It re lects
meaning in a more powerful way. However, due to cultural factors,
translation rarely conveys the metonymy in the TT. Consider the
following example from Surah al-Anaam, ayah 7:
‫ﱢ ﱠ‬ َْ ‫ﱠﱠ‬ َ َ َ َ

Uploaded by S. M. Safi
Example (6:6) ‫ض َﻣﺎ ﻟَ ْﻢ ﻧُ َﻤ ﱢﻜﻦ ﻟﱠ ُﻜ ْﻢ‬ ِ ‫ِﻬﻢ ﱢﻣﻦ َﻗ ْﺮ ٍن ﱠﻣ ﱠﻜﻨﱠﺎ ُﻫ ْﻢ ﻓِﻲ اﻷَْ ْر‬ ِ ‫أَﻟَ ْﻢ ﯾَ َﺮ ْوا َﻛ ْﻢ أَ ْﻫﻠَ ْﻜﻨَﺎ ِﻣﻦ َﻗ ْﺒﻠ‬
‫ﻧﺸ ْﺄﻧَﺎ ِﻣﻦ‬
َ َ‫ِﻬ ْﻢ َﻓﺄَ ْﻫﻠَ ْﻜﻨَﺎﻫُﻢ ﺑِ ُﺬﻧُﻮﺑِ ِﻬ ْﻢ َوأ‬
ِ ‫ارا َو َﺟ َﻌ ْﻠﻨَﺎ اﻷَْ ْﻧ َﻬﺎ َر ﺗَ ْﺠ ِﺮي ِﻣﻦ ﺗَ ْﺤﺘ‬ ‫َوأَ ْر َﺳ ْﻠﻨَﺎ ﱠ‬
ً ‫اﻟﺴ َﻤﺎ َء َﻋﻠَ ْﯿ ِﻬﻢ ﱢﻣ ْﺪ َر‬
‫ﯾﻦ‬ َ ‫ِﻢ َﻗ ْﺮﻧًﺎ‬
َ ‫آﺧ ِﺮ‬ ْ ‫َﺑ ْﻌ ِﺪﻫ‬
‘See they not how many of those before them We did destroy? —
generations We had established on the earth, in strength such as We
have not given to you—for whom We poured out rain from the skies
in abundance, and gave (fertile) streams lowing beneath their (feet):
yet for their sins We destroyed them, and raised in their wake fresh
generations (to succeed them)’. (Ali, 1968, p. 79)

ً ‫ﺎء َﻋﻠَْﯿ ِﻬﻢ ﱢﻣ ْﺪ َر‬


In this ayah, the translator rendered ‫ارا‬ ‫وأَ ْر َﺳﻠْﻨَﺎ ﱠ‬,
َ ‫اﻟﺴ َﻤ‬ َ which is a
metonymy in the SL, as ‘We poured out rain from the skies in
abundance’. The translator could not keep the same metonymy of the
SL due to the cultural discrepancies, as one language may use certain
terms metonymically, while another language cannot. Hence, in the
example above, the translator could do no more than pursue semantic
translation (Ali et al., 2012).
Alliteration is another rhetorical device in which several words that
are close together begin with the same letter or sound (Collins COBUILD
Dictionary, 2006). Alliteration is employed in the Holy Quran to
impress readers and to give them good mental space in which to
consider the meanings of the Holy Quran; it also interweaves meanings
and makes them melodic (Ayyash et al., 2013). The miracle of the
language of the Holy Quran is that alliteration not only serves melodic
purposes, but also serves to support meaning. An example to illustrate
how translation fails to convey the alliteration in the ST is found in the
following example from Surah al-Baqara, ayah 114:

Example (114:2) ۚ ‫اﺳ ُﻤ ُﻪ َو َﺳ َﻌﻰ ﻓِﻲ َﺧ َﺮا ِﺑ َﻬﺎ‬ َ ‫اﷲ أَن ُﯾ ْﺬ َﻛ َﺮ ﻓ‬


ْ ‫ِﯿﻬﺎ‬ ِ‫ﺎﺟ َﺪ ﱠ‬ ‫َو َﻣ ْﻦ أَ ْﻇﻠَ ُﻢ ﻣ ﱠ‬
ِ ‫ِﻤﻦ ﱠﻣﻨَ َﻊ َﻣ َﺴ‬
ٌ ‫ِﯿﻦ ۚ ﻟَ ُﻬ ْﻢ ﻓِﻲ اﻟ ﱡﺪ ْﻧﯿَﺎ ِﺧ ْﺰ ٌي َوﻟَ ُﻬ ْﻢ ﻓِﻲ ْاﻵ ِﺧ َﺮ ِة َﻋ َﺬ‬
‫اب َﻋ ِﻈﯿﻢ‬ َ ‫ﻮﻫﺎ إِﱠﻻ َﺧﺎﺋِﻔ‬ ُ ‫ُﻢ أَن ﯾَﺪ‬
َ ُ‫ْﺧﻠ‬ ْ ‫ﺎن ﻟَﻬ‬ َ ‫أُوﻟَﺌ‬
َ ‫ِﻚ َﻣﺎ َﻛ‬
‘And who does greater evil than he who bars God’s places of
worship, so that His Name be not rehearsed in them’. (Arberry, 1982,
p. 27)

In this example, alliteration is achieved in the authentic Arabic ayah by


repeating the same sound or letter: ‫ م‬. However, the translation lost

Uploaded by S. M. Safi
these melodic sounds (Abdul-Raof, 2004). Melody in the Holy Quran
has a spiritual shadow of meaning, which is felt on reading the
authentic text.
Antithesis is a further rhetorical device that is lost in translation. It
refers to the utilizing of parallelistic structures to show contrast in
meaning; however, translation fails to convey this type of structure.
Consider the following example:

Example (14 :82) ‫ﯿﻢ‬ ٍ ‫( َوإِ ﱠن ا ْﻟ ُﻔ ﱠﺠﺎ َر َﻟﻔِﻲ َﺟ ِﺤ‬13:82) ‫ِﯿﻢ‬ َ ‫إِ ﱠن اﻷَْ ْﺑ َﺮ‬
ٍ ‫ار َﻟﻔِﻲ َﻧﻌ‬
‘As for the Righteous, they will be in bliss; And the Wicked—they
will be in the Fire’. (82:13–14) (Ali, 2006, p. 441)

Examining the ayahs clearly shows the loss in conveying the meaning
of the antithesis that exists in the SL, since ‘Righteous’ is not an
antithesis of ‘wicked’, neither is ‘bliss’ an antithesis of ‘ ire’ (Abdul-Raof,
2004). Similarly, oxymoron is a rhetorical device is employed
effectively in the Holy Quran. It refers to the use of two adjacent
antonyms. Consider the following ayah (7) in Surah at-Talaq:

Example ُ ‫ِﻖ ُذو َﺳ َﻌ ٍﺔ ﱢﻣﻦ َﺳ َﻌﺘِ ِﻪ ۖ َو َﻣﻦ ُﻗ ِﺪ َر َﻋﻠَ ْﯿ ِﻪ ِر ْز ُﻗ ُﻪ َﻓ ْﻠﯿُﻨ ِﻔ ﱠﻘﺎ آﺗَﺎ ُه اﷲﱠُ ۚ َﻻ ﯾُ َﻜﻠﱢ‬
(7:65) ‫ﻒ‬ ْ ‫ﻟِﯿُﻨﻔ‬
‫ﺎﻫﺎ ۚ َﺳﯿَ ْﺠ َﻌ ُﻞ اﷲﱠُ ﺑَ ْﻌ َﺪ ُﻋ ْﺴ ٍﺮ ﯾُ ْﺴ ًﺮا‬
َ َ‫اﷲﱠُ ﻧَ ْﻔ ًﺴﺎ إِﱠﻻ َﻣﺎ آﺗ‬
‘Let the man of means spend according to his means: and the man
whose resources are restricted, let him spend according to what Allah
has given him. Allah puts no burden on any person beyond what He
has given him. After a dif iculty, Allah will soon grant relief’. (Ali, 2006,
p. 406)

In the ayah, an oxymoron exists between the two antonym words ‫ﻋﺴﺮ‬
/AAusran/and ‫ ﯾﺴﺮا‬/usran/. This is due to the nature of the Arabic
language, which allows the object to precede the subject (Abdul-Raof,
2004). Ali’s translation could not keep the same rhetorical device as it
exists in the ST.
Ellipsis is another common loss in the translation of the Holy
Quran. Ellipsis is the term given to instances of anaphora in which a
missing predicate can be understood from context (Johnson, 2001).
The Arabic language permits this type of construction, which English

Uploaded by S. M. Safi
does not allow in the same way (Al-Azab & Al-Misned, 2012). Ellipsis is
one of the very characteristics of the Quranic text, which translation
fails to convey (Al-Azab & Al-Misned, 2012; Ali et al., 2012). Consider
the following example:

Example (31:13) ۗ ‫ض أَ ْو ُﻛﻠﱢ َﻢ ِﺑ ِﻪ ا ْﻟ َﻤ ْﻮﺗَﻰ‬


ُ ‫َوﻟَ ْﻮ أَ ﱠن ُﻗ ْﺮآﻧًﺎ ُﺳ ﱢﯿ َﺮ ْت ِﺑ ِﻪ ا ْﻟ ِﺠ َﺒﺎ ُل أَ ْو ُﻗ ﱢﻄ َﻌ ْﺖ ِﺑ ِﻪ اﻷَْ ْر‬
‫ِﯾﻦ َﻛ َﻔ ُﺮوا‬ ُ ‫ﺎس َﺟﻤِﯿ ًﻌﺎ ۗ َو َﻻ َﯾﺰ‬
َ ‫َال اﻟﱠﺬ‬ َ ‫آﻣﻨُﻮا أَن ﻟﱠ ْﻮ َﯾ َﺸﺎ ُء اﷲﱠُ ﻟَ َﻬﺪَى اﻟﻨﱠ‬ ِ َ‫ﷲ اﻷَْ ْﻣ ُﺮ َﺟﻤِﯿ ًﻌﺎ ۗ أََﻓﻠَ ْﻢ َﯾ ْﯿﺄ‬
َ ‫س اﻟﱠﺬ‬
َ ‫ِﯾﻦ‬ ِ‫َﺑﻞ ﱢﱠ‬
ُ ‫اﷲ ۚ إِ ﱠن اﷲﱠَ َﻻ ﯾُ ْﺨﻠ‬
‫ِﻒ ا ْﻟﻤِﯿ َﻌﺎ َد‬ ِ‫ِﻲ َو ْﻋ ُﺪ ﱠ‬
َ ‫ِﻢ َﺣﺘﱠﻰ ﯾَ ْﺄﺗ‬
ْ ‫ﺻﻨَ ُﻌﻮا َﻗﺎ ِر َﻋ ٌﺔ أَ ْو ﺗَ ُﺤ ﱡﻞ َﻗ ِﺮﯾﺒًﺎ ﱢﻣﻦ دَا ِرﻫ‬ ِ ُ‫ﺗ‬
َ ‫ﺼﯿﺒُﻬُﻢ ﺑِ َﻤﺎ‬
‘Had it been possible for a Lecture to cause the mountains to
move, or the earth to be torn asunder, or the dead to speak, (this
Qur’an would have done so). Nay, but Allah’s is the whole command.
Do not those who believe know that, had Allah willed, He could have
guided all mankind? As for those who disbelieve, disaster ceaseth not
to strike them because of what they do, or it dwelleth near their home
until the threat of Allah come to pass. Lo! Allah faileth not to keep the
tryst’. (Pickthall, 2001, p. 106)
‘If only a Koran whereby the mountains were set in motion, or the
earth were cleft, or the dead were spoken to—nay, but God’s is the
affair altogether. Did not the believers know that, if God had willed, He
would have guided men all together? And still the unbelievers are
smitten by a shattering for what they wrought, or it alights nigh their
habitation, until God’s promise comes; and God will not fail the tryst’.
(Arberry, 1982, p. 151)

In the ST, the result clause is ellipted to create impressive effect on


readers, and this is recurrent in the Holy Quran. However, as seen in the
translations above, this ellipted meaning could not be conveyed in
translation (Al-Azab & Al-Misned, 2012). Even though Pickthall
intervened by adding an extra clause to clarify the ellipted clause, the
translation could not completely convey the meaning. Arabic native
speakers understand that there is an ellipted clause; nonetheless, they
have a clear understanding of the meaning. English translations fail to
re lect the same elliptical structure.
Another example of ellipsis that is lost in translation, as provided
by Ali et al. (2012), is the translation of the following ayah (82) from

Uploaded by S. M. Safi
Surah Yusuf, which reads:

Example 12 :82) ‫ﻮن‬


َ ‫ِﻗ‬ َ َ‫ِﯿﺮ اﻟﱠﺘِﻲ أَ ْﻗ َﺒ ْﻠﻨَﺎ ﻓِﯿ َﻬﺎ ۖ َوإِﻧﱠﺎ ﻟ‬
ُ ‫ﺼ ﺎد‬ َ ‫اﺳﺄَ ِل ا ْﻟ َﻘ ْﺮ َﯾ َﺔ اﻟﱠﺘِﻲ ُﻛﻨﱠﺎ ﻓ‬
َ ‫ِﯿﻬﺎ َوا ْﻟﻌ‬ ْ ‫) َو‬
‘Ask at the town where we have been and the caravan in which
we returned, and (you will ind) we are indeed telling the truth’. (Ali,
2006, p. 153)

