Burlan.gratiela Licență Varianta Finala 2 [623031]

TRANSILVANIA UNIVERSITY OF BRAȘOV
FACULTY OF LETTERS
APPLIED MODERN LANGUAGES FRENCH -ENGLISH

B.A. PROJECT
THE TRANSLATION OF COLLOCATIONS IN
“TURTLES ALL THE WAY DOWN” BY JOHN GREEN

Candidat e,
Burlan Grațiela -Valentina

Supervisor ,
Conf . dr. Oana Tatu

BRAȘOV
2019

2
Table of Contents
ABSTRACT ………………………….. ………………………….. ………………………….. ………………………….. …. 4
CHAPTER ONE: INTRODUCTION ………………………….. ………………………….. ……………………….. 5
1.1. General remarks ………………………….. ………………………….. ………………………….. ………………. 5
1.2. Aim and research question ………………………….. ………………………….. ………………………….. … 6
1.3. Outline of the thesis ………………………….. ………………………….. ………………………….. ………….. 6
CHAPTER TWO: INTRODUCTION TO TRANSLATION STUDIES ………………………….. ……. 8
2.1. Introduction ………………………….. ………………………….. ………………………….. …………………….. 8
2.2. What is translation? ………………………….. ………………………….. ………………………….. ………….. 8
2.3. Types of translation ………………………….. ………………………….. ………………………….. ………….. 9
2.4. Stages of translation ………………………….. ………………………….. ………………………….. ………. 10
2.5. Highlights in the history of translation ………………………….. ………………………….. …………… 12
2.6. The possibility versus the impossibility of translations ………………………….. ………………… 13
2.7. The fide lity of translations ………………………….. ………………………….. ………………………….. . 15
2.8. Translation strategies ………………………….. ………………………….. ………………………….. ………. 17
2.9. Problems of equivalence ………………………….. ………………………….. ………………………….. …. 19
2.10. Cultural aspects of translation ………………………….. ………………………….. …………………….. 21
2.11. Conclusion ………………………….. ………………………….. ………………………….. …………………… 26
CHAPTER THREE: THE TRANSLATION OF COLLOCATIONS ………………………….. ………. 28
3.1. Introduction ………………………….. ………………………….. ………………………….. …………………… 28
3.2. What is a collocation? ………………………….. ………………………….. ………………………….. …….. 28
3.3. Translating collo cations ………………………….. ………………………….. ………………………….. ….. 35
3.4. Collocations for non -native speakers ………………………….. ………………………….. …………….. 41
3.5. Conclusion ………………………….. ………………………….. ………………………….. …………………….. 44
CHAPTER FOUR: THE TRANSLATION OF COLLOCATIONS IN ―TURTLES ALL THE
WAY DOWN‖ BY JOHN GREEN ………………………….. ………………………….. ………………………… 45
4.1. Introduction ………………………….. ………………………….. ………………………….. …………………… 45
4.2. The auth or- John Green ………………………….. ………………………….. ………………………….. …… 45
4.3. The book ………………………….. ………………………….. ………………………….. ……………………….. 46
4.4. Methodology ………………………….. ………………………….. ………………………….. …………………. 47
4.5. Corpus ana lysis ………………………….. ………………………….. ………………………….. ……………… 52

3
4.6. Conclusion ………………………….. ………………………….. ………………………….. …………………….. 55
CHAPTER FIVE: FINAL CONCLUSIONS ………………………….. ………………………….. ……………. 57
5.1. Re commendations and suggestions for future research ………………………….. ………………… 57
REFERENCE LIST ………………………….. ………………………….. ………………………….. ………………. 58
APPENDIX I: QUESTIONNAIRE ………………………….. ………………………….. …………………….. 61
APPENDIX II: CORPUS ………………………….. ………………………….. ………………………….. ………. 63

4

ABSTRACT
The subject matter of the present study is two -fold: on the one hand, translation of collocations in
Turtles all the way down by John Green, and on the other hand, the knowledge students have
about collocations and their translation. Collocations are a major part of any language, and each
word included in a collocation has an impact on the meaning and ―reflect s the linguistic, stylistic
and cultural features o f texts‖ (Al Sughair, 2011: iii). Collocations, as reflected by the theory, are
a cultural aspect that is difficult to translate and define. Taking all these aspects into account we
shall show how complex and diffic ult to translate can sometimes collocatio ns be.
The theoretical background is mostly based on the works of Mona Baker andSusan Bassnett for
the chapter regarding translation studies and on several studies issued over the years on the study
and the translation of collocations. Quantitatively, the analysis and the results have shown that
the most used translation strategy is literal translation, and the second most used is transposition.
The translation of the collocations from Turtles all the way down has been an interesting job, as
Camelia Ghioc (the author of the Romanian translation) has taken the liberty to translate them in
such manner, as to preserve the soul and the mind of John Green.
Key words: translation, collocation, Green, literal, knowledge, learners

5
CHAPTER ONE: INTRODUCTION
1.1. General remarks
Throughout history, there have been several people with different opinions about change.
Aitchison (2001) said that ―e verything in this universe is perpetually in a state of change, a fact
commented on by philosophers and poets through the ages. A flick through any book of
quotations reveals numerous statements about the fluctuating world we live in ‖. Moreover ,
Heraclitus, the philosopher who lived in the VIth century BC , said that ―everything rolls on,
nothing stays still‖ and Edmund Spense r (XVIth century) speaks of ―the ever -whirling wheel of
change, the which all mortal things doth sway‖. Language, like everything else joins the flux.
The German philosopher -linguist Wilhelm von Humboldt noted in 1836: ―There can never be a
moment of true standstill in language, just as little as in the ceaseless flaming thought of men. By
nature it is a continuous process of development‖ . Aitchison said that language change is
―inevitable as the erosion of hills ‖, yet it is not a decay, new vocabulary allo wing language to
respond to newer situations, to adapt itself to the society continuously changing (Aitchison,
2001: 95). After several years during which the discipline known as Translation Studies has not
been taken seriously, in 1990s , translation becam e fundamental act for the human exc hange
(Bassnett, 2013: 1 ). She said that the translator is a traveler, a person who is engaged in the
journey from one source to another (Bassnett, 2013: 1). The growth of translation ‘s field has led
to the study of lang uage and has made the relationship between culture and translation one of the
main interests of linguists. Collo cations and idioms are part of the culture.
According to Wray ( 2005), a big part of our everyday language is ‗formulaic‘, and it comes in
chunk s, fixed or pre -fixed ones, being predictable in form and idiomatic . She also claims that t he
formulaic language, is preferred by the native speakers for an easy communication ; learners rely
on prefabricated language to get started in learning one language , yet they struggle with getting a
hold of them in order to sound native like. Over the years, collocations have been the object of
studies in linguistics and lexicography, in the attempt to define and include them into
dictionaries. According to some auth ors one usage for collocations is translation, due to the
―ability‖ of them to not be characterized ―on the basis of syntactic and semantic regularities‖,
thus not being possible to translate ―on a word by word basis‖ (McKeown & Radev, 1997: 2).

6
1.2. Aim and research question
While theorists have generally decided that collocations are a difficult concept to define or to
include in dictionaries, and they cannot be translated ―on a word for word basis‖, we aim at not
only try ing to give a definition of col locations but also at comparing their translation in the
present day literature.
Therefore, curious about the relation between collocations and translation, and about the
knowledge the students have about them, we have undert aken a study in relation to th ese aspects.
Subsequently, the aim of this research is t o determine how the collocations can be translated and
how much knowledge do students have about collocations. We shall try to answer the following
questions: ‗How can collocations be translated?‘ and ‗Do ESL learners have knowledge about
collocations ?‘
The analysis is carried out on a corpus made up of collocations extracted from Turtles all the
way down by John Green (2017) and on a questionnaire applied to a group of ESL learners,
students at the F aculty of Letters. The book has been written from the perspective of a 16 year
old adolescent that suffers from OCD and anxiety, thus we can imagine, that the difficult part for
the translator, was to preserv e the spirit and the presence of the author .
1.3. Outline of the thesis
The thesis focus es is on translations, collocations and the translation of collocations, as well as
on the non-native speakers knowledge about collocations. Thus, we will approach aspects related
to literary translations in the the oretical framework based on different studies that describe the
work on which this research focuses (the second chapter).
The first theoretical chapter focus es on translation in general (Chapter Two). It is a sort of
introduction to the field of translatio n, and several topics related to this field . It encloses
subchapters about the stages of translations, about translatability and untranslatability, strategies
of translations and the translation of cultural aspects.
The second theoretical chapter (Chapter Three) is centered on collocations and their translation.
It shall focus on the definition of collocations as well as their translation.

7
The fourth and final chapter contain s the actual analysis of the corpus extracted from the book as
well as the analysi s of the questionnaire . The analysis proposed will be both quantitative and
qualitative. Qualitatively, it seeks to explore the translator‘s behaviour and attitude towards the
translation of culture combined with collocation s, while quantitatively it deal s with the statistics
of the result s of the questionnaire.

8
CHAPTER TWO : INTRODUCTION TO TRANSLATION STUDIES
2.1. Introduction
The first chapter of this project deals with theoretical aspects regarding the definition of
translation studies , the types and the stages of translation, the highlights of the translation studies
history, the fidelity and the translatability and untranslatability of some aspects, cultural aspects
and the problems of equivalence.
2.2. What is translation?
Translation is a form of communication where by a message is transferred from a source
language into a target language. If one were to consult a dictionary, the definitions one would get
are:
1. an act, process, or instance of translating: such as a rendering from one language into
another (Merriam -Webster – online dictionary)
2. spoken or written words that have been changed into a different language (Macmillan –
online dictionary )
3. something that is translated, or the process of translating something, from one language
to another (Cambr idge- online dictionary) .
There are different perspectives upon the art of translation. For example, Basil Hatim said that ― it
is not a sterile linguistic exercise but an act of communication ‖ (Hatim, 2014: xi). He also states
that the translator is not ju st practicing the transfer from a source language into a target language,
but he is a mediator of communication (Hatim, 2014) .
Eugene Nida claims that ―Translating consists in producing in the receptor language the closest
natural equivalent to the message of the source language, first in meaning and secondly in style‖
(Nida, 2003: 12) .
Roger Bell offers another definition of translation from Lindsay & Norman (1977) in his book
from 1991:

9
« Traduire c'est énoncer dans un autre langage (ou langue cible) ce q ui a été énoncé dans
une autre langue source, en conservant les équivalences sémantiques et stylistiques. » He
also gives his own translation of this definition: ― Translation is the expression in another
language (or target language) of what has been expre ssed in another, source l angua ge,
preserving semantic and stylistic equivalence ‖ (Bell, 1991: 5) .
J. Catford defines translation as ―the replacement of textual material in one language (SL) by
equivalent textual material in another language (TL)‖ (Catford in Proshina , 2008) .
Newmark says that translation is ―ought to be simple ‖, and that one should be able to say the
same thing in two different languages, but also one can see that translation is ―complicated,
artificial and fraudulent‖ because he claims tha t by using a different language you are turning
into someone you‘re not (Newmark, 1988: 5 ).
Still, there are also some neg ative opinions about translation and some o f them belong to
Lawrence Venut i, who asserts: ―any act of translation is an obvious absurd ity, an effort to go
back in time and restate what used to be an arbitrary state of mind‖ . He also says that translation
is ―an act of terrorism since translation always involves semantic loss and dis traction of the
linguistic form ‖ (Venut i, 1998) .
2.3. Types of translation
Roman Jakobson, in his article ―On linguistic aspects of translation‖ published in 1959
distinguishes three ways of interpreting verbal sign s: they may be translated into other signs of
the same language, into another language or into a nother non-verbal system of symbols.
Consequently, there are three kinds of translation:
1. intralingual translation or rewording – an interpretation of verbal signs by means of other
verbal signs in the same language ; paraphrasing is a type of intralingual translation, and it
is used in order to give explanations or to define things ;
2. interlingual translation or translation proper – an interpretat ion of verbal signs in one
language by means of verbal signs in another language;
3. intersemiotic translation or transmutation – an interpretation of verbal signs by means of
signs of non -verbal sign systems (Jakobson, 1959).