In the Quranic text, there is a metonymy that includes ellipsis; however,


the translator sought to add ‘at’, which is not in the ST, to clarify the
meaning. The ayah shows loss in the translation of the ellipsis. In sum,
rhetorical meaning is sometimes lost in translation due to a
translator’s inability to ind the equivalent lexical item in the TL, or
because of the syntactic structure of the SL. Differences in culture make
it more dif icult for translation to retain the same rhetorical devices
that exist in the TL.
Another feature of the Holy Quran is that names are rarely stated,
with the exception of names of prophets and some angels; this is so as
to re lect the general effect and the universality of its message. The
Quranic text, no doubt, is rich with metaphors, shifts, foregrounding,
metonymy, simile, repetition, ellipsis, al-saj’, puns and so forth (Jaber,
2010). Jaber gives an example of how translators face dif iculty in
translating the Quranic images. Consider the following translation of
ayah 2 in Surah Al Baqarah:

Example (92:2) ۖ ‫اﺳ َﻤ ُﻌﻮا‬ ‫َوإِ ْذ أَ َﺧ ْﺬﻧَﺎ ﻣِﯿﺜَ َﺎﻗ ُﻜ ْﻢ َو َر َﻓ ْﻌﻨَﺎ َﻓ ْﻮ َﻗ ُﻜ ُﻢ ﱡ‬


ْ ‫اﻟﻄﻮ َر ُﺧ ُﺬوا َﻣﺎ آﺗَ ْﯿﻨَﺎ ُﻛﻢ ِﺑ ُﻘ ﱠﻮ ٍة َو‬
َ ِ‫ِﺠ َﻞ ِﺑ ُﻜ ْﻔ ِﺮ ِﻫ ْﻢ ۚ ُﻗ ْﻞ ِﺑ ْﺌ َﺴ َﻤﺎ ﯾَ ْﺄ ُﻣ ُﺮ ُﻛﻢ ِﺑ ِﻪ إ‬
‫ﯾﻤﺎﻧُ ُﻜ ْﻢ إِن ُﻛﻨﺘُﻢ‬ ْ ‫ﺼ ْﯿﻨَﺎ َوأُ ْﺷ ِﺮﺑُﻮا ﻓِﻲ ُﻗﻠُﻮ ِﺑ ِﻬ ُﻢ ا ْﻟﻌ‬ ْ ‫َﻗﺎﻟُﻮا َﺳﻤ‬
َ ‫ِﻌﻨَﺎ َو َﻋ‬
َ ‫ﱡﻣ ْﺆ ِﻣﻨ‬
‫ِﯿﻦ‬
‘And remember We took your covenant and We raised above you
(the towering height) of Mount (Sinai): (Saying): “Hold irmly to what
We have given you, and hearken (to the Law)”: They said: “We hear,
and we disobey:” And they had to drink into their hearts (of the taint)
of the calf because of their Faithlessness. Say: “Vile indeed are the
behests of your Faith if ye have any faith!”’. (Ali, 1968, p. 15)
‘Hold fast by that which We have given you, and hear (Our Word),
they said: “We hear and we rebel.” And (worship of) the calf was made

Uploaded by S. M. Safi
to sink into their hearts because of their rejection (of the covenant).
Say (unto them): “Evil is that which your belief enjoineth on you, if ye
are believers”’. (Pickthall, 2001, p. 20)
‘And when We made a covenant with you and raised the mountain
over you: Take hold of what We have given you with irmness and be
obedient. They said: “We hear and disobey.” And they were made to
imbibe (the love of) the calf into their hearts on account of their
unbelief. Say: “Evil is that which your belief bids you if you are
believers”’. (Al-Bayt, 2008, p. 3)

As seen in the above translations, the Quranic word ‫ َوأُ ْﺷﺮﺑﻮا‬was


translated as ‘drink’ by Ali; as ‘sink’ by Pickthall; and ‘imbibe’ by Ahl-ul-
bait. According to Jaber (2010), the most accurate rendition is that of
Ahl-ul-bait. She assumes that ‘imbibe’ means ‘to absorb or to receive
into the mind’, which is the closest to the ST. It seems that being
among native speakers of the Arabic language, linguists, and teamwork
helped Ahl-ul-Bait to give better translations, as Jaber states. Another
example that was provided by Jaber is the translation of the Quranic
word ‫‘ ﺷﻌﺎﺋﺮ‬/sha’aer/in the following ayah:
ُْ َ ‫َذﻟ‬
Example (22: 32) ‫ﻮب‬ ِ ‫ﻧﻬﻬﺎ ِﻣﻦ ﺗَ ْﻘ َﻮى اﻟﻘُﻠ‬ َ ‫ِﻚ َو َﻣﻦ ﯾُ َﻌ ﱢﻈ ْﻢ َﺷ َﻌﺎﺋ‬
َ ‫ِﺮ اﻟﻠﻬ ِﻪ ﻓَ ِﺈ‬
‘Such (is the Pilgrimage): whoever honors the sacred rites of
Allah, for him it is good in the Sight of his Lord’. (Ali, 1968, p. 224)
‘That (is the command). And whoso magni ieth the offerings
consecrated to Allah, it surely is from devotion of the hearts’.
(Pickthall, 2001, p. 138)
‘That (shall be so); and whoever respects the signs of Allah, this
surely is (the outcome) of the piety of hearts’. (Shakir, 1999, p. 157)

As seen in these examples, Ali translated the Quranic word ‫ َﺷ َﻌﺎﺋِﺮ‬as


‘sacred rites’, Pickthall translated it as ‘the offerings’ and Shakir
translated it as ‘the signs’. According to Jaber, the most accurate
translation is that of Shakir. Having discussed how rhetorical devices
are challenging in translation, certain coping strategies can be
suggested: literal translation, modulation, free translation or functional
equivalent-based translation, and ideational equivalence-based
translation.

Uploaded by S. M. Safi
1. Literal translation: This can be useful when the ST and TT
rhetorical devices are similar. Baker (2011), in her discussion on
idioms, suggested that literal translation can be a helpful strategy
in translation. An example of this is provided by scienti ic journal
of Faculty of Education, Misurata University, Libya, in an English
advertisement that reads ‘Lipton tea can do that’ promoting Lipton
tea as a means of relaxation that can help to solve complex
brainteasers. It was translated literally as ‫ﺷﺎي ﻟﯿﺒﺘﻮن ﯾﻘﺪر ﻋﻠﻰ ﻛﺪه‬.
Although the translation is literal, it conveyed the ST meaning and
the rhetorical impact of the ST. Literal translation can be a useful
strategy in translating certain rhetorical devices but not all of
them. Some culturally bound terms may need alternative
translation strategies.
2.
Modulation: This is a useful strategy in translating culturally
bound rhetorical devices. One example that was given by the
scienti ic journal of the Faculty of Education, Misurata University,
Libya, is the English advertisement ‘add life to life’, which is an
advertisement for a telecom company in the United Arab Emirates.
It is a pun that is dif icult to render. However, the translator was
able to render the ST advertisement as ‫و ﺗﺤﯿﺎ ﺑﻬﺎ اﻟﺤﯿﺎة‬. He used
modulation, which was a successful strategy with which to render
the ST rhetorical device, retaining the same effect in the TL.
3.
Free translation or functional equivalent-based translation:
Sometimes it is dif icult to use modulation or literal translation to
render certain rhetorical devices, which leaves a translator no
option but to attempt to convey the function of the rhetorical
device.
4.
Ideational equivalence-based translation: This can be the last
resort for a translator and involves conveying the idea but omitting
the rhetorical features.

Exercise Translate the following sentences. Then, explain the


strategies employed in translating the igures of speech in the
sentences.

Uploaded by S. M. Safi
1. ‘Her tears were a river lowing down her cheeks.’

2.
‘The boy stalked his brother before inally pouncing on his
prey.’
3.
‘The wind was a howling wolf.’
4.
‘My teacher is a dragon.’
5.
‘My friend is a chicken.’
6.
‘Scholars are shining stars.’
7.
‘The sky is crying.’
8.
‘The noise gave me a headache.’
9.
‘The sun was smiling in the sky.’
10.
‘I am all ears.’
11.
‘This step is the beginning of the end.’
12.
‘The clap of thunder went bang and scared my poor dog.’
13.
‘That kitchen knife will take a bite out of your hand if you don’t
handle it safely.’
14.
‘Ahmed was as white as a sheet after he walked out of the
horror movie.’

References
Abdelwali, M. (2007). The loss in the translation of the Qur’an [Electronic version]. The
Translation Journal, 11(2), 120–125. Retrieved from http://www.accurapid.com/journal/

Uploaded by S. M. Safi
40quran.htm.

Abdul-Raof, H. (2004). The Quran: Limits of translatability. In S. Faiq (Ed.), Cultural encounters in
translation from Arabic. Frankfurt Lodge: Multilingual Matters Ltd.

Abdul-Raof, H. (2005). Pragmalinguistic forms in cross-cultural communication: Contributions


from Qur’an translation. Intercultural Communication Studies, 4, 115–130.

Abedelrazq, Y. I. (2014). Problems of translating homonymy in the glorious Quran: A comparative


analytical study (Unpublished thesis).

Al-Azab, A., & Al-Misned, A. (2012). Pragmatic losses of Qur’an translation: A linguistic approach.
English Language and Literature Studies, 2(3), 42–49. https://doi.org/10.5539/ells.v2n3p42.

Al-Bayt, R. A. (2008). The Qur’an: A new English translation of its meanings. Amman: Royal Aal al-
Bayt Institute for Islamic Thought.

Al-Ghazali, M. F. (2010). Lexical gaps in Arabic -to-English translation. Al-Mustansiriya Journal of


Arts, 1(52), 1–16.

Ali, A., Brakhw, M. A., Bin Nordin, M. Z. F., & ShaikIsmail, S. F. (2012). Some linguistic dif iculties in
translating the Holy Quran from Arabic into English. International Journal of Social Science and
Humanity, 2(6), 588–590. https://doi.org/10.7763/IJSSH.2012.V2.178.

Ali, A., Brakhw, M. A., & Zarirruddin, M. (2014). Transferring polysemic words from Arabic into
English: A comparative study of some samples from the Holy Quran. Australian Journal of Basic
and Applied Sciences, 8(23), 38–43.

Ali, A. Y. (1968/2006). The Holy Qur’an, text, translation and commentary (Trans.). Beirut,
Lebanon: Dar al Arabia.

Al-Munajjid, M. N. (1997). Attradef i ALQuran al Kareem bayna Al Nazariyyah wa Att Atbeeq


[Synonyms in the Holy Quran from a theoretical and a practical perspective]. Beirut: Dar Al Fikr Al
Moaser.

AlQinai, J. (2012). Convergence and divergence in the interpretation of Quranic polysemy and
lexical recurrence. Pan-Paci ic Association of Applied Linguistics, 16(1), 83–109.

Al-Qurtubi, M. S. (2004). Al JamAA liahkam al Qur’an (Tafsir Al Qurtubi). Cairo, Egypt: Dar Al-Fikr.

Al Salem, M. N. (2014). The translation of metaphor from Arabic to English in selected poems of
Mahmoud Darwish with a focus on linguistic issues (Unpublished thesis). The University of Leeds.

Al-Utbi, M. (2011). Translation of ‫ ﻛﺎد‬in the Quran as an instance of lexical gaps. Majallat Kulliatu
Aladab, 1(98), 69–92.

Arberry, A. (1982). The Koran interpreted. Oxford: Oxford University Press.

Asyuti, J. (2008). Muzhir i Olum Allughah wannwa’aha [The science of language and its types]
(3rd ed.). Cairo: Dar Al-Turath.

Ayyash, J., Ahmad, S., & Abdullah, N. (2013). Models of alliteration derivation in the Quran.

Uploaded by S. M. Safi
International Journal of Islamic Thought, 3(June), 113–120. http://journalarticle.ukm.my/6417/
1/11_Jamil_Ayyash_IJIT_Vol_3_Jun_2013.pdf.

Az-Zarkashi, M. B. E. (2006). Al Burhan i Oloum Al Quran [The evidence in the Holy Quran
sciences]. Cairo: Dar al-Hadeeth.

Baker, M. (1992/2011). In other words: A coursebook on translation (2nd ed.). London and New
York: Routledge.

Benfoughal, A. (2010). Students’ dif iculties and strategies in translation: The case of third year
students (PhD thesis). University of Constantine, Algeria.

Bentivogli, L., & Pianta, E. (2000). Looking for lexical gaps. In Proceedings of the Ninth EURALEX
International Congress (pp. 1–6). Trento, Italy. Retrieved from http://dialnet.unirioja.es/servlet/
articulo?codigo=4616045.

Collins, H. (2006). CoBuild advanced learner’s English dictionary. Glasgow, Scotland:


HarperCollins.

Conner, D. (1983). Understanding semantics. Exeter: Exeter University Press.

Cui, Y., & Zhao, Y. (2014). Translation of rhetorical igures in the advertising discourse: A case
study. International Journal of Society, Culture & Language (IJSCL), 2(2), 57–67.

Darwish, A. (2010). Elements of translation. Melbourne: Writescope.

Daryabadi, A. (2007). The glorious Quran: Text translation and commentary. London: Islamic
Foundation.

Emara, S. (2014). Extraordinary vocabulary of the Qur’an and related translation problems.
International Journal of Linguistics, 5(1), 248–272. https://doi.org/10.5296/ijl.v5i1.3309.

Emara, S. A. E. (2013). Extraordinary vocabulary of the Qur’an and related translation problems.
International Journal of Linguistics, 5(1), 248–272. https://doi.org/10.5296/ijl.v5i1.3309.

Geeraerts, D. (2010). Theories of lexical semantics. New York: Oxford University Press Inc.

Ghazala, H. (2008). Translation as problems and solution. Beirut: Dar El-Ilm Lilmalayin.

Gibbs, R. (1994). The poetics of mind: Figurative thought, language, and understanding.
Cambridge: Cambridge University Press.

Ibn Ashour, M. (1984). Tafsir Attahrir Wattanwir [Liberation and enlightenment interpretation
book] ( irst). Tunis: Tunisian Publishing House.

Ilyas, A. I. (2013). Intra-textuality in translating some problematic Qur’anic verses. Arab World
English Journal, 2, 86–95.

Irving, T. (1985). The Noble Qur’an. Cedar Rapids: Iowa Publisher.

Jaber, I. (2010). Translating the genre of Quran: The challenge of translating the inimitable. Journal
of College Education for Women, 21(4), 943–954.

Uploaded by S. M. Safi
Johnson, K. (2001). What VP ellipsis can do, and what it can’t, but not why. In The handbook of
contemporary syntactic theory (pp. 439–479). https://doi.org/10.1111/b.9781405102537.2003.
00016.x.

Kabali, S. (2006). Semantics of metaphor : An overview of Majaz interpretation in the Holy Qur’an
(Unpublished PhD thesis). Islamic University, Malaysia. Retrieved from http://lib.iium.edu.my/
mom2/cm/content/view/view.

Khalifa, M. (1989). The sublime Qur’an and orientalism (2nd ed.). Karachi: International Islamic
Publisher (Pvt.) Ltd.

Khorami, M. (2014). Eloquence of repetition in Quran and Arabic old poetry. Language Related
Research, 5(2), 90–110.

Klepousniotou, E. (2002). The processing of lexical ambiguity: Homonymy and polysemy in the
mental lexicon. Brain and Language, 81(1–3), 205–223. https://doi.org/10.1006/brln.2001.2518.

Klepousniotou, E., & Baum, S. R. (2005). Processing homonymy and polysemy: Effects of
sentential context and time-course following unilateral brain damage. Brain and Language, 95(3),
365–382. https://doi.org/10.1016/j.bandl.2005.03.001.

Kövecses, Z. (2010). Metaphor: A practical introduction. English Linguistics/Journal of the English


Linguistic Society of Japan, 22(1), 205–231.

Lobner, S. (2002). Understanding semantics. Sonipat, India: Replika Press.

Lyons, J. (1977). Semantics. Cambridge: Cambridge Scholars Publishing.

Newmark, P. (1981). Approaches to translation. Oxford: Pergamon.

Newmark, P. (1988). A textbook of translation: Text. Hertfordshire: Pearson Education Limited.

Nida, E., & Taber, C. (1982). The theory and practice of translation (2nd ed.). Leiden: E.J. Brill.

Nugroho, A. B. (1999). Meaning and translation. Journal of English and Education, 2(3), 94–112.

Palmer, F. (1981). Semantics: A new outline. London and New York: Cambridge University Press.

Pickthall, M. (2001). The meaning of the glorious Qur’an. Kuala Lumpur: Islamic Book Trust.

Quinn, A. (1993). Figures of speech: 60 ways to turn a phrase. Utah: Gibbs Smith Publisher.

Rasekh, A. E., Dastjerdi, H., & Bassir, A. (2012). On homonymous expressions in the Qur’an: A case
study of the English translations of the term (fasad). The Journal of International Social Research,
5(22), 136–148.

Sadiq, S. (2008). Some semantic, stylistic and cultural problems of translation with special
reference to translating the glorious Qur ’ ân. Sayyab Translation Journal, 1(1), 37–59.

Said Ghazala, H. (2012). Translating the metaphor: A cognitive stylistic conceptualization (English
– Arabic). World Journal of English Language, 2(4), 57–68. https://doi.org/10.5430/wjel.
v2n4p57.

Uploaded by S. M. Safi
Semino, E. (2008). Metaphor in discourse. Cambridge: Cambridge University Press.

Shakir, M. (1999). The Holy Quran. New York: Tahrike Tarsile Quran.

Shehab, E. (2009). The Problems involved in translating Arabic cognitive synonyms into English.
Majallat Al-Jaam ’Ah Al-Islamiyyah, 17(1), 869–890.

Shunnaq, A. (1992). Functional repetition in Arabic realized through the use of word-strings with
reference to Arabic-English translation of political discourse. NouveltesDe La Fit-Newsletter, 2(1),
5–39.

Simpson, G. B. (1981). Meaning dominance and semantic context in the processing of lexical
ambiguity. Journal of Verbal Learning and Verbal Behavior, 20(1), 120–136.

Stanojević, M. (2009). Cognitive synonymy: A general overview. Facta Universitatis–Linguistics


and Literature, 7(2), 193–200.

Xu, X. (2008). The rhetoric and translation of English advertisement. International Journal of
Business and Management, 3(11), 83–86.

Zaky, M. M. (2001). Translation and language varieties. The Translation Journal, 5(3). Retrieved
from http://translationjournal.net/journal/17theory.htm.

Uploaded by S. M. Safi
© The Author(s) 2020
N. Abdelaal, Translation between English and Arabic
https://doi.org/10.1007/978-3-030-34332-3_5

5. Culture as a Problem in Translation


Noureldin Abdelaal1
(1) University of Nizwa, Nizwa, Oman

Noureldin Abdelaal

Overview
This chapter discusses the problem related to culture-bound terms,
and how to deal with them.
The chapter covers the following topics:
1.
Culturally bound terms;
2.
Strategies to translate culturally bound terms;
3.
Fixed expressions.

Malinowski, a famous anthropologist, was one of the irst


anthropologists to indicate that language can only be understood with
reference to culture (Katan, 1993). In 1923, he coined the term ‘context
of situation’; by this concept, he meant that language could be only
understood with reference to culture and situation. If culture and
situation are clear for interlocutors, language can be understood
(Katan, 1993). Delisle (1988) mentions that one of the merits of
translation is that it relates two cultures to each other, in terms of
thought and perception. Culture, as de ined by Newmark (1988, p. 94)
is a ‘Way of life and its manifestations peculiar to one speech

Uploaded by S. M. Safi
community’. Differences in culture are more problematic in translation
than differences in language structure (Nida, 2000). Nida mentions
three categories of relatedness between language and culture: the irst
category is when the distance between the source and receptor codes
is limited linguistically and culturally, as with the relatedness between
English and French; translating from Hebrew to Arabic is another
example. In this category, problems of translation will occur least
frequently; however, in this category of languages a translator should
not be deceived by cognates, such as that between the English word
‘demand’ and the French word ‘demander’ which may be super icially
thought to be identical even though their meanings are not. The second
type of relatedness between languages is when the SL and the TL are
culturally related but linguistically different; an example of this is
translating from German to Hungarian. This type of relation is less
problematic than the third type, in which the differences are due to
culture. Relatedness between English and Arabic falls into the third
category, as they are quite different languages linguistically and
culturally. This prompts many problem in the translation process that
need to be handled carefully. Thus, deep knowledge of the target culture
is a necessity for a successful translation due to there being a
considerable mismatch between English culture and Arabic culture in
terms of beliefs, customs and traditions (Mares, 2012).
One main component of culture is language and its vocabulary.
Vocabularies attain their meanings from the culture to which they
belong; and, since Arabic is different from English, mastering Arabic
culture is essential for the production of good translation. What could
cause problems for translators may include the religious facts of Arabic
societies, even their names, which have religious signi icance (Mares,
2012). Cultural ambiguity is identi ied by Newmark (1988) as one of
the seven ambiguities of translation. Such ambiguity may result from a
gap in translation. This gap could be grammatical, lexical, or linguistic.
Differences between languages in terms of cultures create what are
referred to as ‘culturally bound’ terms.
Culturally bound terms are particular cultural elements that are
bound to each speci ic language. According to Harvey (2000, p. 2)
‘culture-bound terms refer to concepts, institutions and personnel
which are speci ic to the SL culture’. According to Ordudari (2007),

Uploaded by S. M. Safi
translating culturally bound elements in general, and allusions in
particular, seems to be one of the most challenging tasks to be
performed by a translator. In the following section, culturally bound
terms and their role in creating lexical gaps will be discussed.

5.1 Culturally Bound Terms


Language and culture are part and parcel of each other; they cannot be
separated because they are interwoven. They have a homologous
relationship. To put it another way, language marks cultural identity, or
we could say that language is culture and, thus, translating a language
implies translating a culture. However, culture is complex because it
implies a fuzzy set of attitudes, behavioural conventions, and basic
assumptions and values that are shared by set of people (Spencer-
Oatey, 2000). Furthermore, when the SLs and TLs belong to different
cultural groups, it is truly dif icult to ind terms in the TL that express
the highest level of accuracy possible to the meaning of certain words
(Haque, 2012). Connotations and associations of words in one
language may differ from those in another language, or they may have
different emotive associations. Cultural and social differences affect
the process of translation and make it challenging (Al-Shawi, 2012). Put
simply, there are some words or expressions, especially those that
have a religious context, that are culturally bound terms; they do not
have equivalents in the TL. Some Arabic words—such as ‫ﺧﻠﻮة‬
/khulwah/, ‫ ﻋﻘﯿﻘﺔ‬/aqeeqah/, or ‫ رﺣﻢ ﻗﻄﯿﻌﺔ‬/qatiat rahem/ do not have
equivalents in English. They are culturally bound terms (Bahameed as
cite in Al-Haj, 2014). ‘Culturally bound’ is a broad term that includes a
wide range of expressions such as idioms, collocations and ixed
expressions. However, in this section I use ‘culture-bound terms’ to
refer to words that do not have equivalents in the TL due to cultural
speci icity. The lack of equivalents for such terms creates lexical gaps,
and they can include exotic or emotive expressions. Thus, the only
solution available to a translator is to use transliteration, or to render
such terms through periphrastic translation (Abdul-Raof, 2004; Al-
Azab & Al-Misned, 2012). Examples of culture-bound terms are
‘girlfriend’ and ‘boyfriend’, which can be translated as ‫ﻋﺸﯿﻘﺔ و ﻣﻌﺸﻮق‬,
although they are not complete equivalents of the ST terms. The Arabic

Uploaded by S. M. Safi
terms are used pejoratively to refer to an illegal relationship between a
man and a woman. However, the English words seem to be used
positively to refer to an acceptable and legal relationship (perhaps
close to marriage) between two partners. In the Holy Quran, for
example, it is dif icult to ind equivalents for words such as: ‫اﻟﺼﻤﺪ‬
/assamad/ or ‫ اﻷﺧﻼص‬/alikhlaas/. Another consideration that makes
translating these lexicons dif icult is that they are pregnant with
meaning. That is, even when seeking a periphrastic translation, the
translation falls short of providing a full explanation of the denotative
and connotative meanings of a word. It may be more meaningful to
provide a paraphrased and periphrastic translation simultaneously. In
short, some culture-bound terms can be translated using an equivalent
that has been created in the TL that is considered to be equivalent to
the SL item, or by using other translation strategies such as
‘paraphrasing’, ‘transliteration’, using a ‘functional equivalent’, or
‘borrowing’.

Exercise
Translate the following terms between English and Arabic.
1.
sel ie
2.
fuck
3.
bitch
4.
heavy-smoker
5.
the White House
6.
the Iranian White Revolution
‫ﺳﻬﺮة‬
‫ﻃﺮب‬
‫ﺣﻘﺪ‬
‫ﻋﺪة‬

Uploaded by S. M. Safi
‫اﻟﻨﻘﺎب‬
‫اﻟﻠﻘﯿﻂ‬

5.2 Idioms and Fixed Expressions as a


Problem in Translation
Idioms—culture-bound expressions—do not function as single units,
comprising as they do of multiple words. They are dif icult to de ine or
describe in exact terms. Richards & Schmidt (2002) de ine an idiom, as
‘an expression which functions as a single unit and whose meaning
cannot be worked out from its separate parts’ (p. 246). For example,
take the idiom ‘a little bird told me that’. This expression means ‘I have
discovered this piece of information in my own way’ and, since nothing
further is said, implies an unwillingness to reveal the source if there is
one, or a decision that the source is to remain a mystery. The little bird
cannot be perceived literally to have been the agent. Hence, the
meaning of the idiom has nothing to do with the separate lexemes of
‘bird’, or ‘tell’. However, one view holds that an idiom usually starts as a
phrase that has a literal meaning and which then is used in a igurative
way. In other words, this view holds that there is a relationship
between the components of idioms and their idiomatic or igurative
meaning (Al-Haddad, 1994). Baker (2011) postulates that idioms are
frozen strings of language whose meanings cannot be deduced from
their individual components. Consider the idiom ‘fed up’. Animals and
people can be fed, and there could be other usages involving ‘feed’ or
‘fed’. However, when it comes to being fed up, ‘up’ must follow ‘fed’. It
has to be in the past tense ‘fed up’; one cannot say ‘feed up’, using the
present tense of ‘feed’, as this combination makes no sense.
Fernando (1996, p. 3) states that there are three features that
characterize idioms: compositeness, institutionalization and semantic
opacity. Compositeness refers to the nature of idioms that comprise of
more than one word (i.e. multi-word expressions). Institutionalization
implies that idioms are conventionalized expressions that are a
product of an ‘ad hoc’ situation, or serve certain purposes. Semantic
opacity refers to igurative or non-literal features of idioms, in the