10
Interlingual translati on or translation proper is the transfer from a source languag e to a target
language. Jakobson also states that there is a general problem with any type of translation : ―that
while messages may serve as adequate interpretations of code units or messages, there is
ordinarily no full equivalence through translation ‖ (Jakobson, in Bassnett , 2013: 23) . Jakobson
also claims that th ere are times when the translation may not be possible due to lack of complete
equivalence and the translator might be forced ―to resort to a combination of code units in order
to fully interpret the meaning of a single unit‖ (Jakobson, in Bassnett, 2013: 24). The author also
gives an example in order to demonstrate that: the Russian word ―syr‖ (a food made of fermented
press ed curds) that is translated into English as ―cottage cheese‖, and as a result he claims that
this translation is only ―an adequate in terpretation of an alien code uni t and equivalence is
impossible ‖ (Jakobson, in Bassnett, 2013: 24) .
Bassnet t says that there are term s that have been borrowed from other languages like ―Bon
appétit ‖ from French, which cannot be translated into Englis h so the translated text has to take
into considerations the following in order to determine what to use in English :
1. ―Accept the untranslatability of the SL phrase in the TL on the linguistic level.
2. Accept the lack of a similar cultural convention in the TL.
3. Consider the range of TL phrases available, having regard to the presentation of class,
status, age, sex of the speaker, his relationship to the listeners and the context of their
meeting in the SL .
4. Consider the significance of the phrase in its particular context —i.e. as a moment of high
tension in the dramatic text.
5. Replace in the TL the invariant core of the SL phrase in its two referential systems (the
particular system of the text and the system of culture out of which the text has sprung) ‖
(Bassnett, 2013:31 ).
The translator has to choose the terms in the target language in order to preserve the spirit of the
source language .
2.4. Stages of translation
Marianne Lederer (2014) asserts that the best way to see the process of a translation is during
conference interpreting, and considers that conference interpreting is the best situation in which
one can observe ―how the concrete manifestation from a speech or a text passes through a
translator‘s mind and becomes another concrete manifestation, thereby t ransferring sense‖
(Lederer, 2014: 6) . Lederer quotes Cary w ho claims that ―only the spoken word contains the

11
fullness of human language and it is a mutilation to focus one‘s interest only on what the printed
page can hold [ …]‖ (Cary in Lederer, 2014:7) su ggesting that the interpreter has to deal both
with the thoughts and with their expression . Lederer also quotes Cary‘s opinion regarding the
written code :
―the written text – the only one we have access to – is but a mummy, a faulty and
fragmented copy of the author‘s living thoughts […] when we translate, do we not
sometimes wonder: ‗What does he really mean, this author from whom I am only getting
a partial message […]‖ (Cary in Lederer, 2014: 8)
Cary also believes that the speeches give out their meanin g far easier than written texts, the
author and the receiver of the written texts being connected only by the form , leaving room for
different interpretations. (Cary in Lederer, 2014: 8)
Lederer claims that there are three stages involved in the process of translating.
Stage I . Understanding the meaning: The progressive process applied directly on text, and with
the participation of extra linguistic knowledge, that is part of the translation cognitive universe .
Understanding the meaning is ―arriving at a me ntal representation‖ (Lederer, 2014: 13). Lederer
says that Sperber and Wilson (1986) assume that there are two stages of understanding the texts.
―The first stage is understanding the text‘s language; the second consists of ‗inferring‘ the sense
with hel p from extra -linguistic knowledge ‖ (Lederer, 2014: 16) . Yet, a text is not understood in
phases, first at the level of language, then at the level of discourse but concomitantly . The
general concept it is not observed separately and previously from the spe cific concept. (Lederer,
2014: 16 )
Stage II . Deverbalization: This stage consists in capturing the more or less explici t intentions of
the source text, this being the immediate effect of understanding the meaning. One can remember
the meaning that was rend ered but cannot remember the exact words that were used to render
that message. Deverbalization is less apparent in the case of the translation of written texts, then
in the case of interpretation but it is present.
Stage III . Reverbalization . This stage requires involvement on the part of the translator, a
detachment of the source language constraints and an attempt to find an expression that should
somehow be at the same time faithful to the original text and to the ordinary usage of the target
language. Delisle says that as long as the meaning is understood, its re -expression, the

12
reverbalization is done based on what it is understood, not on the words that are used (Delisle in
Lederer, 2014: 36 ).
Although it is said that there are three stages of transl ation, there is still one more to be taken in
consideration and that is the verification stage . After completing his work , the translator must
check that the sentences are suitable to the target language community; the sentences shou ld not
follow the struc tures and forms of the original . Even though the main ideas may have been
rendered , the translator must be sure that he was creative enough in order for the reader to
experience ―the author‘s ideas and feelings‖ completely and the expression of ideas to be
understood. The aim of this stage is to check that the equivalent is suitable to render the original
meaning. (Lederer, 2014: 38-39)
2.5. Highlights in the history of translation
Translations have been important since ancient times and that can be proven through the
authority of historical documents. The great civilizations of antiquity made use first of them. The
first traces date from the year of 300 BC during the Egyptian old Kingdom.
In Europe, the first translator ‘s name to be recorded was Livius An dronicus, a Greek slave who
in the year of 240 BC translated the Odyssey into Latin verse. The f lourishing of the great
civilizations from the antiquity coincides with periods of intense translating activities. The
Greek s got acquainted to oriental texts. In 300 BC the Romans took over many elements of
Greek culture as well as of Greek religious system.
The golden age of Latin literature was the age of translations and it is from this wide spread
practical activity that the first translations norms emerged . Thus Cicero in his De oratore favored
free translation trusting the artistic potential of his own language as well as the creative gifts of
the translator.
Quintilian is the first writer who perceived the elements of ― contest‖ between the translated
version and the original. In the ninth century , in Britain , King Alfred the Great was the first
important translator of Latin works into English.
It is necessary to mention that for each and every culture, religious texts represented the first
texts to be ev er translated.

13
In Romania, practical needs as well as favorable historical conditions, gave rise to the first wave
of translations. In the period of emancipation from the Slavonic language used in church
services, important philosophical and religious books appeared following Greek models.
In 1559, Deaconu Coresi translated ― Întrebare cre știneasc ă‖ which is regarded as the first, still
available Romanian translation. In 1648 bishop Simion Ștefan translated the New Testament into
Romanian while the first complete of th e Bible, known as ― Biblia de la Bucure ști‖, undertaken in
1688 and alt hough it was written in a language which was not yet under the control of rational
grammar, it still uses a unitary language and shows remarkable literary qualities for that time.
The first Romanian poet translator was Dosoftei, who was also bishop of Mold avia and whose
translation of the ―Psalms of David‖ was a highly influential text appreciated for its aesthetic
qualities.
2.6. The possibility versus the impossibility of translations
For many centuries, translation theorists approached a number of issue s structured as a series of
oppositions closely related to each other:
1. possibility vs. impossibility of translations
2. faithful vs . unfaithful translations
3. literal vs . free translations
4. source language (culture) oriented translations vs . target language(cul ture) oriented
translations
In regard to the possibility versus the impossibility of translations arguments on both sides have
always existed.
The myth and story of the Tower of Babel has always fascinated translators and the students of
translation. There are several authors that ―have wanted to believe in a primordial language that
was lost in the scattering of tongues at Babel and that might be regained through perfect mystical
translation ‖. Therefore, in their perspective, the translator is seen as a ―W orld Savior ‖ who
restores the ―original linguistic unity‖ smashed by th e gods on the plains of Shinar (Robinson in
Baker, 2005) . It has been interpreted as an explanation of the beginning of translations or as a
proof and warning that translations are doom ed to failure. It is also said that the translator is a

14
―traducer ‖ because s/he ― fails to restore us‖ to the state where the translation was not necessary.
Robinson also says that once there was spoken a single language, and that the histor y should
repeat and everybody speak a single language again, a Lingua Franca : Esperanto or English
(Robinson in Baker, 2005) .
The arguments for untranslatability are mostly situ ated at the level of vocabulary. There is a
theory according to which language precedes though t, meaning that each language offers to the
generations to come a ready -made interpretation of reality which is specific to each language.
This theory is called linguistic determinism . An example that supports the theory is the fact that
the Eskimos, for i nstance, use tens of terms for sno w and Arabic people distinguish about 20
terms for camel. Additionally, while most languages distinguish about seven colors in the
spectrum, in Hebrew there are no separate terms for blue. Red, white and black are dissocia ted
but there is only one term for yellow and g reen. The idea is that there exist s a lack of symmetry
between languages at the level of vocabulary, which may be an argument for untranslatability.
There is also, the case of untranslatable terms that relate to a specific geographic, historical,
social -cultural experience and these has been always a challenge for translators. In Romanian,
for instance ―dor‖, ―plai‖, ―tain ă‖, ―spa țiu mioritic‖, in English ―gentleman‖, ―understatement‖,
French: ―charme‖, ―esprit‖ .
―Translatability is the capacity for some kind of meaning to be transferred from one language to
another wi thout undergoing radical change ‖, says Anthony Pym. He also says that the basic
problem for or against translatability is the relation b etween the source -text ―expressions‖ and
―meanings‖ that are maintained in the source language (Pym in Baker, 2005) .
There are also supporters who trust completely in the poss ibility of translation. Given the
universality of languages and experience, they say that general meaning can be rendered in any
language. They basically say that reason, thoughts and even culture have this universal trait
which makes possible the inter -cultural and inter -linguistic communication. There are authors
who put the terms of translatability on the target language, i.e. that the translatability of a
meaning depends on how is that meaning found in the target language and the equivalents for
that te rm. There are terms that may not be found in a language, due to the lack of knowledge of
that concept in the culture and society. In that situation that term will be translated through
paraphrasing, or explanations and definitions of that term.

15
2.7. The fidelity of translations
The issue of fidelity of translation lies at the core of the traditional translation discour se, and it
has been debated for 200 years , during this time favoring either translation in the spirit of the
original o r to the letter of th e original. European historians of tr anslation provided many
examples of historical periods or translations schools that side with one direction or the other. It
all comes down to the distinction between ―word for word‖ and ―meaning for meaning‖
translatio n.
Many scholars consider literal translations ungrammatical, other too grammatical, so it triggers
extreme positions. In the middle age s, translations had the tendency towards literal, the reason
being theological, more exactly the desire to preserve the integrity of the original text, also
because the word of God was sacred a nd unique so it had to be translated literally . On the other
hand, in the XIIth century translators adopted literalism, because if they changed even slightly the
meaning, their trans lations could have been censored or even banned. In the XVIIth century the
German school adopted it as a mean of invigorating the German language and as the only way of
allowing the readers a glimpse of what the original work looks like. Another reason for choosing
literary translations is related to the translator‘s ethics: the translator should be invisible proving
this way the devote dness to the original text and author. However, the same ethics can have a
different interpretation and can display the tr anslator as a powerful voice, wr iting and obviously
translating, constantly reminding the readers that what they are reading is a translation. Another
reason for translating literally may be connected to the lack of confidence in the aesthetic
capacities o f the target language, which is felt to be less suggestive than the source language.
Peter Newmark is the one who called literal translations, ‗semantic translations ‘ and he
enumerates some of the characteristics :
 the importance of the form of t he origina l;
 the attempt to pr eserve the author‘s idiolect and the flavour and tone of the text ;
 can be usually applied in texts which are predominantly expressive in function and in
which form and meaning are tightly linked to each other (drama, autobiographical book
and literary texts in general) (Newmark, 1988)
Literal translations are considered to be source language oriented, source culture oriented.

16
The communicative translation should be perceived in terms of a convention with the audience.
Since the trans lations can never be identical, it should a t least have a similar effect upon the
readers and give them the illusion of an independent value. This idea is present in Eugene Nida‘s
theory of dynamic equivalence , also referred to as functional equivalence . Dynamic equivalence
is defined as ―the degree to which the receptors of the message in the receptor language (target
language) respond to it in the same manner as the re ceptors in the source language‖ (Nida, 1964) .
The response is never identical due to the cultural and historical s ettings that are too different;
still, there should be a high degree of equivalence of response, otherwise the translation will have
failed to accomplish its purpose. It is very important to mention that the more different two
languages and cultures are, the better the conditions for achieving dynamic equivalence.
Some of the characteristics of this type of translations are:
 being smoother, clearer and sim pler for readers to understand;
 transformations occurred both at the linguis tic and at the cultural level ;
 they display creativity ;
 they are time bound (it is possible that after a period of time they would not be any longer
to the taste of the public).
Target -oriented ness has always been regarded as a better translation than the literal tendencies.
The best-known examples of communicative translations were created in t he XVII – XVIIIth
centuries ―les belle infidèles ‖ were created in France and they were translations of classical
works that were linguistically and culturally adapte d to the taste of the target contemporary
readers. However, there are times when the distance between the source and the target texts
exceeds the limits of communicative translations. Thus, in their wish to satisfy the meanings of
the target community and to give free way to their creativity , translators may resort to free
adaptations and imitations and by these means, the translator betrays the source text, and use
them as a pretext of producing texts of their own. For example, Ezra Pound who translated fr om
Chinese without knowing the language or Tudor Arghezi‘s who translated from Baudelaire
without following the content of the original poems.
Nowadays, many contemporary trends in translation studies seem to swing the balance in the
target direction with out losing the sight of the source text. In fact, the source -target dilemma is

17
usually in doubt dramatically and a wide range of contextual factors account for the translators
preference for one method or the other.
2.8. Translation strategies
In 1958, two translation collars named Jean Paul Vinay and Jean Dabelnet published a
significant study entitled ―Compared stylistics of French and English‖ which became a classic in
the field of translation studies. This volume gives future translations a chance to get familiar with
the complex operations that are involved in translating and with the various ways in which
different languages cut out reality. In the authors‘ opinion there are two directions translators
may choose in their work : that of direct/literal decoding techniques and that of the
indirect/oblique one. The choice of direct procedures creates a direct correspondence between
the two languages; in this case translating does not involve a semantic or grammatical
restructuring. As for the indirect procedures, they generate complex restructuring of the
translated units; the changes may vary from switch of grammatical clauses to completely
different perspectives on reality.
Krings (1986:18) defines translation strategy as ―translator‘s potentially conscio us plans for
solving concrete translation problems in the framework of a concrete translation task‖, Loescher
(1991:8) defines it as ―a potentially conscious procedure for solving a problem faced in
translating a text, or any segment of it‖. Venuti (1998: 240) indicat es that translation strategies
―involve the basic tasks of choosing the foreign text to be translated and devel oping a method to
translate it‖.
1. Direct/literal translations
In this case, no formal or conceptual restructuring of extra reality is necessary. There are three
direct translation procedures:
a) Loan borrowing, also called loan transfer : the use of loan implies a transfer of the
source language term into the target language because the target language has no
equivalent for it. Another reas on is that loans can render the exotic atmosphere of the
source language. Also, loans may be chosen by translator so as to convey a sound effect,
or in some cases even brand names can be borrowed for economic reasons