Uploaded by S. M. Safi
sense that meanings of idioms are not the mere sum of their literal
parts. Fernando (1996) adds that these characteristics are not only
characteristics of idioms, but are also shared by other multi-word
expressions, such as collocations, proverbs and idioms.
Mäntylä (2004) argued that ive features were always (i.e.
traditionally) used to characterize idioms. These classical or
traditional features are metaphoricity or igurativeness, analysability
or non-compositionality, ixedness of form, the level of formality, and
being multi-word expressions. Metaphoricity is deemed the most
principal feature of idioms. Non-compositionality indicates that
idioms are dead, whereby their meanings are arbitrary and not
igurative. Fixedness of form signi ies the intolerant syntactic nature of
idioms, as they are frozen. The level of formality is related to the fact
that idioms are considered to belong to informal, spoken language
rather than to formal, written language. Finally, idioms are composed
of more than one word and, thus, they are multi-word expressions.
However, Mäntylä (2004, p. 28) argues against these classical
features of idioms. He believes that idioms are not merely dead, frozen
metaphors, as there are certain idioms that are neither dead nor
frozen. There is much literature that is in line with Mäntylä, and that
rejects the idea that idioms are dead and frozen metaphors (e.g. Gibbs
et al., 1989; Glucksberg, 1993). For example, Gibbs (1993, p. 58) states
that the assumption of some scholars that idioms are dead metaphors
is far from being accurate. He adds that the arbitrary conventions of
usage may determine idioms’ meanings. Take, for example, the idiom
of ‘break a leg’, which means to wish a good luck before a theatrical
performance. This idiom originated from the old superstition that
wishing good luck to someone would be bad luck; hence, over time,
people started to use it and it became ixed as a convention. However,
Mäntylä (2004, p. 29) adds that detecting the link between the origins
of an idiom and its meaning is dif icult because the interrelationship
weakens over time.
As such, the translation of idioms poses many challenges for a
translator whose job necessarily goes beyond merely translating
lexical items from an SL into a TL. It is a process of translating the style
of language, and therefore the culture, of an SL into a TL. Consequently,
it is essential that a translator be cognizant of the cultural variances

Uploaded by S. M. Safi
and the various strategies of discourse in the SL and TL to achieve
optimal accuracy. The hidden structure of the source text should be
analysed through the use of various strategies of discourse by the
translator (Razmjou, 2004). Aldahesh (2017) argues that the semantic,
syntactic and pragmatic complexities of the ST idioms make
translating them a challenging task.
Baker (1992/2011) postulates that the problems in translating
idioms are prompted by a lack of two skills: the ability to recognize and
interpret an idiom correctly, and knowledge of how to render the
various aspects of the meaning of the ST idiom into the TL.
Recognizing an expression as an idiom may not be as easy a task as it
may seem. Translators sometimes fall in the trap of perceiving an
idiom literally, and are thus unable to recognize it as an idiom. While
some idioms lend themselves to literal rendering, others do not. ‘Birds
of a feather lock together’ is an example of an idiom that can be
translated literally and make perfect sense, and yet be understood in a
way in which it perhaps was not intended. Literally, this statement is
true. Sparrows are with sparrows; robins are with other robins; crows
with other crows: various bird species do not mix. However, this
statement is normally intended to describe humans and, these days,
has more to do with the character of a person (whether they are like-
minded or alike in nature) than race or ethnicity, although this idiom
can be used in this way as well. Also, with the idiom ‘to kill two birds
with one stone’, it is possible that a literal application could actually
happen. However, the English idiom, ‘got my goat’ (to get a person’s
goat means to irritate them as in: ‘He’s got my goat’, cannot be
translated literally.
Baker believes that the more dif icult an expression is to
understand in speci ic contexts, the more likely it will be recognized as
an idiom by a translator. For example, the expression ‘Put your money
where your mouth is’ surely must mystify TL readers, and the
translator would realize that this saying is not meant to be taken
literally. However, Baker mentioned that there are certain instances
where idioms can be misleading for a translator. Some idioms can be
interpreted literally and thus a translator may render some kind of
vague, poorly understood meaning—the meaning that is obvious
resulting from the simplest word-to-word direct translation. But the

Uploaded by S. M. Safi
problem is that these idioms may have a different meaning from the
literal meaning of the words. A case in point is the idiomatic
expression ‘go out with’, the meaning of which, if translated literally,
will not correspond with the ST meaning in some contexts. People who
are just starting to date are said to be ‘going on a date’. If they continue
past the irst date, people will ask ‘Are you going to see him/her
again?’, or ‘Are you going to go out with him/her again?’, or ‘Are you
going to go on another date with him/her again?’ Or someone will say
‘They’re going on another date.’ If they continue to date, they’re
‘checking each other out’ (‘seeing how it goes’ and ‘where’ or ‘how far it
goes’). When the dating continues and becomes a pattern, then people
will say ‘they’re dating’. When the two people become a couple, then the
terms ‘boyfriend’ and ‘girlfriend’ will be used. Or people may describe
the couple as ‘being an item’.
Another problem with idioms is that ‘An idiom in the source
language may have a very close counterpart in the target language
which looks similar on the surface but has a totally or partially different
meaning’ (Baker, 2011, p. 70). Baker gives the example of ‘pull his leg’,
which has an equivalent idiom in Arabic: yashab rijluh. (This means
quite literally ‘pull his leg’) However, the Arabic and English idioms
have different meanings. The English idiom means to tease somebody
by misinforming them, and then tell the truth. It means ‘to deceive
someone playfully’; maybe people may tell the truth, if need be, but that
is more after the fact than part of the de inition. For example, an uncle
tells his niece ‘The sun is going to rise and set in the east tomorrow.’
The young niece replies: ‘Really?’ And the uncle says, ‘Nah, I was just
pulling your leg.’ Another usual reply in such a circumstance is ‘Nah, I
was just teasing you.’ In contrast, the Arabic idiom means to deceive
and trick somebody purposefully. Thus, the Arabic version is meant to
be a real deception (not teasing someone) and is, therefore, more
sinister in nature.
According to Baker (2011), there are some challenges in translating
idioms that have nothing to do with the nature of idioms. In other
words, these dif iculties are faced in translating opaque as well as
transparent idioms. These problems are: lack of equivalence, an idiom
in the SL may have simultaneous idiomatic and literal senses, and the
use of idioms in written discourse.

Uploaded by S. M. Safi
1. Lack of equivalence: Some idioms are culture-speci ic and,
therefore, they do not have equivalents in the TL. A case in point
would be ‘Yours faithfully’, which does not have an equivalent
greeting in Arabic. A translator then has to translate it as ‫و ﺗﻔﻀﻠﻮا‬
‫( ﺑﻘﺒﻮل ﻓﺎﺋﻖ اﻷﺣﺘﺮام‬which means ‘Please accept the utmost respect’), or
any other common greeting in Arabic. Therefore, as proposed by
Fenyo, knowledge of the source and target cultures is proximal,
premium and a prerequisite to proper translation. A culture-
speci ic idiom is not necessarily untranslatable. For example, the
English idiom ‘to carry coals to Newcastle’ means ‘something
brought or sent to a place where it is already plentiful’. So, this
means that the action was useless because the material or item
was not needed, or, ‘it is best sent where it can be sold or used’. This
idiom can be translated into Arabic as ‫‘( ﯾﺒﯿﻊ ﻣﯿﺎه ﻓﻲ ﺣﺎرة اﻟﺴﻘﺎﯾﯿﻦ‬to sell
water in the district of water sellers’). The idiom in the SL may
have a similar counterpart in the target language, but the
connotations are different and they may not be pragmatically
transferable.
2.
An idiom in the SL may have idiomatic and literal senses at the
same time that are not represented (at either the literal or
idiomatic level) in many languages. For example, the play on
meaning that exists with the English expression ‘to poke your nose
into something’ is not represented at either the literal or idiomatic
levels in many languages. The play on meaning in this idiom is
different. In English, it is possible to ‘poke your nose into
something’, usually by accident. Perhaps you got whipped cream or
ice-cream on your nose when you were eating a desert, or perhaps
your nose got too close to a lower you smelled. An example of a
literal meaning: ‘I poked my nose in (into) that lower and now it
feels itchy, like I have to sneeze.’ An example of a igurative meaning
is: ‘John is always poking his nose into other people’s business’,
meaning that John pries into other people’s personal affairs.
Another example is ‘to kick the bucket’, which can literally mean ‘to
kick a bucket (of water)’, or idiomatically mean ‘to die’. In Arabic,
however, the idiomatic meaning does not exist.
3. The use of idioms in written discourse, the contexts in which they
can be used and the frequency of their use may be different in the

Uploaded by S. M. Safi
can be used, and the frequency of their use may be different in the
SL and TL. For example, English makes considerable use of idioms
in written discourse, which is not the case in Arabic.
As discussed, translating idioms is challenging and translators
develop their own ways of dealing with it.
Baker (2011) posits that there are some useful strategies that can
be followed when translating idioms. These strategies are: using an
idiom of similar meaning and form, using an idiom of similar meaning
but dissimilar form, borrowing the SL idiom, translation by
paraphrase, translation by omission of the play on the meaning of the
idiom, translation by omission of the entire idiom, and compensation.
1.
Using an idiom of similar meaning and form: This involves using an
idiom in the TL that has roughly the same meaning as the SL idiom
and consists of equivalent lexical items. For example, the Arabic
idiom ‫ رأﺳﺎ ﻋﻠﻰ ﻋﻘﺐ‬has an equivalent in English: ‘head over heels’. So,
the Arabic idiom, when translated, consists of the exact same
number of words and has the same meaning.
2.
Using an idiom of similar meaning but dissimilar form: This
involves using an idiom in the TL that has roughly the same
meaning as the SL. However, it does not have equivalent lexical
items. For instance, the English idiom ‘let bygones be bygones’ is
similar in meaning to ‫اﻟﻠﻲ ﻓﺎت ﻣﺎت‬.
.3
Borrowing the source language idiom: Sometimes, borrowing the
SL idiom can be a way to translate culture-speci ic items. For
example, the idiom ‘get out of my hand’ is sometimes translated
literally into Arabic as ‫ﺧﺮﺟﺖ ﻣﻦ ﯾﺪي‬.
4.
Translation by paraphrase: Another common strategy of
translation is translation by paraphrase, whereby a translator
paraphrases the SL idiom. An example of this is the English idiom
‘a bird in the hand’, which can be translated as ‫‘( ﯾﻐﺘﻨﻢ اﻟﻔﺮﺻﺔ‬seize the
opportunity’).
5. Translation by omission of a play on idiom: ‘This strategy involves
rendering only the literal meaning of an idiom in a context that
f f f f

Uploaded by S. M. Safi
allows for a concrete reading of an otherwise playful use of
language’ (Baker, 2011, p. 84). For example, translating ‘to burn his

boats’ as ‫ﯾﺤﺮق ﻣﺮاﻛﺒﻪ‬. The translation sounds literal but the


idiomatic meaning is still conveyed.
6.
Translation by omission of entire idiom: This strategy of omitting a
whole ST idiom in the TT is followed either when there is no
equivalent for the ST idiom and it therefore cannot be paraphrased,
or because of stylistic reasons.
7.
Compensation: A translator sometimes seeks to omit or play down
some features of an idiom in the ST that occurs at a speci ic point
in the text and present it somewhere else in the TT.

Exercise
Translate the following idioms into English. Then explain the
strategy used in the translation.
‫اﻟﻘﺮد ﻓﻲ ﻋﯿﻦ أﻣﻪ ﻏﺰال‬
‫اﻟﺒﺎب ﯾﻔﻮت ﺟﻤﻞ‬
‫ﻻ ﺣﻮل ﻟﻪ و ﻻ ﻗﻮة‬
‫ﻓﺎر دﻣﻲ ﺑﺴﺒﺐ ﻣﺎ ﻗﺎﻟﻪ‬
‫إﻧﻪ ﻃﻮﯾﻞ اﻟﻠﺴﺎن‬
‫ﻃﺎر ﻋﻘﻠﻪ‬
‫ﺳﻤﻦ ﻏﻠﻰ ﻋﺴﻞ‬
‫ﺛﻘﯿﻞ اﻟﺪم‬
‫ﺧﻔﯿﻒ اﻟﺪم‬
‫ﺑﻨﺖ اﻟﺤﻼل‬
Translate the following English idioms into Arabic. Then, explain
the strategy used in the translation.
1.
‘Got ahead of the game’
2.
‘A leap in the dark’

Uploaded by S. M. Safi
3. ‘A yes-man’

4.
‘All cats are black in the dark’
5.
‘Every cloud has a silver lining’
6.
‘Get a taste of your own medicine’
7.
‘Beat around the bush’
8.
‘Give someone the bene it of the doubt’
9.
‘Pull yourself together’
10.
‘A picture is worth 1000 words’
11.
‘Do something at the drop of a hat.’

5.3 Collocations
Collocations are sometimes culturally bound, as they do not apply to all
languages. Baker (2011, p. 14) de ines collocations as ‘semantically
arbitrary restrictions which do not follow logically from the
propositional meaning of a word’. Put more simply, collocations refer to
the habitual occurrence of words together. For example, we say ‘make
love’, but we do not say ‘do love’. Similarly, ‘bus’ collocates with ‘catch’,
‘miss’, ‘ticket’, ‘by’, ‘on’. So, these restrictions are arbitrary. We can say
‘catch the bus’, but we cannot say ‘hold the bus’. Also, we say ‘rancid
butter’ but ‘rotten eggs’; it cannot work the other way around even
though ‘rancid’ and ‘rotten’ are synonymous. When words collocate,
they may give meaning that is different in the SL; therefore, a different
word may be required as an equivalent in the TL. For example, ‘bend’
means‫ﯾﺜﻨﻲ‬. However, when it collocates with different words, it will give

Uploaded by S. M. Safi
different meanings in the TL. Consider the following examples
(Table 5.1).
Table 5.1 The collocations of ‘bend’

English Arabic
1. ‘Bend his head’ ‫ﯾﺤﻨﻲ رأﺳﻪ‬
2. ‘Bend the law’ ‫ﯾﻠﻮي ﻋﻨﻖ اﻟﻘﺎﻧﻮن‬
3. ‘Bend his leg’ ‫ﯾﺜﻨﻲ ﻗﺪﻣﻪ‬
4. ‘Bend over the sink’ ‫ﯾﻨﻜﺐ ﻋﻠﻰ ﺣﻮض اﻟﻐﺴﯿﻞ‬

As can be seen in these examples, ‘bend’ is translated differently


based on the word with which it collocated. In example 2, it was
translated as two words. Translating collocations, therefore, is not
without its problems. There are many pitfalls and dif iculties in
translating collocations that will be discussed in detail: the engrossing
effect of ST patterning, misinterpreting the meaning of an SL
collocation, the tension between accuracy and naturalness, culture-
speci ic collocations, and marked collocations in the source text.
1.
The engrossing effect of ST patterning
Sometimes a translator becomes engrossed in the ST and produces
an odd collocation in the TL—perhaps by translating an ST collocation
literally. In this situation, a translator needs to detach himself from the
ST patterning by leaving the translation for some time, subsequently
revising it to ensure that it matches with the TL patterning. For
example, a novice translator may translate ‘break the law’ as ‫ﯾﻜﺴﺮ اﻟﻘﺎﻧﻮن‬,
though a more natural translation would be ‫ﯾﺨﺎﻟﻒ اﻟﻘﺎﻧﻮن‬.
2.
Misinterpreting the meaning of an SL collocation
A translator may misinterpret the meaning of an SL collocation
when the SL collocation and the TL collocation are similar in form but
not in meaning. Baker (2011) gives the following example:

Example

Uploaded by S. M. Safi
ST: The industrialist had been struck by his appearance as
someone with modest means.
TT: ‫رﺟﻞ اﻟﺼﻨﺎﻋﺔ ﻣﻈﻬﺮه ﯾﻨﻢ ﻋﻦ اﻟﺘﻮاﺿﻎ و اﻟﺒﺴﺎﻃﺔ‬

In this example, the ST idiom ‘with modest means’ was translated as


‫‘( اﻟﺘﻮاﺿﻎ و اﻟﺒﺴﺎﻃﺔ‬modesty and simplicity’), which shows the in luence of
the TL on the translator’s decision to render the idiom incorrectly. It
should have been translated it as ‫‘( ﻏﯿﺮ ﺛﺮي او ذو دﺧﻞ ﻣﺤﺪود‬poor/limited
income’).
3.
The tension between accuracy and naturalness
Tension is created when a translator is forced to prioritize either
accuracy or naturalness. It is dif icult to maintain both. Baker gives the
example of ‘law’, which can be ‘bad’ or ‘good’. However, a natural
translation for ‘bad/good law’ would be ‫ﻏﯿﺮ ﻋﺎدل‬/‫‘( ﻗﺎﻧﻮن ﻋﺎدل‬fair/unfair
law’), which may be signi icantly different from the SL collocation.
Another striking example that was given by Baker is ‘hard drinks’,
which is likely to be translated into ‘alcoholic drinks’. However, the
collocation ‘hard drinks’ does not include all alcoholic drinks, it only
includes spirits such as whisky, gin and brandy; it does not include
other alcoholic drinks such as beer. In short, a translator needs either
to prioritize accuracy and therefore translate ‘hard drinks’ as ‫ﻣﺸﺮوﺑﺎت‬
‫ ﺛﻘﯿﻠﺔ‬or prioritize naturalness and therefore translate ‘hard drinks’ as
‫ﻣﺸﺮوﺑﺎت ﻛﺤﻮﻟﯿﺔ‬.
4.
Culture-speci ic collocations
There are some collocations that are language-speci ic; they do not
sound natural in other languages. Baker considers that these terms
need to be over-translated in the TL, as more information needs to be
provided in the TL to clarify the ST collocation. Baker gives the
example of the English collocation ‘damaged, dry, and brittle hair’,
which was translated into Arabic as ‫ و أﯾﻀﺎ‬،‫ اﻟﻤﺘﺄذي أو اﻟﺘﺎﻟﻒ‬،‫اﻟﺸﻌﺮ اﻟﻤﻘﺼﻒ‬
‫ﻟﻠﺸﻌﺮ اﻟﺠﺎف أو اﻟﻀﻌﯿﻒ اﻟﺒﻨﯿﺔ او اﻟﻘﺎﺑﻞ ﻟﻠﺘﻜﺴﺮ‬. The ST collocations are culture
speci ic as hair in English can be ‘dry’, ‘damaged’, or ‘brittle’; however, in

Uploaded by S. M. Safi
Arabic it can be ‫ ﻣﺘﻘﻀﻒ‬،‫ ﻧﺎﻋﻢ‬،‫‘( ﺧﺸﻦ‬split-ends’, ‘dry’, ‘oily’, ‘coarse’, and
‘smooth’). In the translation quoted by Baker, the translator opted to
render the ST collocations unnaturally in the Arabic language, which it
could be argued was improper. This is similar to the problem
mentioned earlier regarding the tension between ‘accuracy’ and
‘naturalness’. In this example, the translator prioritized accuracy over
naturalness. It is always the translator’s decision to adopt the
appropriate translation strategy. Sometimes, what is assumed to be
‘accurate’ translation may turn out to be inaccurate, as a literal
translation of an ST term does not necessarily convey accuracy when
translated.
5.
Marked collocations in the source text
Marked collocations are images that are created in the SL and
translating these collocations may be marked in the TL. For example,
‘the sun sank’ as used by John Steinbeck, the Nobel laureate, in his
novel The Red Pony. The writer in this case may ind himself translating
it literally as ‫ ﻏﺮﻗﺖ اﻟﺸﻤﺲ‬to create a similar unmarked collocation in the
TL.

5.4 Strategies to Translate Culturally Bound


Terms
Venuti (1995) argues that translating culture can be approached from
two perspectives: one view holds that the source culture should be
preserved in the TT by following strategies that preserve the ST
elements and providing explanation to cultural items when necessary.
This kind of translation is referred to as ‘exotocized’. Literal translation
can be adopted to translate ST cultural idioms if a translator wishes to
preserve the ST culture. The other perspective is ‘domestication’, which
attempts to render the ST elements into functionally equivalent
elements in the TT. To illustrate the difference between the two
approaches, let us give an example of the ST idiom ‘A bird in the hand is
worth two in the bush’, which can be translated as ‫ﻋﺼﻔﻮر ﻓﻲ اﻟﯿﺪ ﺧﯿﺮ ﻣﻦ‬
‫أﺛﻨﯿﻦ ﻋﻠﻰ اﻟﻐﺼﻦ‬. This is an ST oriented translation that preserved all the

Uploaded by S. M. Safi
elements of the ST. However, if we wished to translate it into a
culturally equivalent idiom, we may translate it as ‫ﻋﺼﻔﻮر ﻓﻲ اﻟﯿﺪ ﺧﯿﺮ ﻣﻦ‬
‫ﻋﺸﺮة ﻋﻠﻰ اﻟﺸﺠﺮة‬. Actually, the two translations seem close to each other;
the only difference is that the word ‘bush’ ‫ اﻟﻐﻀﻦ‬was translated into
‫( اﻟﺸﺠﺮة‬tree) to adapt the TT culture. A more striking example would be
as follows:

Example
ST: Hold your horses; we still have plenty of time.
TT1: ‫ ﻟﺪﯾﻨﺎ اﻟﻜﺜﯿﺮ ﻣﻦ اﻟﻮﻗﺖ‬،‫أﻣﺴﻚ ﺧﯿﻮﻟﻚ‬.
TT2: ‫ﺗﺮﯾﺚ ﻓﻠﺪﯾﻨﺎ وﻗﺖ ﻛﺎﻓﻲ‬.

As can be seen, TT1 observes the ST cultural norms and values.


However, the translation may sound ambiguous and unclear. On the
other hand, TT2 is a TT equivalent expression that conveys the same
meaning but without the stylistic and idiomatic effect that exists in the
ST.
Graedler in Ordudari (2007) sets rules to translate culture, which
include: creating a new word, explanation, preserving the SL term
intact, and opting for an alternative word from the TL.
a.
Creating a new word: This can be achieved by transliteration or
borrowing. For example, the words ‫ ﻣﺴﻠﻢ‬،‫ ﺟﻬﺎد‬،‫ ﺣﺞ‬are all borrowed
from Arabic to English. Also, the English words ‘supermarket’ and
‘toilet’ are borrowed from English to Arabic.
b.
Explanation: A translator may need to explain the term through
glossing or a footnote. For example, the word ‫ ﺧﻠﻮة‬may be
transliterated and then explained as ‘being alone with a foreigner’,
which is different from ‫اﻟﺨﻠﻮة اﻟﺸﺮﻋﯿﺔ‬, which means to have ‘the right
to stay alone with one’s own wife’.
c.
Preserving the SL term intact: This can be achieved through
borrowing (as discussed with regard to creating a new word.
d. Opting for a word in the TL that seems similar to, or has the same
‘relevance’ as, the SL word. For example, translating ‫ ﺻﻼة‬as ‘prayer’.

Uploaded by S. M. Safi
In a similar vein, Harvey (2000) proposed four techniques for
translating culturally bound terms: functional equivalence, formal
equivalence, transcription or borrowing, and descriptive or self-
explanatory translation.
1.
Functional equivalence: This refers to rendering a referent in the
SL culture into a functionally equivalent referent in the TL. For
example, translating ‫ ﻋﻠﻰ أﺣﺮ ﻣﻦ اﻟﺠﻤﺮ‬as ‘on pins and needles’.
Another example is translating ‘For he’s a jolly good fellow for he’s
a jolly good fellow’ as ‫ﺳﻨﺔ ﺣﻠﻮة ﯾﺎ ﺟﻤﯿﻞ ﺳﻨﺔ ﺣﻠﻮة ﯾﺎ ﺟﻤﯿﻞ‬. Another example
is translating ‘kick the bucket’ as ‫ﻟﻘﻰ ﺣﺘﻔﻪ او ﻣﺼﺮﻋﻪ‬.
2.
Formal equivalence (or ‘linguistic equivalence’) refers to a word-
for-word translation. This type of literal translation can be
appropriate in some contexts.
3.
Transcription or borrowing refers to reproducing or, where
necessary, transliterating the original term.
4.
Descriptive or self-explanatory translation refers to using generic
terms in lieu of the culturally bound terms to convey the meaning;
for example, translating ‫اﺑﻦ ﻟﺒﻮن‬, which refers to two-year old
camels, as a ‘camel’ to convey the generic meaning.

Exercise
Translate the following collocations into Arabic.
1.
‘Peaceful death’
2.
‘Pass a law’
3.
‘Pay a visit’
4.
‘Brain death’
5. ‘The throes of death’

Uploaded by S. M. Safi
6.
‘Wonderments and bewilderments’
7.
‘Hale and hearty’
8.
‘Wealthy and well’
9.
‘On the alert’
10.
‘The great mountains’

Exercise
Examine the following ST collocations and their translations, and
then explain the strategy used and the extent to which it was
effective in conveying the ST collocational meaning (Source Al
Sughair, 2011).

‘Rat trap’ ‫ﻓﺦ ﺟﺮذان‬


‘An impulse of cruelty’ ‫داﻓﻊ ﻗﺴﻮة‬
‘Resentful eyes’
‘Flying heavily’
‘Screen door’
‘He looked secretly’
‘To risk lives’
‘High-priced’
‘War drums’
‘Off-colour joke’
‘Tip of the tongue’
‘Delicate foods’
‘My hot cheeks’

Uploaded by S. M. Safi
5.5 Pym’s Typology of Translation Solutions
Pym (2018) proposed a typology for translation solutions; this
typology is assumed to be a comprehensive typology that helps deal
with the various problems faced when translating: copying words,
copying structure, perspective change, density change,
resegmentation, compensation, cultural correspondence and text
tailoring.
1.
Copying words: This is a kind of transcription, exoticism,
transliteration or transference; for example, translating ‘internet’
as ‫اﻧﺘﺮﻧﺖ‬.
2.
Copying structure: This is similar to the structural calque proposed
by Vinay and Darbelnet (1958) (see Vinay and Darbelnet’s model in
Chapter 2, for examples).
3.
Perspective change: This is similar to Vinay and Darbelnet’s
modulation; for example, translating ‘keep the door closed, please’
as ‫رﺟﺎءا ﻻ ﺗﻔﺘﺢ اﻟﺒﺎب‬.
4.
Density change: This can be applied by employing strategies that
help distribute the information across a greater textual space.
These strategies may be explicitation or one-to-many translation.
In other words, one word that is lexically dense or semantically
complex is rendered into many words using an explicitation or
paraphrase strategy.
5.
Resegmentation: This includes changing the order of sentences or
paragraphs; it also includes breaking down complex sentences or
joining simple sentences together.
6.
Compensation: This is when ‘A value is rendered with resources
different from those of the start text and in a textual position or
linguistic level that is markedly different from that in the start text
(Pym, 2018, p. 44). This may include notes, glossing, or similar.
7. Cultural correspondence: This happens when ‘corresponding
referents are held to be in different special or temporal locations,

Uploaded by S. M. Safi
referents are held to be in different special or temporal locations,
as opposed to cases where the same referent is given different
expressions but remains in the one location (p. 44). For example,
translating ‘hi’ as ‫اﻟﺴﻼم ﻋﻠﯿﻜﻢ‬. This covers what Vinay and Darbelnet
call ‘adaptation’ and ‘equivalence’.
8.
Text tailoring: This includes the deletion and addition of material
on the grammatical or semantic levels.