18
b) Loan translation or calque : the source language components are preserved as such, but
they are literally translated , in other words, calques are literal translations at the level of
the phrase. Thus, the English compound word ―superman‖ becomes ―super om‖ in
Romanian, ―The European Union‖ becam e ―Comunitatea Europeană‖ .
c) Literal translation : this implies a one -to-one transfer of the language structure, which is
possible in the cases of direct word correspondence, and identical syntactic structure.
This is the case of ―perfect linguistic equivalen ce‖ which is seldom met and should be
double -checked .
2. Indirect translations
The four indirect translation procedures are:
a) Transposition: it consists in replacing the grammatical clause of the items in the source
language with another grammatical clause in the target language without altering the
meaning of the message. For instance: verbs were replaced by no uns: ―we deliver‖ →
―livrări la domiciliu‖; adverbs replaced by verbs: ― he merely nodded‖ → ―se mul țumi s ă
încuviin țeze numai‖; adjectives were replaced by nouns: ―in the early XIXth century‖→
―la începutul secolului al XIX -lea‖; adverbs were replaced by n ouns: ―it is popularly
supposed that‖→ ― lumea crede că ‖; verbes replaced by prepositions: ―reports indicate
that‖→ ―potrivit rapoartelor‖
b) Modulation: it is more complex than the transposition because it involves reshaping of
the semantic perspective; more exactly it determines sh ifts in focus and point of view.
For instance: ―to take French leave ‖→ ―a o șterge englezește ‖, ―to go dutch ‖ → ―a plăti
jumi-juma ‖, ―as white as the sheets‖→ ―alb ca varul ‖, ―as good as gold‖→ bun ca p âinea
caldă, ―dress rehearsal‖ → ―repetiție generală‖
c) Equivalence : in a narrow sense it implies the preservation in the target language of the
source language situation by using completely different structural and stylistic means. It
involves the message as a whole and not only part s of it. It may occur in the translation of
idioms, proverbs and sayings. For instance: ―Lupul își schimbă blana dar nă ravul ba‖ →
―The fox may go gray but never good‖, ―a ochi doi iepuri dintr -un foc‖→ ―to kill two
birds with one stone‖, ―nu se auzea nici mu sca‖→ ―you could hear a pin drop‖, ―la
paștele cailor‖ → ―when the pigs fly‖

19
d) Adaptation: it may occur whenever the source language situation is non -existent in the
target language and has to be replaced by another situation. For instance, the situation in
Romania of kissing a lady‘s hand out of respect; there are two possible translations:
preserving the gesture and saying ―he kissed her hand‖ together with a footnote
explaining the significance of the gesture in Romania; or the second possibility would be
replacing all together this gesture by a similar respectful gesture in the English culture
like ―raising one‘s hat‖.
Newmark (1988b) claims that there is a difference between translation methods and translation
procedures. He writes that, ―[…] while tran slation methods relate to whole texts, translation
procedures are used for sentences and the smaller units of language ‖ (Newmark, 1988b: 81).
2.9. Problems of equivalence
Equivalence is a controversial central concept of translation studies. Kenny Dorothy claims in
her article published in Mona Baker‘s Encyclopedia of Translation -Studies that different authors
approach differently the top ic of equivalence. S ome of them, like Nida, Catford, Pym, ―define
translation in terms of equivalence relations ‖ while ot hers repudiate the notion of equivalence
asserting that it is either ―irrelevant‖ or that it is ―harmful‖ for translation studies (Kenny in
Baker, 2005 : 77). Yet, Kenny claims that Baker is somewhere in between all these beliefs and
that she is using the t erm of equivalence ― for the sake of convenience ‒ because most translators
are used to it rather than because it has any theoretical status‖ (Baker, 1992 : 5-6). Baker also
claims that ―although equivalence can usually be obtained to some extent, it is influenced by a
variety of linguistic and cultural f actors and is therefore always relative‖ (Baker, 1992: 6). Fact
that is also stated by Kenny who says that ―equivalence is variously regarded as a necessary
condition for translation an obstacle to progress in translation studies or a useful category for
describing translations‖ (Kenny in Baker, 2005: 77) . Popovič distinguishes four types of
equivalence in his definition of translation equivalence:
―1. Linguistic equivalence , where there is a homogeneity on the linguistic level of both
SL and TL texts, i.e. word for word translation
2. Paradigmatic equiva lence , where there is equivalence of ‗the elements of paradigmatic
expressive axis‘ i.e. elements of grammar, which Popovič sees as being a higher category
than lexical equivalence

20
3. Stylistic (translational) equivalence , where there is ‗functional equiva lence of elements
in both original and translation aiming at an expressive identity with an invariant of
identical meaning‘.
4. Textual (syntagmatic) equivalence , where there is equivalence of syntagmatic
struct uring of a text, i.e. equivalence of form an d shape‖ (Popovič in Bassnet , 2005: 33)
(Bassnett ‘s interpretation of Popovič‘s words ).
Nida claims that there are two types of equivalence, formal and dynamic. Formal equivalence
―focuses attention on the message itself , in both form and content. In such a translation one is
concerned with correspondences as poetry to poetry, sentence to sentence, and concept to
concept‖ . Nida names this type ―gloss translation‖ where the translator tries to make the TL text
as understandable and maintaining the form and content of SL text. Dynamic equivalence is
based on the principle of equivalent effect, i.e. that TL text transmit the same message as the SL
text, thusly trying to maintain the same relation between the receivers and TL message as in the
case of the origi nal receivers and SL message . ―A translation of dynamic equivalence aims at
complete naturalness of expression and tries to relate the rece ptor to modes of behavior relevant
within the context of his own culture ‖ (Nida, 1964: 159).
Bassnett claims that e quivalence ― overall results from the relation between signs themselves, the
relationship between signs and what they stand for, and the relationship between signs, what they
stand for and those who use them‖ (Bassnett, 2005: 3 5). She is backing up her a ssertion with an
example regarding the translation of the shock value of blasphemous expressions from Spanish
or Italian, rendering this value being possible by the use of ―substituting expressions with sexual
overtones to produce a comparable shock effect (e .g. porca Madonna ‒ fucking hell )‖ (Bassnett,
2005: 36 ). James Holmes (in Bassnett, 2005: 36) considers that the use of the term equivalence is
―perverse‖, since ― to ask for sameness is to ask too much‖ . Bassnett states also that in
translation, equivalence ―should not be a pproached as a search for sameness, since it cannot even
exist between two TL versions of the same text, let alone between the SL and the TL version‖
(Bassnett, 2005: 37 -38). In addition, she claims that Popovič‘s four types of equivalence offer a
base to wards an approach that sees it as ―a dialectic between the signs and the structures within
and surrounding the SL and TL texts‖ (Bassne tt, 2005: 37 -38).

21
2.10. Cultural aspects of translation
According to Katan, not everybody agrees about ―the existence or relevance of cultural
differences in translation‖ (Katan in Munday, 2009: 74). He also says that the re are thre e
interconnected problem areas. The definition of culture is the first area of controversy. Initially,
according to Katan, culture was simple. I ts first meaning was referring to ―the humanist ideal of
what was civilized in a developed society‖ (Katan in Munday, 2009: 74); then a second meaning
has occurred: ―the way of life of a people‖ (Katan in Munday, 2009: 74) . The second
controversial area is related to the historical division between people who see language and
culture as two separate entities and those who perceive language as culture. The first category
sees translation as ―a universalist encoding -decoding linguistic activity, transferring meaning
from the SL to the TL‖ (Katan in Munday, 2009: 75 ), where ―culture and any cultural differences
can be carried by the language without significant loss‖ (Katan in Munday, 2009: 75). On the
other hand, people from the second category, like Nida , believe that ―the context actually
provides more distinction of meaning than the term being analyzed‖ ; therefore the meaning is
bargained between readers within their contexts of culture, not being ‗carried ‘ by the language
(Katan in Munday, 2009: 75). The th ird area is related to the other two controversies and is the
―importance of the ‗culture filter‘ in translation‖ (Katan in Munday, 2009: 75). The cultural
filter is related to the ―ways in which groups organize their shared […] perception of the world‖
(Katan in Munday, 2009: 75). According to Katan, House (2006) defines cultural filter as ―a
means of capturing cognitive and socio -cultural differences‖ that are to be applied by translators.
Katan says also, with the help of others before him, that the tas k of a translator is ―to transfer
terms and concepts in the source text abroad with the minimum los s, so that ‗what you get‘ in the
source text is equivalent to ‗what you get‘ in the target text‖ (Katan in Munday, 2009: 78- 79).
As stated by Katan, if two cultures ‗have reached a comparable degree of development‖ , there is
no justification for which ―the meaning, reader response and uptake‖ to not be universal (Katan
in Munday , 2009 : 79). New ark is calling this ―the cultural value‖ of translation: ―to assi st in the
spreading of culture throughout the world‖ (Newmark, 1981: 184 -5).
Culture is defined by Peter Newmark as ―the way of life and its manifestations that are peculiar
to a community that uses a particular language as its means of expression‖ (Newma rk, 1988: 94 ).
In addition, he does not perceive language as ―a component or feature of culture‖ because he

22
believes that if it were so, the translation would be impossible. He says that language does have
imbedded some cultural aspects in grammar (genders of inanimate nouns) , in lexis and forms of
address . He also claims that if a text is becoming more specific about natural phenomena , ―the
more it becomes embedded in cultural features‖ and this leads to translation problems. Newmark
states that the hardes t words that are to be translated are the more general ones like: ―love,
temperance, temper* right, wrong ‖. As well , he says that the customs are difficult to translate
literally because this would distort the meaning due to the use of ―ordinary l anguage ( ‗topping
out a building‘ , ‗time‘, ‗gentlem en‘, ‗please1 ‘, ‗mud in your eye ‘)‖ for their description
(Newmark, 1988: 95).
―Translation is a kind of activity which inevitably involves at least two language s and two
cultural traditions‖ (Toury , 1978:20 7). As Toury implies, a translator has to make sure that he
conveys the cultur al aspects from the source text, in the most appropriate way, in the target
language. Yet either a translator may subjugate himself/herself to the source text‘s norms or to
the norms active in the target culture ; this means if the translator chooses the first instance , ―the
translation will tend to subscribe to the norms of source text, […] language and culture‖ , this
depicting ―the pursuit of adequate translation‖. On the other hand , if the second stance is
adopted , the norms of the target language/culture are set off and ―subscription to norms
originating in the target culture determines […] acceptability ‖ (Toury, 1978: 208) .
As each community, each people have different beliefs, and each country may have different
beliefs in every corner; and therefore it is important for the translators to search information of
the source language before communicating by transferring the message with the underlying
concepts of culture. Marianne Lederer says ―foreign literature which calls on human universals
has the potential to be understood by any reader‖ , and that the translator has to provide to the
readers from the target culture/language ―additional pieces of knowledge […] to unlock the do or
which leads to the knowledge‖ of the social norms and traditions evoked by foreign literature
(Lederer, 2014: 127) . By translating cultural aspects, the translator helps the readers to complete
their knowledge they were lacking in the beginning. There is not actually a unique, general
solution to transfer culture , and ―the relevant solution is always ad hoc and depends heavily on
the portion of text to be translated‖ (Lederer, 2014: 128). Bassnett quotes Lotman who says that
―No language can exist unles s it is steeped in the context of culture; and no culture can exist

23
which does not have at its center, the structure of natural language ‖ (Lotman in Bassnett, 2005:
23). Therefore, Bassnett continues and says that language is ―the heart within the body of
culture‖ and that the relationship between language and culture re sults ―in the continuation of
life-energy‖ . Moreover, she says that ―the translator treats the text in isolation from the culture at
his peril‖ (Bassnett, 2005:23). Some foreign realities ar e causing trouble to the translator s that
are transferring culture.
1. Idioms are lexical items consisting of several words, with a meaning that cannot be
deduced from individual words. Not all the idiomatic phrases are difficult to understand,
but the probl em idioms are the ones composed of simple words, each of which we (as
foreigners) understand.
e.g. He was beside himself. → ‗beside oneself‘= having lost one‘s self -control because of
the emotion.
2. Words with several meanings : any word that has more than one meaning is potentially
troublesome. Some of the most dangerous misunderstandings arise when everyday words
are used in specialized fields .
e.g. t he mathematician‘s use of ―argument‖ = demonstrație, parametru, variabilă
independentă
the statistician‘s etc. ―random‖ = aleatoriu, stocastic, arbitrar, accidental
the communications engineer‘s ―noise ‖ =paraziți, perturbații
The difficult part about translating this type of words is that i t is not immediately
apparent . To tackle these unexpected misunderstandings, the translator must first be
aware that they may occur and then he uses common sense in d eciding whether to accept
a familiar meaning or check whether another is possible.
3. Sub-technical vocabulary: Troublemakers are words like ―average‖, ―approximate‖,
―effect‖, ―combination‖, ―determine‖. They are needed in most fields of study and are
theref ore worth attention. The problem these words raise is conceptual rather than
linguistic.
4. Transfer of meaning: Metaphor and similar kinds of transferred meaning raise
problems at the level of the meaning , i.e. it is not obvious. Metaphor and similar
kinds of transferred meaning are always potential problems. Like idioms, they do not
mean what at first glance they seem to mean. This is not only the case when reading

24
literature, but e.g. a student of accountancy may be puzzled by ‖galloping inflation‖,
―profi ts wiped out‖, ―fringe benefits‖, etc. Metaphor may lie in a single word or be
far more extensive (and so not really a lexical problem at all). It always involves an
implicit comparison between A and B, so one method of handling it is to analyze
what A and B have in common that is relevant to the context. The problem of
interpretation is increased when the metaphorical term is used as a form of elegant
variation. The reader must not only understand the metaphor as such, but also
recognizes that it has the s ame referent as other non -figurative terms.
5. Text-structuring words are words that are used as signals telling the reader to fill out
their meaning, usually from info elsewhere in the text.
e.g.: The issue will not be resolved by such methods .
The meaning is not clear unless the issue and the methods can be identifi ed. Sometimes
text-structuring words relate to background knowledge the reader is expected to share:
e.g.: Recent events suggest that Brown‘s views are correct.
We need to know what events are referred to and what Brown‘s views are.
Text-structuring words can refer forward as well as back (cataphoric vs. anaphoric
reference).
e.g. Various explanations of the phenomena can be offered.
While ‗ phenomena‘ relates back to something just described, ‗ explanations ‘ signals
what will probably follow. Being able to fill out the meaning of such words is a
crucial skill for interpretation of text.
6. Irony is more a problem of pragmatics rather than lexis . The words may be simple,
but the way the writer uses t hem is not. The difficulty is the mismatch between the
apparent meaning and the writer‘s hidden intention. It is one of the most difficult
uses of language. The only way to handle it seems to be at the level of a whole text,
since interpretation depends on assessment of the writer‘s atti tude from other
evidence in the text. By using irony, the speaker is echoing a point o f view in order
to display some attitude towards it. In translation , it is sometimes difficult to reach