For pedagogical purposes, Pym condensed this typology into three


categories: copying, expression change and material change: these are
summarized in Table 5.2.
Table 5.2 Typology of translation solutions (Pym, 2018, p. 45)

Copying Copying words Copying sounds


Copying morphology
Copying script
Copying structure Copying prosodic features
Copying ixed phrases
Copying text structure
Expression Perspective change Changing sentence focus
change Changing semantic focus
Changing voice
Renaming an object
Density change Generalization/speci ication
Explicitation/implicitation
Multiple translation
Resegmentation Joining sentences
Cutting sentences
Re-paragraphing
Compensation New level of expression
New place in text (notes, paratexts)
Cultural Corresponding idioms
correspondence Corresponding units of measurement, currency,
etc.
Relocation of culture-speci ic referents
Material change Text tailoring Correction/censorship/updating

Uploaded by S. M. Safi
Omission of material
Addition of material

5.6 Conclusion
Translation is a necessity and the notion that a text cannot be
translated can be given no credence. Every type of text or genre is
translatable. A translator, however, needs to identify the appropriate
approach and strategies for the translation of a speci ic text.
Researchers also need to explore the mechanisms and procedures that
can improve the quality of translation, rather than focusing their
efforts on criticizing translations and translators. Based on the
situation a translator faces, they can develop appropriate strategies to
deal with emerging problems.

References
Abdul-Raof, H. (2004). The Quran: Limits of translatability. In S. Faiq (Ed.), Cultural encounters in
translation from arabic. Frankfurt Lodge: Multilingual Matters.

Aldahesh, A. Y. (2017). On idiomaticity in English and Arabic: A cross—linguistic study. Journal of


Language and Culture, 4(2), 23–29. https://doi.org/10.5897/JLC2013.0220.

Al Sughair, Y. (2011). The translation of lexical collocations in literary texts, The American
University of Sharjah.

Al-Azab, A., & Al-Misned, A. (2012). Pragmatic losses of Qur’an translation: A linguistic approach.
English Language and Literature Studies, 2(3), 42–49. https://doi.org/10.5539/ells.v2n3p42.

AL-Haddad, K. (1994). Investigating dif iculties faced by advanced Iraqi students of English in and
using English idioms (Unpublished MA thesis). College of Arts, University of Baghdad.

Al-Haj, A. A. M. (2014). The Cultural agenda of translation & Arabization: Aspects of the problems.
Express, an International Journal of Multi Disciplinary Research, 1(2), 1–14.

Al-Shawi, M. A. (2012). Strategies for translating idioms from Arabic and vice versa. Journal of
American Arabic Academy for Sciences and Technology, 3(6), 139–147.

Baker, M. (1992/2011). In other words: A coursebook on translation (2nd ed.). London and New
York: Routledge.

Delisle, J. (1988). Translation: An interpretive approach. Canada: University of Ottawa Press.

Fernando, C. (1996). Idioms and idiomaticity. Oxford: Oxford University Press.

Uploaded by S. M. Safi
Gibbs, R. W. (1993). What do idioms really mean? Journal of Memory and Language 31, 485–506.

Gibbs, R. W., Nayak, N. P., & Cutting, J. C. (1989). How to kick the bucket and not decompose:
Analyzability and idiom processing. Journal of Memory and Language, 28, 576–593. https://doi.
org/10.1016/0749-596X(89)90014-4.

Glucksberg, S. (1993). Idiom meanings and allusional content. In C. Cacciari & P. Tabossi (Eds.),
Idioms: Processing, structure, and interpretation (pp. 3–26). Hillsdale, NJ: Lawrence Erlbaum
Associates.

Haque, M. Z. (2012). Translating literary prose: Problems and solutions. International Journal of
English Linguistics, 2(6), 97–111. https://doi.org/10.5539/ijel.v2n6p97.

Harvey, M. (2000). A beginner’s course in legal translation: The case of culture-bound terms.
ASTTI/ETI, 2(24), 357–369.

Katan, D. (1993). The English Translation of 1 l Nome della Rosa and the Cultural Filter, Umberto
Eco, Claudio Magris: Autori e Traduttori a Confronto. In L. Avirovic & J. Dodds (Eds.), Campanotto
Editore, Udine, pp. 149–165.

Mäntylä, K. (2004). Idioms and language users: The effect of the characteristics of idioms on their
recognition and interpretation by native and nonnative speakers of English. Dissertation
Abstracts International, C: Worldwide. Retrieved from http://search.proquest.com/docview/
85637231?accountid=8330%5Cnhttp://library.anu.edu.au:4550/resserv?genre=dissertations+&
+theses&issn=&title=Idioms+and+Language+Users:+The+Effect+of+the+Characteristi
cs+of+Idioms+on+Their+Recognition+and+Interpretation+by+N.

Mares, R. (2012). Cultural dif iculties in translations from English into Arabic. Analele Universitatii
Crestine Dimitrie Cantemir, Seria Stiintele Limbii, Literaturii Si Didactica Predarii, 1(33), 1–7.

Newmark, P. (1988). A textbook of translation. Hertfordshire: Pearson Education Limited.

Nida, E. (2000). Principles of correspondence. In L. Venuti (Ed.), The translation studies readers
(pp. 126–140). London and New York: Routledge (First published in 1964, Leiden, Holland: E.J.
Brill).

Ordudari, M. (2007). Translation procedures, strategies and methods. Translation Journal, 3(5),
781–789.

Pym, A. (2018). A typology of translation solutions. Journal of Specialized Translation, 30, 41–65.

Razmjou, L. (2004). To be a good translator. Translation Journal, 8(2). Retrieved from https://
translationjournalnet/journal/28eduhtm.

Richards, J., & Schmidt, R. (2002). Longman dictionary of language teaching and applied
linguistics. London: Longman Publishing Group.

Spencer-Oatey, H. (Ed.). (2000). Culturally speaking: Managing rapport through talk across
cultures. London and New York: Continuum.

Venuti, L. (1995). The translator’s invisibility. Abingdon, UK: Taylor & Francis. https://doi.org/10.
4324/9780203360064.

Uploaded by S. M. Safi
Vinay, J. P., & Darbelnet, J. (1958/2004). A methodology for translation (trans.) by J. C. Sager and
M.-J. Hamel. In L. Venuti (ed.) (2004) The translation studies reader (pp. 128–37). London and
New York: Routledge.

Uploaded by S. M. Safi
Bibliography
Abdel Haleem, M. A. S. A. (2004).The Qur’an . Oxford: Oxford University Press Inc.

Abdel Haleem, M. A. S. A. (2005).The Qur’an . Oxford: Oxford University Press.

Abdelwali, M. (2007). The loss in the translation of the Qur’an [Electronic version].The
Translation Journal, 11 (2), 120–125. Retrieved from http://www.accurapid.com/journal/
40quran.htm .

Abdul Aziz, M. M. (2008). The translation of some rhetorical devices in Al-Fātiha Sūra into
English.Adab Al-Ra idayn, 51 (2), 41–70.

Abdul-Raof, H. (2004). The Quran: Limits of translatability. In S. Faiq (Ed.),Cultural encounters in


translation from Arabic . Frankfurt Lodge: Multilingual Matters Ltd.

Abdul-Raof, H. (2005). Pragmalinguistic forms in cross-cultural communication: Contributions


from Qur’an translation.Intercultural Communication Studies, 14 (4), 115–130.

Abdul-Raof, H. (2010). Qur’anTranslation, discourse, texture and exegesis . London and New York:
Routledge.

Abedelrazq, Y. I. (2014).Problems of translating homonymy in the glorious Quran: A comparative


analytical study (Unpublished thesis).

Abu Musa, M. (1987).Characteristics of structures: Analytical study for issues in semantics . Cairo:
Maktabat Wahbah.

Ad-Darwish, M. (1992).IAArab Al Quran (Parsing of the Holy Quran) . Hams: Al Yamamah


Publishing House.

Ahmed, M. F. (2006). Investigating some semantic problems in the translation of the Holy
Quran.Adab Al-Ra idayn, 2 (43), 61–72.

Ahmed, M. F. (2008). Approaches to denotative and connotative meanings in the translations of the
Holy Quran.Adab Al-Ra idayn, 2 (50), 1–30.

AL-Alawi, Y. (1982).The secrets of rhetorics . Cairo: Dar Al-Kutub Al-Alamyah.

Al-Azab, A., & Al-Misned, A. (2012). Pragmatic losses of Qur’an translation: A linguistic
approach.English Language and Literature Studies, 2 (3), 42–49. https://doi.org/10.5539/ells.
v2n3p42 .

Al-Bayt, R. A. (2008).The Qur’an: A new English translation of its meanings . Amman: Royal Aal al-
Bayt Institute for Islamic Thought.

Aldahesh, A. Y. (2017). On idiomaticity in English and Arabic: A cross—linguistic study.Journal of


Language and Culture, 4 (2), 23–29. https://doi.org/10.5897/JLC2013.0220 .

Al-Ghazali, M. F. (2010). Lexical gaps in Arabic -to-English translation.Al-Mustansiriya Journal of


Arts, 1 (52), 1–16.

Uploaded by S. M. Safi
Al-Ghazalli, M. F. (2012). A study of the English translations of the Qur’anic verb phrase: The
derivatives of the triliteral.Theory and Practice in Language Studies, 2 (3), 605–612. https://doi.
org/10.4304/tpls.2.3.605-612 .

AL-Haddad, K. (1994). Investigating dif iculties faced by advanced Iraqi students of English in and
using English idioms (Unpublished MA thesis). College of Arts, University of Baghdad.

Al-Haj, A. A. M. (2014). The Cultural agenda of translation & Arabization: Aspects of the
problems.Express, an International Journal of Multi Disciplinary Research, 1 (2), 1–14.

Ali, A., Brakhw, M. A., Bin Nordin, M. Z. F., & ShaikIsmail, S. F. (2012). Some linguistic dif iculties in
translating the Holy Quran from Arabic into English.International Journal of Social Science and
Humanity, 2 (6), 588–590. https://doi.org/10.7763/IJSSH.2012.V2.178 .

Ali, A., Brakhw, M. A., & Zarirruddin, M. (2014). Transferring polysemic words from Arabic into
English: A comparative study of some samples from the Holy Quran.Australian Journal of Basic
and Applied Sciences, 8 (23), 38–43.

Ali, A. Y. (1968/2006).The Holy Qur’an, text, translation and commentary (Trans.). Beirut,
Lebanon: Dar al Arabia.

Almubark, A. A., Manan, D. A., & Al-Zubaid, K. (2014). The hindrances in translating speci ic
cultural concepts from Arabic into English.IOSR Journal of Humanities and Social Science, 19 (3),
166–173. https://doi.org/10.9790/0837-1932166173 .

Al-Munajjid, M. N. (1997).Attradef i ALQuran al Kareem bayna Al Nazariyyah wa Att Atbeeq


[Synonyms in the Holy Quran from a theoretical and a practical perspective]. Beirut: Dar Al Fikr Al
Moaser.

Al-Najjar, M. F. (1984).Translation as a correlative of meaning (Unpublished PhD thesis).


Bloomington: Indiana University.

AlQinai, J. (2012). Convergence and divergence in the interpretation of Quranic polysemy and
lexical recurrence.Pan-Paci ic Association of Applied Linguistics, 16 (1), 83–109.

Al-Qurtubi, M. S. (2004).Al JamAA liahkam al Qur’an (Tafsir Al Qurtubi) . Cairo, Egypt: Dar Al-Fikr.

Al Salem, M. N. (2014).The translation of metaphor from Arabic to English in selected poems of


Mahmoud Darwish with a focus on linguistic issues (Unpublished thesis). The University of Leeds.

Alshaje’a, H. (2014). Issues in translating collocations of the Holy Qur’an.Language in India, 14


(8), 49–65.

Al-Shawi, M. A. (2012). Strategies for translating idioms from Arabic and vice versa.Journal of
American Arabic Academy for Sciences and Technology, 3 (6), 139–147.

Al Sughair, Y. (2011).The translation of lexical collocations in literary texts . Sharjah: The


American University of Sharjah.

Al-Utbi, M. (2011). Translation of ‫ ﻛﺎد‬in the Quran as an instance of lexical gaps.Majallat Kulliatu
Aladab, 1 (98), 69–92.

Uploaded by S. M. Safi
Al-Zamakhshari, A. A.-Q. (2000).Al Kashshaf AAn haqaiq ghawamed attanzeel [The revealer of
facts of obscure revelations]. Cairo, Egypt: Arabic Publishing House.

Arberry, A. (1982).The Koran interpreted . Oxford: Oxford University Press.

As-Sa i, A. B. (2011).Translation theories: Strategies and basic theoretical issues . Amman: Dar
Amwaj.

Asyuti, J. (2008).Muzhir i Olum Allughah wannwa’aha [The science of language and its types]
(3rd ed.). Cairo: Dar Al-Turath.

Ayyash, J., Ahmad, S., & Abdullah, N. (2013). Models of alliteration derivation in the
Quran.International Journal of Islamic Thought, 3 (June), 113–120. http://journalarticle.ukm.
my/6417/1/11_Jamil_Ayyash_IJIT_Vol_3_Jun_2013.pdf .