25
the degree of irony perceptible in the source text, even if the propositional content is
faithfully reflected. Recognising an ironic intention is crucial for a translator and it
will condition his output.
7. Advertisements: The issue of translatability has always been a debate regarding the
translatability of the poetic discourse as part of translating within the communication
process. Roman Jakobson considers that, due to the nature of the poetic discourse,
translation proper is impossible, and that ‗only creative transposition is possible‘
(Jakobson, 1959: 238). This is true far beyond poetry: in any discourse whose form
of the language code bears some significance: advertising, political slogans, puns
(there are translators hired by advertising companies to translate ads for
multinational co mpanies).
8. Slang: It is generally speaking, representative for the spoken language. The slang has
the following characteristic s: to be poetic or creative ; it appeared with the desire to
be secretive; it is a form of expressiveness; it is believed that it i s used to express
group membership; the use of slang can also show relaxation. Some people believe
that using slang they appear to be up -to-date or clever. Even so, not all slang is
equally acceptable in all social situations . In order to preserve the effe ct and
especially not to affect the social context , slan g should be translated by slang, for
which the translator needs cultural knowledge . For example for ‗ money ‘ there is
dough, lace, lettuce, the green, stuff, sawbuck (for a ―ten‖) , double -sawbuck (for a
twenty) , grand (for a thousand) → Ro: verzișori, mălai, marafeți, cașcaval, lovele,
biștari ; for police officers: bobby, dick (idiot, cretin), pig, scuffer, filth, plod → Ro:
polițai, șacal, copoi, curcan, gabor, sticlete . Colloquialisms are words with different
meanings for different pe ople. Slang can be accepted as part of colloquialisms; the
difference between them being that slang usually denotes words, while
colloquialisms are significant for words, phrases, sentences. Cliché s are slang
expressions that have lost their color in time
9. Jokes, puns : Puns represent the humorous use of a word that has two meanings or of
different words that sound the same. The translation of puns is highly difficult if not
impossible since the humorous effect gets lost, the joke entailing rather an

26
explanat ion than a translation proper. Thus, the initial intention can usually not be
preserved in the target language.
Katan claims that translation as ―intercultural communication requires treating the text itself
as only one of the cues of meaning ‖ (Katan in M unday, 2009: 91). He also claim s that there
are some factors that are ―silent‖, ―hidden‖ and ―unconscious‖ that determine how the text
will be understood . Through translation, a new text comes to life, text that will be read
―according to a different map o r model of the world, through a series of different set of
perception filters‖ and therefore the translator needs to meditate ( Katan in Munday, 2009:
91, author‘s emphasis) .
According to Harvey, who defined culture -bond terms as the terms which ―refer to concepts,
institutions and personnel which are specific to the SL culture‖ (p.2), there are four major
techniques for translating them:
1. Functional Equivalence : it refers to using a referent in the TL culture whose function is
similar to that of the sou rce language (SL) referent.
2. Formal Equivalence or „linguistic equivalence ‟: It means a 'word -for-word' translation.
3. Transcription or ‗borrowin g‘ (i.e. reproducing or, where necessary, transliterating the
original term): it stands at the far end of SL -oriented strategies. If the term is formally
transparent or is explained in the context, it may be used alone. In other cases, particularly
where no knowledge of the SL by the reader is presumed, transcription is accompanied by
an explanation or a transla tor's note.
4. Descriptive or self-explanatory translation: It uses generic terms (not CBTs) to convey
the meaning. It is appropriate in a wide variety of contexts where formal equivalence is
considered insufficiently clear. In a tex t aiming towards specia lized audience , it can be
helpful to add the original SL term to avoid ambiguity. (Harvey in Ordudari, 2007)
2.11. Conclusion
Taking into account all the theoretical aspects in this chapter, we have covered topics that
are yet to be finished, as every year more studies on these topics are published . One thing

27
that a translator must bear in mind when working on a translation is t o fit the translated text
in the target language community, and to preserve as much as possible the soul of the
original text‘s au thor.

28
CHAPTER T HREE : THE TRANSLATION OF COLLOCATIONS
3.1. Introduction
This chapter deals with collocations and particularities regarding the defining of
collocations, their translation and the struggle these combinations represent to learners . The
first subchapter contains information about the theor etical aspects of collocations, in
particular the definition of these combinations of words as well as their classification and
theories brought by different authors along the years. The second subchapter deals with the
translation of collocations and the opinions of different authors about the approaches needed
to be considered when a translator is to translate collocations. The final subchapter talks
about different authors opinions about the learning of co llocations and their importance in
the study and the understanding of the foreign language.
3.2. What is a collocation?
A collocation consists of two or more words that often go together. Th ese combinations
sound natural to native speakers of English, but they are rather difficult for non -native, due
to their difficult to guess meaning. Each word in a collocation con tributes to the overall
meaning of the compound , collocation . If it were to check a dictionary, it is possible to get
similar yet differ ent definitions for collocation:
1. The habitual juxtaposition of a particular word with another word or words with a
frequency greater than chance (Oxford Online Dictionary) ;
2. A noticeable arrangement or conjoining of linguistic elements (such as words)
(Merriam -Webster Online Dictionary) ;
3. the fact that a word collocates with other words (Mac millan Online Dictionary)
4. The combination of words formed when two or more words are often used together
in a way that sounds correct (Cambridge – online dictionary ).
Accordin g to Gyllstad , the first time the term was used , yet not as a linguistic term, was in
the 17th century, by Francis Bacon in his Natural History in 1627 . As a linguistic term was
used for the first time a century later , by Harris ―who used it to refer to th e linear

29
constellation of words‖ (Palmer, 1933 in Gyllstad, 20 07: 6). As Gyllstad states, Mel′čuk
defines collocations as:
―A COLLOCATION AB of language L is a semantic phraseme of L such that its
signified ‗X‗ is constructed out of the signified of one of its two constituent
lexemes —say, of A —and a signified ‗C‗ [‗X‗ = ‗ A + C‗] such that the lexeme B
expresses ‗C‗ only contingent on A ‖ (Mel′čuk, 1998:30 in Gyllstad, 2013: 14 ).
Gyllstad continues by saying that the formulation ―B expresses ‗C‘ only contingent on A‖
encloses four subtypes of colloca tion:
―a) Collocations con taining a delexical (or ‗support‗, ‗light‗) verb (e.g. give a look,
launch an appeal );
b) Collocations containing a dependent lexeme meaning which only occurs with one
or a few lexemes (e.g. black coffee, French window );
c) Collocations containing a depe ndent lexeme meaning (intensifiers) that can be
used together with other lexemes in the same sense, but its meaning cannot be
expressed by a possible synonym (e.g. strong coffee );
d) Collocations in which one lexeme is dependent on the other lexeme becau se the
meaning of the latter is utterly specific (e.g. the horse neighs, rancid butter )‖
(Gyllstad, 2007: 14).
Furthermore, Gyllstad claims that Kjellmer views collocations as frequency -based , but he
considers that he implies certain syntactic structures in his analyses. Kjellmer defines a
collocation as ―a sequence of words that occurs more than once in identical form [in a text
corpus] and which is grammatically well -structured‖ (Kjellmer, 1987 in Gyllstad, 20 07: 16 ).
He claims also that the kinds of col locations retrieved based on this definition are ― to be,
had been, one of and United States‖ (Gyllstad, 2007: 16).
Jiang Ming -Zhuan says that a collocation ―[…] is a recurrent combination of words that co –
occur more frequentl y than expected by chance ‖ (Jiang et al. , 2012: 246 ). According to
Jiang and his colleagues, c ollocations can be classified according to different criteria: for
example, ― into lexical and grammatical by the nature of their constituents ‖ or according to
word positions in order to disting uish between ―rigid collocations and elastic collocations. ‖
Habitually , a collocation consists of a base word and a collocate and considering that
collocations are broadly used , ―knowing the right collocate for the base word plays an

30
important role in seco nd language learning as well as in machine translation‖ (Jiang et al.,
2012: 246).
Frank Smadja (1993) claims that ―natural languages are full of collocations, recurrent
combinations of words that co -occur more often than expected by chance and that
corres pond to arbitrary word usages‖ (Smadja, 1993: 14 3). He also says that collocations
vary extensively ―[…] in the number of words involved, in the syntactic categories of the
words, in the syntactic relations between words, and in how rigidly the individual words are
used together‖ (Smadja, 1993: 144).
Smadja claims that the most exhaustive definition is in the lexicographic work of Benson
and his colleagues (Benson, 1990): ―a collocation is an arbitrary and recurrent word
combination ‖ (Benson in Smadja, 1993). Regardless, he claims that the definition ―does not
cover some aspects and properties of collocations that have consequences for a number of
machine applications‖ (Smadja, 1993: 146) . As a result, Smadja identifies four properties of
collocations:
 are arbitrary : he believes that as collocations have this property, ―they must be
readily available in both languages f or effective machine translation‖ (Smadja, 1993:
146).
 are domain -dependent : according to Smadja, domain specific collocations are
numerous, and technical jargons being often totally incomprehensible for the non –
expert people , as they contain a large number of technical terms and for the familiar
term there is the tendency to be used differently ;
 are recurrent : this property refers to the fact that these combinations are not
exceptions, but quite the opposite that they are perpetually repeated in a given
context . For instance, word combinations like ‗ to make a decision, to hit a record, to
perform an operation ‘ are characteristic to language, a nd combinations like ― to buy
short," "to ease the jib ‖ are typical of specific domains, and are repeatedly used in
specific contexts;
 are cohesive lexical clusters : by cohesive cluster, they refer to the fact that ―the
presence of one or several words of the collocations often implies or suggest the rest
of the collocation‖ (Smadja, 1993: 146).

31
Smadja has also come up with three types of collocations based on the fact that they come in
a large variety of forms and the words that are involved can vary very much :
1. Predicative relations : consist of two words repeatedly used together in a similar
syntactic relation , and these lexic al combinations are the most flexible type of
collocation , yet they are hard to be distinguished as they frequently coincide to word
sequences in the corpus that are interrupted;
2. Rigid nouns phrases incorporate uninterrupted sequences of words such as " stock
market," "foreign exchange," "New York Stock Exchange," "The Dow Jones average
of 30 industrials” (Smadja, 1993: 148). This type of collocations can consist of
nouns and adjectives, along with closed class words, and are the most rigid type of
collocations. Rigid noun phrases cannot be fragmented without losing their meaning ,
also they refer to important concepts in a domain and se veral of them can be used to
express the same concep t (Smadja, 1993: 148).
3. Phrasal templates: imply idiomatic phrases that contain one, several or no empty
slots; they are phrase -long collocations. This type of collocations mentions the parts
of speech o f the words that can fill the empty slots and are representative for a given
domain. In contrast to the other two types , phrasal templates are useful for language
generation (Smadja, 1993: 148).
According to Choueka (in Smadja) , a collocation is a ―seque nce of adjacent words that
frequently appear together‖ (Smadja, 1993: 149). Smadja claims that in theory, the word
combinations can be any length, but actually, they are formed of two to six words. Church
and Hanks (1989) define collocations as a pair of c orrelated words; that is, according to
Smadja ―a collocation is a pair of words that appear together more often than expected‖
(Smadja, 1993: 150).
According to Benson, Benson, and Ilson (1986a), collocations can be split in two main
categories : lexical collocations and grammatical collocations. These two categories are
different as a consequence of the words that are involved. Lexical collocations embody
syntagmatic affinities among open class words such as verbs, nouns, adjectives, and adverbs.
On the o ther hand , grammatical collocations generally comprise at least one closed class
word among particles, prepositions, and auxiliary verbs (Smadja, 1993: 171).

32
According to Benson, grammatical collocations resemble lexical collocations , meaning that
both ―correspond to arbitrary and recurrent word -occurrences‖ (Benson in Smadja, 1993:
172).
Yusra Al Sughair claims that collocations ―reflect the linguistic, stylistic and cultural
features of texts‖ ( Al Sughair, 2011: iii) .
McKeown and Radev claim that coll ocations are ―a lexical phenomenon that has linguistic
and lexicographic status as well as utility for statistical natural language paradigms‖ , which
means that they include commonly used word pairs and phrases, for which the general
syntactic or semantic rules do not apply (McKeown & Radev , 1997: 2). According to them
collocations have been the object of studies in linguistics and lexicography , in the attempt to
define and include them into dictionaries. Acquiring fluency in using collocations i s difficult
due to their ubiquitous use. Even so, once fluency is acquired, collocations can have
different usages , like disambiguation, ― […] both word sense and structural disambiguation .
This task is based on the principle that a word in a particular se nse tends to co-occur with a
different set of words than when it is used in another sense‖ (McKeown & Radev , 1997: 2).
Another usage for collocations is translation , due to the ―ability‖ of them to not be
characterized ―on the basis of syntactic and semantic regularit ies‖, thus not being possible to
translate ―on a word by word basis‖ (McKeown & Radev , 1997: 2). They also say that
collocations cannot be easily defined and that they are usually discussed by comparison with
free word combinations and idiomatic expressio ns, collocations being somewhere in the
middle. A free word combination can be characterized using general rules; taking into
consideration the semantic constraints on words that have a specific syntactic relation with
the headword. On the contrary, an idiom is a fixed word combination, to which general rules
do not apply, and its meaning cannot be figured out from the words compiling it, and its
structure is fixed (McKeown & Radev , 1997: 2). Collocations are in between these two
extreme s and it is difficul t to differentiate categories. McKeown & Radev claim that ―a word
combination fails to be classified as free word and is termed a collocation when the number
of words which can occur in a syntactic relation with a given headword decreases to the
point wher e it is not possible to describe the set using semantic regularities ‖ (McKeown &
Radev, 1997: 2).