Az-Zarkashi, M. B. E. (2006).Al Burhan i Oloum Al Quran [The evidence in the Holy Quran
sciences]. Cairo: Dar al-Hadeeth.

Baker, M. (1992/2011).In other words: A coursebook on translation (2nd ed.). London and New
York: Routledge.

Baker, M. (2004). The status of equivalence in translation studies: An appraisal. In Z. Yang


(Ed.),English-Chinese comparative study and translation (p. 1). Shanghai: Foreign Languages
Education Press.

Bassnett, S. (2005).Translation studies .Vasa (3rd ed.). New York: Routledge (Taylor & Francis e-
Library).

Beekman, J., & Callow, J. (1974).Translating the word of god: With scriptures and topical indexes .
Grand Rapids, MI: Zondervan Publishing House.

Bell, R. T. (1991).Translation and translating: Theory and practice . London and New York:
Longman.

Benfoughal, A. (2010).Students’ dif iculties and strategies in translation: The case of third year
students (PhD thesis). University of Constantine, Algeria.

Bentivogli, L., & Pianta, E. (2000). Looking for lexical gaps. InProceedings of the Ninth EURALEX
International Congress (pp. 1–6). Trento, Italy. Retrieved from http://dialnet.unirioja.es/servlet/
articulo?codigo=4616045 .

Bock, J. K. (1986). Syntactic persistence in language production.Cognitive Psychology, 18 (3), 355–


387. https://doi.org/10.1016/0010-0285(86)90004-6 .

Cambridge University Press. (2008). Cambridge online dictionary . Retrieved at April 23, 2008,
from the website temoa: Open Educational Resources (OER) Portal at http://www.temoa.info/
node/324 .

Catford, J. C. (1965).A linguistic theory of translation: Language and language learning (1st ed.).
Oxford: Oxford University Press.

Chesterman, A. (2001). Proposal for a Hieronymic Oath.The Translator, 7 (2), 139–154.

Uploaded by S. M. Safi
Chesterman, A. (2012). Catford revisited. InShall we play the festschrift game? (pp. 25–33). Berlin
and Heidelberg: Springer. https://doi.org/10.1007/978-3-642-30773-7 .

Chomsky, N. (1977). Recent contributions to the theory of innate ideas: Summary of oral
presentation. In J. Searle (Ed.),Philosophy of language . Oxford: Oxford University Press.

Collins COBUILD English Grammar . (2005). Glasgow: HarperCollins.

Collins, H. (2006).CoBuild advanced learner’s English dictionary . Glasgow, Scotland:


HarperCollins.

Conner, D. (1983).Understanding semantics . Exeter: Exeter University Press.

Cruse, D. A. (1997).Lexical semantics . Cambridge: Cambridge University Press.

Cui, Y., & Zhao, Y. (2014). Translation of rhetorical igures in the advertising discourse: A case
study.International Journal of Society, Culture & Language (IJSCL), 2 (2), 57–67.

Darwish, A. (2010).Elements of translation . Melbourne: Writescope.

Daryabadi, A. (2007).The glorious Quran: Text translation and commentary . London: Islamic
Foundation.

Delisle, J. (1984). L’Analyse du Discours comme Methode de Traduction. (Theorie et pratique)


Initiation i la traduction francaise des textes pragmatiques anglaisn (Model for Translation-
Oriented Text. Analysis). Editions de l’Unuversite d'Ottawa.

Delisle, J. (1988).Translation: An interpretive approach . Ottawa, ON, Canada: University of


Ottawa Press.

Dickens, J., Hervey, S., & Higgins, I. (2005).Thinking Arabic translation: A course in translation
method—Arabic to English . London: Routledge.

Dweik, B., & Abu Shakra, M. (2010). Strategies in translating collocations in religious texts from
Arabic into English.Atlas Global Journal for Studies and Research, 5 (1), 5–41.

Emara, S. (2014). Extraordinary vocabulary of the Qur’an and related translation


problems.International Journal of Linguistics, 5 (1), 248–272. https://doi.org/10.5296/ijl.v5i1.
3309 .

Emara, S. A. E. (2013). Extraordinary vocabulary of the Qur’an and related translation


problems.International Journal of Linguistics, 5 (1), 248–272. https://doi.org/10.5296/ijl.v5i1.
3309 .

Even-Zohar, I. (1979). Polysystem theory.Poetic Today, 1 (1978), 1–2.

Even-Zohar, I. (1997). Itamar Even-Zohar: Polysystem studies 1990.International Journal for


Theory and Analysis of Literature and Communication, 11 (1), 88.

Even-Zohar, I. (2000). The position of translated literature within the literary polysystem. In L.
Venuti (Ed.),The translation studies reader (pp. 192–197). London: Routledge.

Uploaded by S. M. Safi
Fernando, C. 1996.Idioms and idiomaticity . Oxford: Oxford University Press.

Frawley, W. (1984). Prolegomenon to a theory of translation. In W. Frawley (Ed.),Translation:


Literary, linguistic, and philosophical perspectives . London and Toronto: Associated University
Presses.

Geeraerts, D. (2010).Theories of lexical semantics . New York: Oxford University Press Inc.

Gentzler, E. (2001).Contemporary translation theories . London and New York: Routledge.

Ghali, M. (2005).Towards understanding the ever–glorious Qur’an . Cairo: Dar An–Nashr Liljami.

Ghazala, H. (2008).Translation as problems and solution . Beirut: Dar El-Ilm Lilmalayin.

Gibbs, R. (1994).The poetics of mind: Figurative thought, language, and understanding .


Cambridge: Cambridge University Press.

Gibbs, R. W. (1993). What do idioms really mean?Journal of Memory and Language 31 , 485–506.

Gibbs, R. W., Nayak, N. P., & Cutting, J. C. (1989). How to kick the bucket and not decompose:
Analyzability and idiom processing.Journal of Memory and Language, 28 , 576–593. https://doi.
org/10.1016/0749-596X(89)90014-4 .

Glucksberg, S. (1993). Idiom meanings and allusional content. In C. Cacciari & P. Tabossi
(Eds.),Idioms: Processing, structure, and interpretation (pp. 3–26). Hillsdale, NJ: Lawrence
Erlbaum Associates.

Guillaume, A. (1990).Islam . London: Penguin Books.

Halliday, M. A. K. (1971). Linguistic function and literary style: An inquiry into the language of
William Golding’sThe inheritors . In S. B. Chatman (Ed.),Literary style: A symposium (pp. 330–
365). London and New York: Oxford University Press.

Halliday, M. A. K. (1972/1994).An introduction to functional grammar (2nd ed.). London:


Cambridge University Press.

Halliday, M. A. K. (1985).Functional grammar (2nd ed.). London: Arnold.

Halliday, M. A. K. (1992). Language theory and translation practice.Rivista Internazionale Di


Tecnica Della Traduzione, 1 (1), 15–25.

Halliday, M. A. K. (2001). Towards a theory of good translation. In E. Steiner & C. Yallop


(Eds.),Exploring translation and multilingual text production: Beyond content (pp. 13–18). Berlin
and New York: Mouton de Gruyter.

Halliday, M. A. K. (2007).Language and education . London: Continuum.

Halliday, M. A. K., & Hasan, R. (1976).Cohesion in English . London: Longman Publishing House.

Halliday, M. A. K., & Matthiessen, C. (2014).Halliday’s introduction to functional grammar (4th


ed.). New York: Routledge (Taylor & Francis e-Library). https://doi.org/10.4324/
9780203431269 .

Uploaded by S. M. Safi
Halliday, M. A. K., Macintosh, A., & Strevens, P. (1965).The linguistic sciences and language
teaching . London: Longman Publishing House.

Haque, M. Z. (2012). Translating literary prose: Problems and solutions.International Journal of


English Linguistics, 2 (6), 97–111. https://doi.org/10.5539/ijel.v2n6p97 .

Harvey, K. (1995). A descriptive framework for compensation.The Translator, 1 (1), 56–86.

Harvey, M. (2000). A beginner’s course in legal translation: The case of culture-bound


terms.ASTTI/ETI, 2 (24), 357–369.

Hatim, B., & Mason, I. (1993).The discourse and the translator (4th ed.). London and New York:
Longman Inc.

Hatim, B., & Mason, I. (2005).The translator as communicator . London: Routledge.

Hatim, B., & Munday, J. (2004).Translation: An advanced resource book . London and New York:
Routledge (Taylor & Francis e-Library).

Hervey, S., & Higgins, I. (1992).Thinking translation: A course in translation method—French-


English . New York: Routledge (Taylor & Francis).

Hodges, P. (2009). Compare and contrast two theoretical approaches to translation in


Zainurrahman.The theories of translation from history to procedures . Language and Education.
Retrieved from https://scholar.google.com/scholar?hl=enandq=De+Waard+NidaandbtnG=
Searchandas_ylo=andas_vis=0#8 .

Holz-Mänttäri, J. (1984).Translatorisches Handeln. Theorie und Methode . Helsinki: Academiae


Scientarum Fennicae.

House, J. (1997).Translation quality assessment: A model revisited . Tübingen: Gunter Narr.

House, J. (2001, January). Translation quality assessment: Linguistic description versus social
evaluation.Meta: Journal Des Traducteurs, 46 (2), 243–257. https://doi.org/10.7202/003141ar .

House, J. (2015). Translation quality assessment: Past and present. InTranslation: A


multidisciplinary approach (pp. 241–264). London and Chicago: Palgrave Macmillan.

Hurford, J. R., Heasley, B., & Smith, M. B. (2014).Semantics: A coursebook . Cambridge: Cambridge
University Press.

Ibn Ashour, M. (1984).Tafsir Attahrir Wattanwir [Liberation and enlightenment interpretation


book] ( irst). Tunis: Tunisian Publishing House.

Ilyas, A. I. (2013). Intra-textuality in translating some problematic Qur’anic verses.Arab World


English Journal, 2 (Special Issue on Translation), 86–95.

Irving, T. (1985).The Noble Qur’an . Cedar Rapids: Iowa Publisher.

Irving, T. (1988).The Noble Qur’an . Cedar Rapids: Iowa Publisher.

Jaber, I. (2010). Translating the genre of Quran: The challenge of translating the inimitable.Journal

Uploaded by S. M. Safi
of College Education for Women, 21 (4), 943–954.

Jakobson, R. (1959/1966/2000). On linguistic aspects of translation. In R. A. Brower (Ed.),On


translation . Cambridge, MA: Harvard University Press.

Johnson, K. (2001). What VP ellipsis can do, and what it can’t, but not why. InThe handbook of
contemporary syntactic theory (pp. 439–479). https://doi.org/10.1111/b.9781405102537.2003.
00016.x .

Kabali, S. (2006).Semantics of metaphor: An overview of MAJAZ interpretation in the Holy Qur’an


(Unpublished PhD thesis). Islamic University, Malaysia. Retrieved from http://lib.iium.edu.my/
mom2/cm/content/view/view .

Kalakattawi, F. A. (2005).Lexical relation with reference to polysemy in translation (Unpublished


paper). College of Education, Jadda, Saudi Arabia.

Katan, D. (1993). The English translation of 1l Nome della Rosa and the Cultural Filter. In L.
Avirovic & J. Dodds (Eds.),Umberto Eco, Claudio Magris: Autori e Traduttori a Confronto (pp.
149–165). Udine: Campanotto Editore.

Kehal, M. (2010).Problems in English Arabic translation of reference pragmatic aspects


(Unpublished MA thesis). Mentori University- Constontine, Algeria. Retrieved from http://bu.umc.
edu.dz/theses/anglais/KEH1119.pdf .

Kenny, D. (1998). Equivalence. In M. Baker (Ed.),Routledge encyclopaedia of translation studies .


London and New York: Routledge.

Khalifa, M. (1989).The sublime Qur’an and orientalism (2nd ed.). Karachi: International Islamic
Publisher (Pvt.) Ltd.

Khan, M. M., & Al-Hilali, T. U. (1996).Interpretation of the meanings of the Noble Qur’an . Riyadh:
Maktaba Dar-us-Salam.

Khorami, M. (2014). Eloquence of repetition in Quran and Arabic old poetry.Language Related
Research, 5 (2), 90–110.

Klepousniotou, E. (2002). The processing of lexical ambiguity: Homonymy and polysemy in the
mental lexicon.Brain and Language, 81 (1–3), 205–223. https://doi.org/10.1006/brln.2001.2518
.

Klepousniotou, E., & Baum, S. R. (2005). Processing homonymy and polysemy: Effects of
sentential context and time-course following unilateral brain damage.Brain and Language, 95 (3),
365–382. https://doi.org/10.1016/j.bandl.2005.03.001 .

Koller, W. (1976/1979).Einführung die Übersetzungswissenschaft . Heidelberg: Quelle and Meyer.

Koller, W. (1989). Equivalence in translation theory. In A. Chesterman (Ed. and Trans.),Readings in


translation theory (pp. 99–104). Helsinki: Oy Finn Lectura Ab.

Koller, W. (1995). The concept of equivalence and the object of translation studies.Target, 7 (2),
191–222.

Uploaded by S. M. Safi
Kövecses, Z. (2010). Metaphor: A practical introduction.English Linguistics/Journal of the English
Linguistic Society of Japan, 22 (1), 205–231.

Larson, M. (1998).Meaning-based-translation . Oxford: University Press of American Inc.

Leuven-Zwart, K. V. (1989).Translation and originals: Similarities and dissimilarities I .


Amsterdam: John Benjamins.

Levý, J. (1967). Translation as a decision process. In L. Venuti (Ed.),To honour Roman Jakobson on
the occasion of his seventieth birthday (Vol. 2, pp. 1171–1182). The Hague: Mouton.