33
Because collocations fall in between free -word combinations and idioms , there have been
problems with the place and the manner of their illustration as part o f dictionaries. Their
placement is determined by the word that functions as base ; the base carrying the meaning
and generat ing use for the collocation. Collocations are commonly described as inconsistent,
specific to language and dialect, frequent in conte xt. The inconsistence denotes the fact that
replacing one word of the collocation w ith a synonym may result in an ―infelicitous lexical
combination‖ (McKeown & Radev, 1997: 5). They offer as an example the following: ―a
phrase such as make an effort is acc eptable, but make an exertion is not; similarly, a running
commentary , commit treason , warm greetings are all true collocations, but a running
discussion , commit treachery , and hot greetings are not acceptable lexical combinations ‖
(McKeown & Radev, 1997: 5).
As Baker states, ―every word in a language can be said to have a range of items with which
it is compatible, to a greater or lesser degree‖ (Baker , 1992 : 49) . The range refers to the
combination of other words that are normally combined with the word under consideration.
She claims that some words have a larger collocational range than others (Baker , 1992 : 49).
According to her and others before her, two key factors can have an impact on the
collocation range of an item. The first is the level of speci ficity : ―the more general a word is,
the broader its collocational range; the more specific it is, the more restricted its
collocational range‖ (Baker, 1992 : 50). For example, the verb bury should have a much
larger collocational range rather than any of i ts hyponyms, like inter or entomb ; that is
because only people can be interred , but one can bury people , a treasure , one‘s head , face,
feelings and memories. The second factor that has an impact on the collocational range is the
polysemy of the word. ―Most words have several senses and they tend to attract a different
set of collocates for each sense‖ (Baker, 1992 : 50). For instance, the verb run when it has the
meaning to ‗manage‘ collocates with words like company, institution, and business. When it
means ‗operate or provide‘ it collocates with words like service and course . Still, as stated
by Baker, it is appropriate to say that the other way around it is also true, i.e. ―the
collocat ional patterning of a word determines its different senses‖ (Baker, 1992: 50).
Collocations can be used to specify constraints to ambiguous words in order to disambiguate
their sense. Two major types of constraints have been investigated: one uses the general idea

34
according to which the sense of a n ambiguous word can be deri ved depending on the context
in which it occurs ; and the second type can be obtained when ―pairs of translations of the
word in an aligned bilingual corpus are considered‖ (McKeown & Radev, 1997: 12).
Vaclav Brezina, Tony McEnery and Stephen Wattam claim t hat there have been proposed
three criteria to identify collocations : (1) distance, (2) frequency, (3) exclusivity . The
distance defines ―the span around a node word where we look for collocates‖ (Brezina,
McEnery & Wattam, 2015: 140) and the span can be o f one word, if one were interested in
the adjectives that preceed a noun in English, or of four or five words, if one‘s were
interested in more general associations . This span is named ‗collocation window‘ . The
frequency of use is an important indispensabl e pointer of the normal pattern of a word
association. The authors assert that Gries (2013) claimed that there are three more criteria
that should be considered: (4) directionality, ( 5) dispersion and (6) type -token distribution
among collocates (Brezina, McEnery & Wattam, 2015: 140).
Directionality deals with the fact that ―the strength of the attraction between two words is
rarely symmetrical‖ (Brezina, McEnery & Wattam, 2015: 140) . They say t hat the word
affair has a stronger relationship with love than the other way around because love appears
more often with other words than affair , than vice versa. Dispersion is the dispersal of ―the
node and the collocates in the corpus‖ (Brezina, McEnery & Wattam, 2015: 140). The type –
token distribution among collo cates in addition to the strength of a collocation it also takes
into consideration the ―level of competition for the spot(s) around the head word ‖ (Brezina,
McEnery & Wattam, 2015: 141). The authors thought that in addition to the already six
features, it should also be added a seventh one: the connectivity between individual
collocates. They say that the collocates do not appear in isolation but they are being a part of
―a complex network of semantic relationships which ultimately reveals their meaning an d
the semantic structure of a text or corpus‖ (Brezina, McEnery & Wattam, 2015: 141).
Hatim and Mason claim that collocation , as one of the devices of lexical cohesion , is not
entirely mechanical but also supplies the ―powerful evidence of intentionality and text -type
focus‖ (Hatim&Mason, 2014: 204). Halliday and Hasan describe the process involved in
collocation:

35
―In general, any two lexical items having similar patterns of collocation ‒ that is ,
tending to appear in similar contexts ̶ will generate a cohesive force if they occur in
adjacent sentences‖
(Halliday and Hasan 1976, in Hatim&Mason , 2014 : 204 )
The autho rs affirm that the main idea is ―similar contexts‘ , going on and claiming that what
might be ‗a natural collocation‘ for one person may be less so for another. According to this
statement, collocations can be indicators to a predetermined meaning that cann ot be
expressed by other ways.
Baker says that collocations are ―in fact a direct reflection of the material, social, or moral
environment in which they occur ‖ (Baker, 1992: 49) . According to her collocations reflect
the ―cultural setting in which they ar e embedded‖ (Baker, 1992: 49). She exemplifies her
statement through the frequent English collocation buy a house , and stating that this
collocation is very rare in German because the habit of house -buying is different in these
two cultures.
3.3. Translat ing collocations
Barnbrook claims that for a proper translation of collocations it is necessary t o understand
them in both languages . According to him, Baker analyses collocations of the verb deliver in
English and their equivalents in Arabic; in Arabic t he phrase deliver a baby has a complete
change of focus, the equivalent verb being used with the object woman to describe the
process of birth ( Baker , 1992 in Barnbrook et al, 2 013: 130 ). Bakes goes on and comments
this by saying that this fact suggests th at differences at the level of collocational patterning
are not only about using a different verb with a given noun but also they implicate ―different
way of portraying an event‖ (Baker, 1992 in Barnbrook et al, 2013: 130).
Al Sughair claim s that it is gen erally known that the duty of translation may involve some
sort of diminishment in the meaning due to several re asons (poor choice of words, words
with multiple meanings, the differences in the lexical and grammatical systems of the
languages) (Al Sughair, 2011: 9). Sarikas states that the loss in translation ―is reported to be
a continuum b etween overtranslation (increased detail ) and undertranslation (increased
generalization) in the translation task‖ (Sarikas in Al Sughair, 2011: 9); thus, the translator

36
should be careful in delivering the veracious equivalent in the target language of
collocations.
Frank Smadja, Kathleen R. McKeown, and Vasileios Hatzivassiloglou claim that
collocations are difficult to translate especially for non -native speakers, most ly because they
are ―opaque and cannot be translated on a word -by-word basis‖ (Smadj a, McKeown &
Hatzivassiloglou, 1996: 1 ). Same authors describe collocations as modern -days
hieroglyphics , referring to the approach Jean -Francois Champollion used in order to crea te
the first comprehensive hieroglyphics dictionary and they developed a program na med
Champollion which ―given a pair of parallel corpora in two different languages and a list of
collocations in one of them, automatically produces their translation s‖ (Smadj a, McKeown
& Hatzivassiloglou, 1996: 1). This program us es ―[…] a sentence -aligned parallel bilingual
corpus, translates collocations in the source language into collocations in the target
language‖ (Smajda, McKeown & Hatzivassiloglou, 1996: 2) . The authors proceed and
mention different authors who have approached the translation of collocations through
different techniques like ―sentence alignment (Gale and Church 1991b; Brown, Lai, and
Mercer 1991; Simard, Foster and Isabelle 1992; Gale and Churc h 1993; Chen 1993), word
alignment (Gale and Church 1991a; Brown et al. 1993; Dagan, Church, and Gale 1993;
Fung and McKeown 1994; Fung 1995b), alignment of groups of words (Smadja 1992;
Kupiec 1993; van der Eijk 1993), and statistical translation (Brown e t al. 1993) ‖ (Smadja ,
McKeown & Hatzivassiloglou, 1996: 3). Sentence alignment techniques are mostly used a s
pre-processing stage, before the actual translation. Nevertheless, translation can be
attempted by using only statistical techniques.
Because col locations cannot be translated on a word -by-word basis , a speaker should instead
be aware of the meaning of the entire phrase in the source language and to be acquainted
with the common phrase that is typically used in the target language. Albeit collocati ons are
unpredictable on syntactic and semantic rules, they can be apprehended through repeated
usage . Collocations in which the verb functions as a support verb, issue most of their
meaning from the object in context not from their own semantic features . These types of
collocations are flexible, the constraint is between the verb and its object, and any number of
words may appear between these two elements. Collocations include also rigid groups of

37
words that do not change irrespective of the context they occur in , such as compounds
(Smadja , McKeown & Hatzivassiloglou, 1996: 5).
In order to show the difficulties that collocations cause for translations, the authors use two
sentences:
―(1a) ‗ Mr. Speaker, our Government has demonstrated its support for thes e important
principles by taking steps to enforce the provisions of the Charter more vigorously ‘
(lb) ‗Monsieur le Président, notre gouvernement a prouv é son adhésion ces importants
principes en prenant des mesures pour appliquer plus systématiquement les préceptes de la
Charte ―(Smadja , McKeown & Hatzivassiloglou, 1996: 5).
They say that even though the sentences are quite simple, the automatic translation of 1a as
1b raises several problems. One problem is raised by the presence of collocations. For
examp le, the English collocation to demonstrate support is translated as prouver son
adhésion . The translation uses words that do not coincide to individual words in the source
sentence, and the translation ―prouver son adhesion ‖ conveys the same meaning as the
source phrase. There are similar problems caused by groups of words like ‗ to take steps to‘;
‗to enforce provisions ‘, and they are identified as collocations for various reasons. For
instance, ‗to take steps to ‘ is a collocation because ‘to take’ is used as a support verb for the
noun steps. In the French translation en prenant des mesures it is used the Frenc h term for
‘take’ but the object is the translation of a noun that it is not appear in the source , mesures.
These collocations are flexible and they exhibit variations in word order. The example given
prove that collocations are ― […] domain dependent, often forming part of a sublanguage .
For example, Mr. Speaker is the proper way to refer to the Speaker of the House in the
Canadian Parliament when spea king English. The French equivalent, Monsieur le President ,
is not the literal translation but instead uses the translation of the term President ‖ (Smadja ,
McKeown & Hatzivassiloglou, 1996: 5 -6). The authors claim that the acquisition of the
ability to translate collocations would be an advantage for sublanguage translation. (Smadja ,
McKeown & Hatzivassiloglou, 1996: 6)
The authors claim that a translator should know to make a difference when ―[…] a full
phrase or pair of words must be considered for transl ation and when a word -by-word

38
technique will suffice‖ . In order to do that a translator should consider two tasks: ―1.
Identify collocations, or phrases which cannot be translated on a word -by-word basis, in the
source language. 2. Provide adequate tra nslation for these collocations ‖ (Smadja , McKeown
& Hatzivassiloglou, 1996: 6) . However, it is not sufficient just to have general knowledge
about the two languages , it is necessary to know as well the expressions used in the
sublanguage , since it have been observed that ―[…] idiomatic phrases have different
translations in restricted sublanguage than in general usage . In order to produce a fluent
translation of a full sentence, it is necessary to know the specific translation for each of the
source collocation s.‖ (Smadja , McKeown & Hatzivassiloglou, 1996: 6).
The authors ‘ approach , during their ‗experiment‘ with Champollion, was based on the
supposition that no collocation is ambiguous in the source language and has a distinctive
translation in the target langu age. On this assumption, they thought they could ignore the
context of the collocations and their translations and base their decisions only on the ―[…]
patterns of co -occurrence of each collocation and its candidate translations across the entire
corpus‖ (Smadja , McKeown & Hatzivassiloglou, 1996: 7).
They say as a conclusion based on their study that the ability to provide translations for
collocations is important for three main reasons:
 Given that they are opaque constructions, they cannot be translated on a word -by-
word basis , and the translations must be given for the entire phrase
 They are domain dependent , meaning that every domain comprises an assortment of
phrases that have specific meanings and translations that apply only in the given
domain
 If it were to take a quick look at a bilingual dictionary , shows that ―correspondences
between collocations in two languages are largely unexplored‖ (Smadja , McKeown
& Hatzivassiloglou, 1996: 34).
Al Sughaid says tha t Bahumaid conducted a research on determini ng the procedures that
translator s use in ―rendering collocations when they are not familiar with their equivalents in
the target language‖ . He found that there are four procedures used: ― (1) using a synonym or
near-synonym of the collocation, (2) going fo r a literal translation, (3) avoiding translating

39
the collocation at all, and (4) giving the meaning or explanation of the collocation‖ (Al
Sughaid, 2011: 10 -11). Al Sughaid claims the common procedure is giving the mea ning of
the collocation, second comes the usage of a synonym , after comes the literal translation and
finally the abolition of the collocation.
In 2012, Jiang M.Z., Yen T.X., Huang C.C., Chen M.H. and Chang J.S claim that the
method proposed by Smadja et al in 1996 , to translate collocation may lead to translations of
collocations that are not collocations in the target language. They say that automatically that
a collocation should also be a collocation in the target language , such as a verb -noun
collocation should be translated by a structu re that is also a verb -noun collocation in the
target language (Jiang et al., 2012: 249).
Smadja (1993) says that e very language establishes its own collocational constraints , and as
a result, the collocations become even more difficult to non -native speak ers. According to
him, the translations for collocation should be provided to them , and ―should not be word –
based but rather expression -based‖ (Smadja, 1993: 174). He also considers that bilingual
dictionaries are unfitted in dealing with this type of prob lems; and they provide ―context –
sensitive translations to ambiguous words (e.g., " number " or " rock") or highly complex
words such as " make ," "have " etc‖ (Smadja, 1993: 174). He also claims that the coverage of
the dictionaries ―is limited to semantic varia nts, and lexical collocations are generally
omitted‖ (Smadja, 1993: 174).
Hatim and Mason say that the translation of collocations is causing major problems even to
experienced translators. A collocation may pass unnoticed and the source language will
interfere , and the result would be ―an unnatural collocation‖ which flaws the translated text.
They also claim that in case of translating into native language the problem of interference
of the source language can be surpassed by ―vigilance and by careful re vising‖
(Hatim&Mason, 2014: 204 ).
Baker claims that statements about collocations refer to what is typical or untypical , not
what is admissible or inadmissible, therefore in her opinion there is no impossible
collocation (Baker in Barnbrook et al, 2013: 1 30). Barnbrook based on Baker‘s statement
says that this flexibility enables native speakers to improve their language use by the use of