Levy, J. (1976). Translating as a decision process. In L. Venuti (Ed.),The translation studies reader
(pp. 148–189). London: Routledge.

Lobner, S. (2002).Understanding semantics . Sonipat, India: Replika Press.

Lyons, J. (1977).Semantics . Cambridge: Cambridge Scholars Publishing.

Manfredi, M. (2008). Translating text and context: Translation studies and systemic functional
linguistics (Vol. 1) Translation theory (2nd ed., p. 97). Bologna: Centro di Studi Linguistico-
Culturali (CeSLiC). https://doi.org/10.6092/unibo/amsacta/2441 . In: Quaderni del CeSLiC.
Functional Grammar Studies for Non-Native Speakers of English.

Manfredi, M. (2014).Translating text and context: Translation studies and systemic functional
linguistics. Volume 2: From theory to practice (p. 158). Bologna: Centro di Studi Linguistico-
Culturali (CeSLiC). https://doi.org/10.6092/unibo/amsacta/3219 . In: Quaderni del CeSLiC.
Functional Grammar Studies for Non-Native Speakers of English.

Mäntylä, K. (2004).Idioms and language users: The effect of the characteristics of idioms on their
recognition and interpretation by native and nonnative speakers of English . Dissertation
Abstracts International, C: Worldwide. Retrieved from http://search.proquest.com/docview/
85637231?accountid=8330%5Cnhttp://library.anu.edu.au:4550/resserv?genre=dissertations+&
+theses&issn=&title=Idioms+and+Language+Users:+The+Effect+of+the+Characteristi
cs+of+Idioms+on+Their+Recognition+and+Interpretation+by+N .

Mares, R. (2012). Cultural dif iculties in translations from English into Arabic.Analele Universitatii
Crestine Dimitrie Cantemir, Seria Stiintele Limbii, Literaturii Si Didactica Predarii, 1 (33), 1–7.

Mishra, P. (2009, September). Strength for today and bright hope for tomorrow etymological
analysis of the English language words.Language in India, 12 , 63–75.

Munday, J. (2001).Introducing translation studies . New York: Routledge.

Munday, J. (2008).Introducing translation studies: Theories and applications (2nd ed.). London
and New York: Routledge.

Munday, J. (2009).The Routledge companion to translation studies . London and New York:
Routledge.

Newmark, P. (1981).Approaches to translation . Oxford: Pergamon.

Newmark, P. (1988).A textbook of translation: Text . Hertfordshire: Pearson Education Limited.

Uploaded by S. M. Safi
Newmark, P. (1991).About translation . Clevedon: Multilingual Matters.

Newmark, P. (1998).Approaches to translation . Singapore: Prentice Hall.

Newmark, P. in Munday, J. (2009).The Routledge companion to translation studies . London and


New York: Routledge.

Ni, L. (2009). For translation and theories.English Language Teaching, 2 (2), 78–83.

Nida, E. (1964).Toward a science of translating: With special reference to principles and


procedures involved in Bible translating . Leiden: E.J. Brill.

Nida, E. (2000). Principles of correspondence. In L. Venuti (Ed.),The translation studies readers


(pp. 126–140). London and New York: Routledge ( irst published in 1964).

Nida, E., & Taber, C. (1982).The theory and practice of translation (2nd ed.). Leiden: E.J. Brill.

Nida, E. A. (1959/1975). Principles of translating as exempli ied by Bible translating. In A. S. Dil


(Ed.),Language structure and thought: Essays by Eugene A. Nida. Stanford: Stanford University
Press.

Nida, E. A. (1994). Translation: Possible and impossible. Turjuman: revue de traduction et


d'interprétation=Journal of Translation Studies, 3 (2), 147–163.

Nida, E. A., & Taber, C. R. (1969).The theory and practice of translation . Leiden: E.J.Brill.

Nord, C. (1989). Loyalty instead of loyalty. Proposals for a functional translation typology.Living
Languages, 34 (3), 100–105.

Nord, C. (1997).Translating as a purposeful activity . Manchester: St. Jerome.

Nord, C. (2005).Text analysis in translation (2nd ed.). Amsterdam: Editions Rodopi B.V.

Nord, C. (2008). Persuading by addressing: A functional approach to speech-act


comparison.Southern African Linguistics and Applied Language Studies, 26 (2), 283–293.

Nugroho, A. B. (1999). Meaning and translation.Journal of English and Education, 2 (3), 94–112.

Ogden, M., & Richards, I. A. (1923).The meaning of meaning . New York and London: Harcourt
Brace Jovanovich.

Ordudari, M. (2007). Translation procedures, strategies and methods.Translation Journal, 3 (5),


781–789.

Palmer, F. (1981).Semantics: A new outline . London and New York: Cambridge University Press.

Pickthall, M. (2001).The meaning of the glorious Qur’an . Kuala Lumpur: Islamic Book Trust.

Panou, D. (2013). Equivalence in translation theories: A critical evaluation.Theory and Practice in


Language Studies, 3 (1), 1–6.

Popovič, A. (1976).Dictionary for the analysis of literary translation Edmonton . Alberta:

Uploaded by S. M. Safi
Department of Comparative Literature, University of Alberta.

Pym, A. (2001). Introduction: The return to ethics in translation studies.The Translator, 7 (2),
129–138.

Pym, A. (2004).The moving text: Localization, translation, and distribution . Amsterdam and
Philadelphia: John Benjamins.

Pym, A. (2018). A typology of translation solutions.Journal of Specialized Translation, 30, 41–65.

Quine, W. V. O. (1959). Translation and meaning. In R. A. Brower (Ed.),On translation . Cambridge,


MA: Harvard University Press (Reprinted in L. Venuti (Ed.). (2000).The translation studies reader
(pp. 94–112). London: Routledge).

Quinn, A. (1993).Figures of speech: 60 ways to turn a phrase . Utah: Gibbs Smith Publisher.

Rasekh, A. E., Dastjerdi, H., & Bassir, A. (2012). On homonymous expressions in the Qur’an: A case
study of the English translations of the term (fasad).The Journal of International Social Research,
5 (22), 136–148.

Razmjou, L. (2004). To be a good translator.Translation Journal, 8 (2). Retrieved from https://


translationjournalnet/journal/28eduhtm .

Reiss, K. (1971).Möglichkeiten und Grenzen der Übersetzungskritik . Munich: M. Hueber [Trans. E.


F. Rhodes. (2000).Translation criticism: Potential and limitations ]. Manchester: St. Jerome and
American Bible Society.

Reiss, K. (2004). Type, kind and individuality of text: Decision making in translation. In L. Venuti
(Ed.),The translation studies reader (2nd ed., pp. 168–179). New York: Routledge.

Richards, J., & Schmidt, R. (2002).Longman dictionary of language teaching and applied
linguistics . London: Longman Publishing Group.

Sadiq, S. (2008). Some semantic, stylistic and cultural problems of translation with special
reference to translating the glorious Qur’ân.Sayyab Translation Journal, 1 (1), 37–59.

Sadiq, S. (2010).A comparative study of four English translations of Sûrat Ad-Dukhân on the
semantic level (1st ed.). Newcastle: Cambridge Scholars Publishing.

Said Ghazala, H. (2012). Translating the metaphor: A cognitive stylistic conceptualization (English
– Arabic).World Journal of English Language, 2 (4), 57–68. https://doi.org/10.5430/wjel.
v2n4p57 .

Sale, G. (1734).The Koran . London: Frederick Warne.

Semino, E. (2008).Metaphor in discourse . Cambridge: Cambridge University Press.

Shakernia, S. (2014). Study of Nida’s (formal and dynamic equivalence) and Newmark ’s (semantic
and communicative translation) translating theories on two short stories.Merit Research Journal
of Education and Review, 2 (1), 1–7.

Shakir, M. (1999).The Holy Quran . New York: TahrikeTarsile Quran.

Uploaded by S. M. Safi
Sharifabad, E., Mahadi, T. S., & Kenevisi, M. S. (2012). Linguistic ambiguity in the Holy Qur’ān and
its English translations.International Journal of Linguistics, 4 (3), 352–362.

Shehab, E. (2009). The problems involved in translating Arabic cognitive synonyms into
English.Majallat Al-Jaam ’Ah Al-Islamiyyah, 17 (1), 869–890.

Shunnaq, A. (1992). Functional repetition in Arabic realized through the use of word-strings with
reference to Arabic-English translation of political discourse.NouveltesDe La Fit-Newsletter, 2 (1),
5–39.

Simpson, G. B. (1981). Meaning dominance and semantic context in the processing of lexical
ambiguity.Journal of Verbal Learning and Verbal Behavior, 20 (1), 120–136.

Snell-Hornby, M. (1988/1995).Translation studies: An integrated approach . Amsterdam and


Philadelphia: John Benjamins.

Spencer-Oatey, H. (Ed.). (2000).Culturally speaking: Managing rapport through talk across


cultures . London and New York: Continuum.

Stanojević, M. (2009). Cognitive synonymy: A general overview.Facta Universitatis—Linguistics


and Literature, 7 (2), 193–200.

Steiner, G. (1975/1998).After Babel: Aspects of language and translation . Oxford and New York:
Oxford University Press.

Toury, G. (1980).In search of a theory of translation (p. 159). Tel Aviv: The Porter Institute for
Poetics and Semiotics, Tel Aviv University.

Venuti, L. (1995).The translator’s invisibility . Abingdon, UK: Taylor & Francis. https://doi.org/10.
4324/9780203360064 .

Venuti, L. (2000). Translation, community, Utopia. In L. Venuti (Ed.),The translation studies reader
(pp. 468–488). London: Routldge.

Venuti, L. (2004).The translation studies reader . London and New York: Routledge.

Vermeer, H. (1986). Übersetzen als kultureller transfer. In M. Snell-Hornby (Ed.) (1990),Linguistic


transcoding or cultural transfer? A critique of translation theory in Germany (pp.79–86). In S.
Bassnett & A. Lefevere (Eds.).

Vinay, J. P., & Darbelnet, J. (1958/2004). A methodology for translation. In J. C. Sager & M.-J. Hamel
(Trans.) and L. Venuti (Ed.),The translation studies reader (pp. 128–137). London and New York:
Routledge.

Wang, J. (2010). A critical discourse analysis of Barack Obama’s speeches.Journal of Language


Teaching and Research, 1 (3), 254–261. https://doi.org/10.4304/jltr.1.3.254-261 .

Wendland, E. R. (2012). Framing the frames: A theoretical framework for the cognitive notion of
“Frames of Reference.”Journal of Translation, 6 (1), 27–50.

Whang, Y. C. (2004). To whom is a translator responsible—Author or reader? In S. E. Porter & R. S.


Hess (Eds.),Translating the Bible: Problems and prospects (pp. 46–62). New York: Continuum and

Uploaded by S. M. Safi
T&T Clark International.

Widdowson, H. (1978).Teaching language as communication . Oxford: Oxford University Press.

Wilss, W. (1982).The science of translation: Problems and methods . Tübingen: Gunter Narr.

Wilss, W. (1996).Knowledge and skills in translator behaviour . Amsterdam: John Benjamins.

Xu, X. (2008). The rhetoric and translation of English advertisement.International Journal of


Business and Management, 3 (11), 83–86.

Yinhua, X. (2011). Equivalence in translation: Features and necessity.International Journal of


Humanities and Social Science, 1 (10), 1989–1991.

Zaky, M. M. (2001). Translation and language varieties.The Translation Journal, 5 (3). Retrieved
from http://translationjournal.net/journal/17theory.htm .

Zhuanglin, H. (1988).A course of linguistics . Peking: Peking University Press.

Index
A
Aesthetic function
Appellative
Audiomedial texts
B
Baker, M.
C
Catford, J.C.
Chomsky, N.
The Cognitive approach
Collocations
The Communicative stage
Communicative translation
Covert translation
Culturally bound terms
Culture
Culture-speci ic collocations
D

Uploaded by S. M. Safi
Darbelnet, J.
Delisle, J.
Direct and oblique translation
Dynamic equivalence
E
The Engrossing effect of source text patterning
Expressive function
F
Fixed expressions
Formal equivalence
Form-based
The Functional hierarchy of translation problems
Functions of texts
G
Gender
Ghazala, H.
Grammatical category
Grammatical problems
H
Halliday, M.A.K.
Harvey, M.
Homonymy
I
Idioms
Informative function
Informative texts
Interlingual translation
Intersemiotic translation
Intralingual translation
Invariance
J
Jakobson, R.
Juliane House

Uploaded by S. M. Safi
K
Katan, D.
Koller, W.
L
Larson, M.
Levy, J.
Lexical ambiguity
Lexical gaps
Linguistic stage
M
Malinowski
Manfredi, M.
Marked collocations in the source text
Meaning-based
Metalingual function
Munday, J.
N
Newmark, P.
Nord, C.
O
Operative texts
Overt and covert translations
Overt translation
P
Passivization
Phatic function
Polysemy
Polysystems
The PolySystems theory
Popovič, A.
Problem in translation
Pym, A.
R

Uploaded by S. M. Safi
Reiss, Katharina
Rhetorical devices
The Role of the ST analysis
S
Semantic ield
Semantic translation
Shifting
Skopos
Skopos theory
Steiner, George
Strategies to translate culturally bound terms
Synonymy
Syntactic order
T
Tense
The Tension between accuracy and naturalness
Translation brief
Translation methods
Translation procedures
Translation unit
V
Van Leuven-Zwart
Venuti, L.
Vermeer, Hans
Vinay, J.P.
Vocative function

Uploaded by S. M. Safi

You might also like