40
both unmarked and marked collocations. Yet, according to him, if the translators are not
acquainted with the difference between these collocations, they might bump into some
problems in their translation process (Barnbrook et al, 2013: 130 -1).
Baker also claims that the differences in the ―collocational patterning‖ of the source and
target language lead to drawbacks and c an present several pro blems in translation. She is
offering a couple of problems that are comm on and encountered in the translation of non –
literary texts.
The engrossing effect of source text patterning : It is not hard to suppose that the translator
cannot be disoriented by the differences in the surface patterning between the source and
target language , if a collocation , which renders the same or similar meaning as the one from
the source language , can be identified in the target language text. She claims that some
translators may get ―quite engrossed in the source text and may produce the oddest
collocations in the target language for no justifiable reason‖ (Baker, 1992 : 54). The
confusion between source and target patterns is a trap that can be bypassed a s soon as the
translator is aware of the potential impact that the collocational patterning from the source
text can have on him. According to Baker, one can detach oneself from the source text by
putting aside the draft translation for a while. Neverthele ss, it is advisable for the translator
to avoid transferring source -language patterns that may be untypical for the target language,
except for the case where there is a good reason (Baker, 1992 : 55).
Misinterpreting the meaning of a source -language coll ocation: A collocation can be
misinterpreted in the source text as a cause of the involvement of his/her mother tongue.
This interference appear when a source -language collocation seems to be familiar because
―it corresponds in form to a common collocation in the target language ‖ under the
supposition that the translator is working into his/her mother tongue from a foreign language
(Baker, 1992: 55).
The tension between accuracy and naturalness : While translating un-marked source –
language collocations into the target language, a translator might try to produce a collocation
that is normal in the target language while simultaneously maintaining the meaning that is
associated with the source collocation. Still, this ideal, cannot always be achieved,

41
sometimes a tension connected to the choice between ―what is typical and what is accurate‖
(Baker, 1992: 56) . An adequate collocation in the target language supposes a slightly change
in the meaning , that can be minimal or ―particularly significant in a given conte xt‖ (Baker,
1992: 56). In translation, a certain level of ―loss, addition or skewing of meaning‖ is
frequently inevitable . It is important of course, for the translation to be accurate, but it is also
important to take into consideration the fact tha t ―the use of common target -language
patterns which are familiar to the target reader plays an important role in keeping the
communication channels open‖ (Baker, 1992: 57).
Culture -specific collocations: The cultural setting in which they occur can be reflected by
some collocations. In the case in which the two settings, the source and target language, are
different, the source text might contain collocations that render something that can be
unusual to the target reader. Culture -specific collocations express concepts that are not
easily accessible to the target reader (Baker, 1992: 59 -60). It is possible that the translator to
choose for ―accuracy of meaning, or for what appears to them to be accuracy of meaning, at
the expense of all else‖ (Baker, 1992: 61) .
Marked collocations in the source text: Baker claims that sometimes there are used
uncommon combinations of words so that new images to be created. It is preferable that the
marked collocations that are translated, to be marked as well in the target language . This is
however , biased by the constraints of the target language and the purpose of the translation.
3.4. Collocations for non -native speakers
As reported by Gyllstad (2007 ), second language learnes need to a quire both vocabulary and
grammar , consideri ng them ―indispensable aspects of knowledge‖ (Gyllstad, 2007: 1). He
sustains his statement by quoting Wilkins (1972) ―Without grammar very little can be
conveyed, without vocabulary nothing can be conveyed‖ (Wilkins, in Gyllstad, 2007: 1).
Further on, he claims that the knowledge of lots of words represent s an advantage for a
learner, yet certain ―learner categories‖ need to acquire ―native -like command of an L2 ‖
(2007: 1). He attests that for advanced learners like ―university -level students, teacher
students, translators and other professionals ‖ the possession of a vast vocabulary simply is

42
not enough. This is seconded by Wray‘s statement ―To know a language you must know not
only its individual words, but also how they fit together‖ (Wray, 2005:143 ).
Many authors claim that collocations represent a big problem for learners. Farrokh (2012)
claims that ―the most problematic areas for foreign language learning is collocation‖ as it
often is seen as an inconsistent, overwhelming and annoying obstacle to fluen cy of a native
speaker . Thus, it has been recognized as one of the ways t o differentiate native speakers and
non-native speakers (Farrokh, 2012: 56).
Darvishi states that several researchers have considered the significance of collocations and
the necessi ty of teaching collocations in EFL courses (English as a Foreign Language). She
mentions that these researchers have pointed out some benefits of learning collocations like
―increasing learners' language competence, enhancing learners' communicative compet ence,
and being toward native -like fluency‖ (Darvishi, 2011: 52 , based on Brown, 1974;
Nattinger, 1980, 1988; Channell, 1981; Bahns & Eldaw, 1993; Howarth, 1998 ). Yet, there
were studies that showed that learners have more faults in their writing and speak ing
because of their scarceness of collocation competence . He gives an example of Chinese
learners, which instead of saying ―take medicine‖ they say, ―eat medicine‖ as in Chinese the
noun ―medicine‖ often collocates with the verb ―eat‖ instead of ―take‖ (D arvishi, 2011: 52)..
Darvishi claims that Brown (1974) was among the few pioneers that proposed t o include
collocations in English as Second Language classroom. Brown also emphasizes on the fact
that the learning of collocation improve learners both vocab ulary and oral comprehension
and reading speed. Brown claims that the study of collocations can offer to the students the
―freedom to learn from their own and others‘ mistakes‖ ( 1974: 1 ). Furthermore, Brown
brings up that ―learning collocations enables le arners gradually to realize language chunks
used by native speakers in speech and writing and to get the feel of using words in natural
combination with other words as well‖ (Darvishi, 2011: 54 ). Matsuno Kakuzo (2017) quotes
Singleton (2000) who stated tha t ―we need to know about collocational patterns in order to
function smoothly in lexical terms in either our mother tongue or any other language we
may know‖ (Singleton, 2000 in Matsuno, 2017: 62). Barnbrook claims that along the years
of development of En glish teaching it had appeared the recognition that collocation is as
unit, ―an important element of the language knowledge that learners need to acquire‖

43
(Barnbrook et al, 2013: 120). He continues and states that in order to develop app ropriate
materials for teaching collocations, it is necessary to understand the importance of
collocations in language acquisition (Barnbrook et al, 2013: 120 -121). This idea is seconded
by a quote from Lewis (2000) focused on ―methods which incorporate and teach
collocation ‖ (Barnbrook et al, 2013: 121):
The single most important task facing language learners is acquiring a sufficiently
large vocabulary. We now recognise that much of our ‗vocabulary‘ consists of
prefabricated chunks of different kinds. The single most impor tant kind of chunk is
collocation. Self -evidently, then, teaching collocation should be a top priority in
every language course. (Lewis, 2000:8 in Barnbrook et al, 2013: 121).
Hill claims that the lack of collocational competence leads students into gramma tical
mistakes, due to the creation of long utterances because of the lack of knowledge of
collocations appropriate in that context . Later, the teacher focuses on correcting those
mistakes, not realizing that the reason for them is not the ―faulty grammar but a lack of
collocation‖ (Hill, 2000 in Barnbrook et al, 2013: 122 ). Hill sums up this problem in the
simple phrase ―less grammar, more lexis‖ ( Hill, 2000 in Barnbrook et al, 2013: 122 ).
Several studies show that learners make more mistakes and react mo re slowly , than native
speakers or that the intuition and the fluency with collocations of advanced learners d o not
seem to be equal to that of native speakers (Siyanova & Schmitt: 2008 , Yamashita & Jiang:
2010 , in Matsuno, 2017: 62 ). Yet, Wolter and Gylls tad(2013) found that advanced learners
are more sensitive to ―frequency effects ‖ for collocations in the second language. Also, as
reported by Matsuno, Kjellmer (1991) noted that native speakers make use of large
prefabricated sections, whereas « [t]he lea rner, on the other hand, having automated few
collocations, continually has to create structure»‖ (Kjellmer, 1991, in Matsuno, 2017: 62).
Kjellmer says at some point in his work that as the native speaker makes a hesitation pause
after long stretches of wo rds, whereas a ―moderately fluent leaner‖ makes one after a few
words. The le arner not only does he have preset few collocations, but he also has to plan his
thought units (Kjellmer, 1991: 124). Kjellmer considers that lexical items should be taught
and le arned not on their own but in their context of use (Kjellmer, 1991: 12). Gyllstad claims
that collocations are ―inherently formulaic‖ ; according to him, Wray suggests that
collocations are for native speakers, formulaic sequences where as for learners , they are not.

44
In Wray‘s perspective collocations are seen by adult learners as ―separate items (words)
which may become paired‖, and also that when learners encounter sequences like major
catastrophe they tend to separate the meaning and that they do not accumulate information
on the two in a combination (Wray, 2002 in Gyllstad, 2007: 23). According to Wray, the
thing that causes trouble to the learners is the process of combination of the words, due to
the variety of options: big, large , major , huge etc.
3.5. Conclusion
Taking all of these aspects into consideration, this chapter has treated not only theoretical
aspects regarding the definition of collocations but also the approaches needed to be taken
into consideration when a translator has to translate co llocations , and what problems might
one translator encounter in the translation of these combinations. In addition, it dealt with
the opinion of different authors in what concerns the study and the learning of collocations,
and their importance in the flue ncy of a learner of a foreign language. As Lewis (2000) says
the language acquisition is based on ― prefabricated chunks‖ and these chunks are
collocations , thus in his opinion the teaching of collocation should be included in all courses
of EFL.

45
CHAPTE R FOUR : THE TRANSLATION OF COLLOCATIONS IN “TURTLES
ALL THE WAY DOWN” BY JOHN GREEN
4.1. Introduction
The focus of this chapter shall be the analysis of the way collocations are to be translated
from John Green‘s book ―Turtles all the way down‖ in Romania n. Firstly, some
informations about the author will be necessary in order to have an idea about the
perspective he approaches. Secondly, a short synopsis about the book shall be added in order
to understand the story he approached.
4.2. The author – John G reen
John Green is a n American author who has published seven books, treating stories with
adolescents and their adventure through adolescence and dealing with their problems and
other people‘s problems. He was born on 24 August 1977 , in Indianapolis, Ind iana but after
he was born, his parents moved to Orlando, Florida. He majored in English and religious
studies , and he had spoken about bullying and how this made his teen -age period miserable.
He originally wanted to become a priest, but after a period of five months working in a
children‘s hospital, period in which he has worked with children suffering from deathly
illnesses , made him want to become an author and later on to write ―The Fault in Our Stars‖ .
As an answer to the question, ― Where do you get y our ideas for your books? ‖ Green, on his
website says that his books do not have ―capital -i Ideas‖, but the ideas come from ―lower
case-I ideas‖. For instance , he says that for Looking for Alaska it all started from thinking
―whether there was meaning to s uffering, and how one can reconcile one‘s self to a world
where suffering is so unjustly distributed‖ . For Paper Towns, it started from the idea of how
people are fascinated with ―manic pixie dream girls‖ and their continuous ―misimagining of
each other‖ . In a video he posted on YouTube on his channel vlogbrothers he confessed of
suffering from OCD this being the s tarting idea for Turtles all the way down . He says that
for him the obsessive comes before compulsive as he experiences ―obsessive thought
spiral s‖, intrusive thoughts that he d oes not want to have but he can not control. He also
claims that he started writing this book from trying to find ―a form and an expression for this
non-sensorial experience of living inside of thought spirals‖ (―What OCD Is Like (for Me)‖,

46
July 2017). His books have been translated in over fiftyfive languages . In Romania they
have been translated and published at Editura Trei.
4.3. The book
Turtles all the way down appeared in October 2017 and it is the author‘s fifth solo novel.
The author claims that the starting point for this book was his own struggle with this mental
illness. He stated in a vlog posted on YouTube that it all started from him thinking of how it
would be like living as a sixteen years old girl, trying to be a good daughter, a good friend, a
good student and also living with the terrifying thought spirals that she cannot escape from .
This story‘s main character is Aza Holmes , a troubled sixteen years old girl, who suffers
from OCD and has anxiety . Aza‘s men tal illness manifests under a fear of the human
microbiome , she is constantly worrying about infection, in particular by C. diff ( Clostridium
difficile ) ―which can be fatal‖ (Green, 2017: 11 ). She still also mourns the loss of her father ,
dead for eight ye ars, with whom she had a strong relationship. The other important
characters of the book are Aza‘s mother, Daisy Ramirez, Mychal Turner and Davis Pickett
Jr. The plot of the book starts with the disappearance of Russell Pickett, a billionaire
magnate inves tigated for fraud. Drawn by the reward offered for any information that might
lead to Pickett‘s arrest, Daisy drags Aza into a search for the missing billionaire. Davis, who
is aware that the two girls are after the reward, tries to stop them from searchin g his father
by giving them the reward of $100000 with money hidden by his father around the house.
From this point on, Davis and Aza start a relationship , but in which Aza considers that she
cannot commit, as she believes that she cannot overcome her anxi ety, fact that prevents them
from having a real relationship . Daisy Ramirez is Aza‘s childhood best friend, and she
writes Star Wars related fanfictio n. Daisy always insists on Aza reading her stories, but her
best friend refuses to do so. After long insistence, Aza reads Daisy‘s stories, only to find out
that she is using them as a method to vent for her frustrations about Aza‘s behaviour. Their
relationship gets to a tough point, but they rekindle their friendship after the accident they
have with the car . At an underground art exhibition inside an unfin ished drainage tunnel
system of Pogue's Run (that Pickett's company was responsible for), Aza and Daisy go
exploring on their own, where they finally solve the mystery and realize that Pickett had run
to the very place they were. After noticing a bad smell emanating from the area, they suspect

47
that the billionaire had already died. Aza tells Davis of their discovery. He places an
anonymous tip to the police, who find the body. Given the loss of their parents and home
(their mother had died years prior), added with the fact that their father had left his entire
fortune to his pet tuatara, Davis and his younger brother Noah decide to relocate to
Colorado, where Noah would attend a boarding school. As Davis and Aza say their
goodbyes, she reflects on the open possibilities of her future .
The translation of the book is published at Editura Trei and it is called ―Un șir infinit de
țestoase‖ by Camelia Ghioc.
4.4. Methodology
The data regarding the students‘ knowledge regarding collocations was gathered through a
questionnaire attached in Appendix. The questionnaire was applied to first year stude nts studying
at Faculty of Letters, at Transilvania University of Brașov . The test was applied to a group of
thirty -six students, and it consisted of exercices containing multiple collocations in context and
outside context. The s tatistic data concerning t he answers for the questionnaire represented in the
following chart.

As some of the online dictionaries I check ed might say, a collocation represents two or more
words that often go together and sound correct as such. As part of the test, the students w ere
asked to provide a d efinition/explanation of what a collocation is in their opinion. Out of the
thirty -six students, only twenty -two students tried to give a definition. Six of them defined the
collocation as an expression, or as two or more words that have a meaning . Some other six
answers that were similar were that a collocation is a group of words that have a meaning Solved the entire
questionnaire ,
22
Did not solve
exercise I, 14
Solved less the
50% out of
exercise II, 5Solved less
than 50% of
exercise III ,
18Solved less than
50% of exercise
IV, 18Translated
correctly less
than 50% of ex
V, 13

48
together . After analyzing the completed tests, the definition that was close est to reality was two
or more words combined in order to form an expression, a phrase.

The second exercise that the students were asked to solve was to match words in order to make
up collocations . The collocations that they were meant to form were the following: powerful
engine , have breakfast , make an effort , watch TV , substantial meal (also accepted substantial
breakfast ), bitterly cold , pitch dark , strictly forbidden , make mistakes . Out of the 36 students
that have answered the test, twelve of them have managed to make up less than 50% of the
collocations.

As a third exercise the students wer e asked to translate ten collocations taking into consideration
the context in which they appear. The English collocations used in the questionnaire were the
following:
1) break one’s promises (Daisy Ramirez won‘t brea k her promises, but she will break your
heart) Gave a
definition,
22Defined
collocation as
an expression ,
6Defined
collocation as a
group of words,
6Gave a proper
definition, 1
No collocation
made up , 1 Two to four
collocations
made up, 11
Over five
collocations
made up,
24Nine out of
nine
collocations
made up, 2

49
2) carrying a vague but constant fear (I remember being eleven and carrying a vague but
constant fear.)
3) to give a speech (She had to give a speech in front of the whole city.)
4) to kill time (To kill time that mor ning, she went through the files from the computer.)
5) to take care of (Before coming to the concert, he had to take care of some things.)
6) to put to sleep ‒ for this, two context were given (1. She had to put to sleep her old dog. 2.
This song is putting me to sleep. )
7) to get back (After a while, they had to get back to homework.)
8) to spoil the fun (She spoiled the fun by telling me the end of the movie to soo n.)
9) to leave a bad impression (By doing the opposite of what John has told her, left a bad
impression on everyone.) .
The students were asked to translate the expressions taking into consideration the context. They
had to complete the task without consult ing a dictionary. Out of the thirty -six students:
1) The expression break one‟s promises was translated as a-și încălca promisiunile by twenty –
four persons (other variants given: a-și rupe promisiunile , a se ține de promisiune , a-și respecta
promisiunile );
2) The phrase carrying a vague but constant fear has caused troubles to most of the students as
twelve st udents did not even try to provide a translation for this phrase. Some other twelve
students have translated this phrase as aveam o frică vagă, dar constantă , with variants for the
term fear: sentiment de frică , teamă , temere . Few of the students translate d the verb carrying as
simțeam or even purtam . However, one of the students had a more poetic approach of the
translation: Îmi amintesc că la 11 ani eram cuprins de o stare constantă și vagă de frică. The
context in which this phrase appear s is taken from Turtles all the way down , and in the Romanian
version this context is translated as ―î mi aminteam că aveam unsprezece ani și purtam în suflet
o frică vagă, dar permanentă‖ (Green translated by Ghioc , 2017: 1).
3) the expression to give a speech was transl ated as a ține un discurs by twenty -six students, the
variants given were a vorbi în public , a susține un discurs , two of the students did not understand
the collocation, thus translating it as a vorbit repede , a avut o conversație ;

50
4) the phrase to kill time was translated by fourteen of them as a pierde timpul , alternative
translations being a face să treacă timpul / a trece timpul , a omorî timpul , a omorî plictiseala , a-
și ocupa timpul ;
5) the idiom to take care of was translated as a rezolva …, with t he variant a se ocupa de (ceva) ,
a avea grija de (ceva) , a pune în ordine ;
6) the expression to put to sleep proofed to be difficult, as out of thirty -six students, only twelve
persons knew the euphemistic meaning of the expression that of killing an anima l painlessly,
often with an injection; translating it as a eutanasia , most of the students translating the
expression as a pune la somn , a-și adormi , a-și culca câinele. Two of the students even
translated it in a more sensitive manner: a-i curma viața , a-și lua rămas bun . As for the first
meaning of the expression, everyone translated it as a adormi ;
7) the phrasal verb to get back was tra nslated as a se întoarce by twenty -one students,
alternative given a se reapuca , a reveni la ;
8) the expression to spoil the fun was translated by most of the student as a strica distracția ,
alternative translations a stricat farmecul filmului , a distruge distracția , a strica/distruge cheful ;
9) the translation for the phrase to leave a bad impression were a lăsa o impresie proastă/rea ,
a face o impresie proastă/rea , a crea o impresie proastă . One of the students have translated it as
a lăsa un gust amar .
By analysing the whole exercise, not only the before mentioned collocations caused troubles to
some of the students, but also some of the context s they appeared in. For instance, the phrase
break your heart from the first sentence that was to be transla ted actually represented a problem
for some of the students. The entire sentence in which this phrase appears is extracted from the
book, and it is translated by Camelia Ghioc as follows Daisy Ramirez nu-și calcă niciodată
cuvântul, dar îți va călca inima -n picioare . Most of the students translated break your heart as
îți va frânge inima , very few of them have used the verb a rupe . Yet, some other students had a
different word choice and not only had th ey used a different verb but also they chose to
translated differently the entire syntagme . The translations were as follows: a.[…] te va răni ;
b.[…] îți va sfâșia inima ; c.[…] te va face să suferi ; d.[…] îți va distruge inima ; e.[…] îți va

51
Over five
correct
collocations,
10
Half of the
collocations,
8Less than
half of
them, 18distruge sufl etul. Another sequence for which there was given variants by the students was by
doing the opposite of what John has told her . One constant in most of the translation s, has
been the use of the Romanian gerund făcând and the use of the verb a spune . Howeve r, there
have been some students who even though they maintained the gerund, they translated the rest of
the sequence differently. Some of them maintained the sequence used by other făcând opusul a
ceea ce i -a spus John , but they used the verb a zice or a cere; others ha d chosen a different
approach for the translation of the sequence. Two of the students felt the need to add the
adjective s total and exact in order to emphasize the impact of the sentence. Two of the students
chose to use the negative in ord er to express the meaning of the sentence. One other student
opted for the use of a noun phrase instead of the usual verb phrase the phrase being translated as
făcând opusul vorbelor lui John […].
The fourth exercise was one that necessitated knowledge about collocations out of the context.
They were given possible co llocational range of eight words/phrases: life, environment , age, go
through , attention , memorie s, have , value . Out of the thirty -six students that solve d the
exercise, only ten of them succeeded to cross for more than four cases, the correct odd-one. Yet,
most of the students only
succeeded to cross only one to
three normal collocations. At a
closer analysis of the results, none
of the students succeeded in
guess ing all the eight collocational
range correctly. Although s ome of
the students that managed to made
up more than half of the
collocations at exercise no II,
managed as well to get a high scor e on this exercise , others that had a good score on that
exercise, have only done correctly up to four collocational ranges. Yet, most of the students have
maintained their score, proving a lack of outside -of-context collocations knowledge.
As a final ex ercise of the test, the students were asked to find the Romanian equivalent to ten
collocations. Some of them have been already used in the third exercise, but others were

52
completely new for this questionnaire. The ten collocations that were to be translat ed were the
following: to kill time , middle finger , to go through , to look at someone dead in the eye , fried
food, backseat bench , hold one’s hand , bubble gum , to commit assault , hardcover books . The
collocations that seemed to cause a lot of trouble
to the students were to commit assault , to look at
someone dead in the eye, hardcover books as
only six of the students managed to translate all
three of them correctly in Romanian. The main
translations given to these three were a te uita fix
în ochii cuiva , a privi pe cineva drept /fix în ochi ,
a fixa cu privirea ; a agresa , a comite un asalt;
cărți cu coperte cartonate , cu coperte tari .
However, even the concept s are quite simple, and
common, bubble gum and backseat bench also
represente d a problem to few of t he students, as
some either did not translate them or came up with some translation like gumă balonată/ baloane
de gumă for bubble gum and bucata din spate , centură de siguranță , spătarul băncii for
backseat bench .
After analysing the tests and student s‘ answers , most of the students even though they did not
provide a definition for collocations, succeeded to solve correctly many of the exercises, some of
them better than some of the students that provided a definition. By analysing the results of the
questionnaire, it may be said that the students, lack theoretical knowledge of collocations, being
able to translate and derive them thanks to the context and general level of English.
4.5. Corpus analysis
Another phase of the present research has been that of gathering a corpus made up of eighty –
seven elements, after reading the book in English as well as in Romanian. The corpus is inserted
in the Appendix along with the translation extracted from the Romanian version of the book. As
another step of this research , a categorization o f the collocation based on the theoretical
information mentioned earlier in the paper, on section 3.2, shall be made . Less than
five correct
translations
, 6
Five
correct
translations
, 6Over six
correct
translations
, 20Ten correct
translations
, 4

53
Throughout the history of the study of collocations, two theorists have made a classification of
collocations. One of them is Gyllstad based on the mathematical definition of Mel′čuk who says
that ―B expresses ‗C‘ only contingent on A‖ , and the other one is Smadja who came up with
these subclasses of collocations based on the fact that they come in a large variety of forms and
the words that are involved can vary very much .
Below, a classification of the collocations from the corpus, based on these two definitions shall
be made . Based on the subtypes Gyllstad came up with , we may classify the collocations from
the corpus as follows:
a) Collocations containing a delexical (or _support„, _light„) verb : ‗break her promise ‘,
‗fall in love ‘, ‗get a tattoo ‘, ‗to look up from her phone ‘, ‗to get back ‘, ‗put to sleep ‘, ‗to
look over ‘, ‗blow a kiss ‘, ‗to get laid ‘, ‗to look at someone dead in the eye ‘, ‗to commit
assault ‘, ‗hold my hand‘ , ‗to face criminal charges‘ ;
b) Collocations containing a dependent lexeme meaning which only occurs with one or
a few lexemes : ‗press conference ‘, ‗bank accounts ‘, ‗police report ‘, ‗ghost town ‘, ‗blog
entry ‘, ‗sexual harassment ‘, ‗criminal investigation ‘, ‗blood cells ‘, ‗tomato sauce ‘, ‗lawn
mower ‘, ‗rounding error ‘, ‗biological evidence ‘, ‗meteor shower ‘, ‗seat belt ‘, ‗passenger
seat‘, ‗backseat bench‘, ‗burner phone‘, ‗college fund‘, ‗steering wh eel‘, ‗ bubble gum ‘,
‗light pollution‘ , ‗executive assistant‘ , ‗law firm‘, ‗criminal investigation‘, ‗ middle
finger ‘, ‗‘
c) Collocations containing a dependent lexeme meaning (intensifiers) that can be used
together with other lexemes in the same sense, but it s meaning cannot be expressed
by a possible synonym : ‗true love‘, ‗fried food‘, ‗best friend‘, ‗legal guardian‘, ‗golf
cart‘, ‗frozen rain‘, ‗dirty money‘, ‗hardcover books‘, ‗mental health‘ ;
d) Collocations in which one lexeme is dependent on the other lexe me because the
meaning of the latter is utterly specific : ‗leafless branches‘, ‗parking spot‘, ‗parking
lot‘, ‗far away‘, ‗missing person report‘, ‗veggie burger‘, ‗summer vacation‘, ‗fierce
embrace‘, ‗ soul song ‘, ‗test tube‘, ‗nervous habit‘ .
According t o Smad ja‘s theory the collocations ‗leafless branches ‘, ‗night -vision picture ‘, ‗dining
room table ‘, ‗awkward silence ‘, ‗arc of light ‘, ‗nervous habit ‘, ‗silly game ‘, ‗home computer ‘ fall
into the category of predicative relations . Most of these collocatio ns have been translated

54
through literal translation , but for example, ‗fierce embrace ‘ has been translated through
transposition îmbrățișându -se cu foc , thus changing the grammatical clause of the collocation
without altering the meaning. One might h ave th e same mental representation while reading both
the English context and the Romanian one1. Some other collocations like for example ‗true love ‘,
‗parking spot‘, ‗parking lot‘, ‗best friend‘, ‗hospital gown‘, ‗seat belt‘, ‗lawn mower‘ , ‗rounding
error‘ fall into the category of rigid nouns phrases . These have been mostly translated through
literal translation . Collocations like ‗break her promise‘, ‗fall in love‘, ‗get a tattoo‘, ‗to kill
time‘, ‗put to sleep‘, ‗to take care of‘, ‗blow a kiss‘, ‗to look at s omeone dead in the eye‘ , ‗to lose
one‘s mind‘ fall under the category of phrasal templates , and have been translated through
calque and literal translation.
While reading the books and comparing the contexts in w hich some of these collocations occur, I
observed that during the translation, the translator has chosen to modulate the collocation , to
omit it altogether , or had to enlarge the context of the collocation in Romanian . For instance , for
‗state officials‘ the translation provided by the Romanian tra nslator is oficiali de la nivelul
autorității statului , which in my opinion could have been translated, considering the context of
occurrence2, as funcționari de stat .
Another curiosity found during the analysis and the study of the original and of the tr anslation
was in the context of ‗criminal charges‘. The context in which it occurs is ―Pickett might be
allowed to stay in them without having to face criminal charges at home‖ and it was translated as
lui Pickett i s-ar fi permis să locuiască în ele fără să fie trimis înapoi ca să fie judecat . Even
though I consider the translation quite accurate, and fit to the bigger context, I would have
changed it a bit as follows Pickett ar fi putut locui acolo fără să fie trimis acasă pentru a fi
judecat . In addition , another ‗ curiosity‘ concern s the translation of ‗school day‘. The context of
occurrence is ‗I believe your school day ends at two fifty -five, so you should have adequate time
to get there‘ and it‘s translation is Din câte știu terminați orele la două cincizeci și cinci, deci ar

1‗a lavender T -shirt […] featuring Han Solo an d Chewbacca in a fierce embrace‘ (en.) translated
as un tricou de culoarea lav andei […] pe care erau Chewbacca și Han Solo îmbrățișându -se cu
foc (p. 99)
2 ―the company had bribed a bunch of state officials in exchange for contracts to build a better
sewer overflow system in Indianapolis ‖ (p. 45) translated as ―[…] compania mituise mai mul ți
oficiali de la nivelul autorității statului ca să primească contractele pentru a construi un sistem
mai bun de deversare a excesului de deșeuri din canalizarea orașului Indianapolis‖

55
trebui să aveți destul timp ca să ajungeți acolo . As it can be seen in the context, the concept of
school day is a cultural bound term, and it had to be translated through an equivalent available in
Roman ian culture.
One other particularity about the translation of collocations can be seen in the case of ‗to reach
over‘ which can usually be translated as a (se) întinde , but in the context ―the movie was called
Jupiter Ascending , and it was both ridiculous and kind of awesome. A few minutes in, I reached
over to hold his hand, and it felt okay‖ it is omitted3. Not only that the concept of reaching over
is not expressed, but also the translator felt the need to mention with who Aza is in the scene. In
my opin ion, considering that the whole scene is centred on Aza and Davis, it is unnecessary to
mention his name.
The translator in one particular case has chosen a more poetic and more appropriate translation.
For example in the context ‗ He wrote of love and lon ging/ That often got him laid I‘m sure,
/Which was the poem‘s sole intent‘ , the translator has chosen to translate the phrase ‗got him
laid‘ as i-a adus la pat multe femei , which is an appropriate language for young readers , maybe
underage ones.
In the ca se of the ‗soul song‘ the translator has chosen to maintain the term as such, using a loan
borrowing technique, cântec soul as the concept of soul music is not present in the Romanian
culture , therefore there is no appropriate word to translate it. The context of ‗awkward silence‘
was translated Hai să continuăm în persoană tăcerea asta stingheră , being an interesting
translation for this collocation. Yet, the translation , in my opinion could also have been tăcere
stânjenitoare or tăcere ciudată . In the cas e of ‗nervous habit‘, it was translated through
transposition as obicei căpătat pe fond nervos .
4.6. Conclusion
Taking all these aspects into account , we have seen how complex and difficult to translate can
sometimes collocations be. We have also seen tha t some of the students lack knowledge about
collocations, not being able to work out the ir meaning or the proper translation. Collocations, as
it can be deduced from the theoretical background elaborated in the previous chapters, are a

3 Filmul se chema Ascensiunea lui Jupiter și chiar era în acela și timp ridicol și oarecum mi șto.
După câteva minute de la începutul filmului, l -am luat de mână pe Davis și a fost OK . (p. 181)

56
major part of any la nguage, and each word compounding the collocation has an impact on the
meaning and ―reflect the linguistic, stylistic and cultural features of texts‖ (Al Sughair, 2011: iii).
We have classified the collocations from the corpus based on two theories, of Sma dja and
Gyllstad, and in addition, we tried to determine the translation strategy used by Camelia Ghioc in
the process of translating Turtles all the way down . Collocations, as seen from the theory, are a
cultural aspect that is difficult to translate and define.
The questionnaire, the answers given to it and the analysis of the corpus has allowed us to come
to the conclusion that students need more knowledge about collocations, not only of their use but
also about their definition. Moreover, we have seen that collocations can be classified in two
categories, and can be translated through techniques as literal translation and transposition, rarely
through calque , sometimes even being omitted, as a result of the nonnecessity of their use in the
target langua ge. The translation of the collocations from Turtles all the way down has been a n
interesting job, as Camelia Ghioc has taken the liberty to translate them in such manner, in order
to preserve the soul and the mind of John Green.

57
CHAPTER FIVE: FINAL CON CLUSIONS
As mentioned earlier, this thesis aim ed at determining how collocations can be translated and
how much knowledge do students have about collocations . In relation to these aspects, a study
has been undertaken . The analysis was carried out on a corp us consisting of collocations
excerpted from Turtles all the way down and on a questionnaire applied to a group of ESL
learners , students at the Faculty of Letters.
Thanks to our study, we were able to answer the initial research questions. The answer to the
first research question (how can collocations be translated?) , as seen from the theory and the
analysis of the corpus and of the results, that they can be translated through transposition and
literal translation , or as C. Ghioc did, they can be omitted altogether. When it comes to
answering the second research question, concerning the knowledge students have about
collocations, we can say that the answer is a disappointing one , as few students, from those that
we have questioned, have succeeded to solve and answer all the questions from the questionnaire
applied. The research shows on one hand that few students have knowledge about collocations
and their translation an d on the other hand that collocations can be translated through
transposition or litera l translation.
5.1. Recommendations and suggestions for future research
Translation requires attention and diligence from the translator to avoid the delivery of an
awkward collocation and potentially an incorrect meaning to the target audience. It is
recommended that translators sh ould avoid calque translation as much as possible for the
translation of collocations in literary texts and opt for other strategies that help in producing
dynamic equivalence to produce texts with literary style in the TL. Collocations are a
phenomenon that can be exploited and investigated in many other genres. Other domains like
business, sports reporting, legal language, and science and technology can be further
investigated.

58
REFERENCE LIST
1. Al Sughair, Yusra. The transl ation of lexical collocations in literary texts . Phd Diss.
2011.
2. Baker, Mona. " In Other Words: A textbook on translation. " London and New York:
Routledge, 1992
3. Baker, M ona. (Ed.) . ―Routledge Encyclopedia of Translation Studies ‖ London:
Routledge, 1998
4. Barn brook, Geoff, Oliver Mason, and Ramesh Krishnamurthy. Collocation:
Applications and implications . Springer, 2013.
5. Bassnett, Susan. Translation studies . Routledge, 2013.
6. Bell, Roger . Translation and translating . Vol. 56. London: Longman, 1991.
7. Brezina, Vacl av, Tony McEnery, and Stephen Wattam. "Collocations in context: A
new perspective on collocation networks." International Journal of Corpus
Linguistics 20.2 (2015): 139 -173.
8. Brown, Dorothy F. "Advanced vocabulary teaching: The problem of collocation."
RELC journal 5.2 (1974): 1 -11.
9. Darvishi, Sohrab . The investigation of collocational errors in university students‘
writing majoring in English. In International Conference on Education, Research
and Innovation (Vol. 18, No. 12, pp. 52 -57), 2011.
10. Gyllstad, H enrik. Testing English Collocations : Developing Receptive Tests for Use
with Advanced Swedish Learners. Språk – och litteraturcentrum, Lunds universitet,
2007
11. Green, John. Turtles all the way down . Penguin UK, 2017 translated by Ghioc, C at
Editura Trei, 2017
12. Hatim, Basil, and Ian Mason. Discourse and the Translator. Routledge, 2014.
13. Jakobson, Roman. "On linguistic aspects of translation." On translation 3: 30 -39,
1959
14. Jiang, Ming -Zhuan Tzu -Xi Yen, Chung -Chi Huang, Mei -Hua Chen, and Jason S.
Chang. "Translatin g Collocation using Monolingual and Parallel Corpus." In
Proceedings of the 24th Conference on Computational Linguistics and Speech
Processing (ROCLING 2012), pp. 246 -260. 2012.

59
15. Kjellmer, G. O. R. A. N. ―A mint of phrases ‖ IN English Corpus Linguistics: St udies
in Honour of Jan Svartvik ed. K. Aijmer, et al., 111 -127. (1991).
16. Krings, Hans P. "Translation problems and translation strategies of advanced
German learners of French (L2)." Interlingual and intercultural communication
(1986): 263 -276.
17. Lederer, Mar ianne. Translation: The interpretive model. Routledge, 2014.
18. Löscher, Wolfgang. "Translation performance, translation process and translation
strategies." Tuebingen: Guten Narr (1991): 8.
19. Matsuno, Kazuko. "Processing collocations: Do native speakers and sec ond language
learners simultaneously access prefabricated patterns and each single word?."
Journal of the European Second Language Association 1.1 (2017).
20. McKeown, Kathleen R., and Dragomir R. Radev. "Collocations." Handbook of
Natural Language Processing . Marcel Dekker: 1 -2, 2000
21. Newmark, Peter. Approaches to translation (Language Teaching Methodology
Senes) . Oxford: Pergamon Press, 1981.
22. Newmark, Peter. A textbook of translation . Vol. 66. New York: Prentice hall, 1988.
23. Newmark, Peter. (1988b). Approaches to Translation. Hertfordshire: Prentice Hall
24. Nida, Eugene A. Toward a science of translating: With special reference to principles
and procedures involved in Bible. EJ Brill, 1964.
25. Nida, Eugene A., and Charles Russell Taber. The theory and practice of tran slation.
Vol. 8. Brill, 2003.
26. Ordudari, Mahmoud. "Translation procedures, strategies and methods." Translation
Journal 11.3 (2007): 8.
27. Proshina, Zoya. "Theory of translation (English and Russian)." Z. Proshina. –
Vladivostok: Eastern University Press, 2008
28. Smadja, Frank. "Retrieving collocations from text: Xtract." Computational
linguistics 19.1 (1993): 143 -177.
29. Smadja, Frank, Kathleen R. McKeown, and Vasileios Hatzivassiloglou. "Translating
collocations for bilingual lexicons: A statistical approach." Comput ational linguistics
22.1: 1 -38, 1996.

60
30. Toury, Gideon. "The nature and role of norms in translation." The translation studies
reader 2, 2000.
31. Vinay, Jean -Paul, and Jean Darbelnet. "Stylistique comparée de l‘anglais et du
français." Paris/Montréal: Didier/Bea uchemin, 1958.
32. Venuti, Lawrence. The scandals of translation: Towards an ethics of difference.
Routledge, 2002.
33. Venuti, Lawrence. Strategies of translation. In M. Baker (Ed.), Encyclopedia of
translation studies (pp. 240 -244). London and New York: Routledg e, 1998
34. Wray, Alison. Formulaic language and the lexicon. Cambridge University Press,
2005.
―What OCD is Like (for Me) ‖ at https://youtu.be/jNEUz9v5RYo : 7 May 2019
Biographic data about John Green at http://www.johngreenbooks.com/ : 7 May 2019

61
APPENDIX I: QUESTIONNAIRE

I. What is in your opinion a collocation? Provide a definition/explanation.

II. Match the words in column A with the words in column B in order to mak e up
collocations.
A B
1. substantial a. engine
2. make b. TV
3. pitch c. mistakes
4. watch d. meal
5. strictly e. breakfast
6. make f. cold
7. bitterly g. dark
8. have h. an effort
9. powerful i. forbidden
III. Translate the following collo cations taking into consideration the context:
1. ―Daisy Ramirez won‟t break her promises , but she will break your heart.‖

2. ―I remember being eleven and carrying a vague but constant fear .

3. She had to give a speech in front of the whole city.

4. To kill time that morning, she went through the files from the computer.

5. Before coming to the concert, he had to take care of some things.

6. She had to put to sleep her old dog.

7. This song is putting me to sleep .

8. After a while, they had to get back to homework.

9. She spoiled the fun by telling me the end of the movie to soon.

10. By doing the opposite of what John has told her, left a bad impression on everyone.

62
IV. Cross out the word in each set, which does not form a normal collocation.
1. live /lead/go/ have a comfortable life
2. a firm/ familiar/ pleasant/ stable environment
3. feel /look/ talk/ show your age
4. go through a crisis/stage/divorce/problem
5. undivided/perfect/careful/close attention
6. share/wake up/ rekindle/ stir up memories
7. have adventures/ a baby visitors/ a n increase
8. sentimental/ sensitive/ practical/ outstanding value

V. Find the Romanian equivalent to the following collocations:
1. ―to kill time‖=
2. middle finger =
3. to go through =
4. to look at someone dead in the eye =
5. fried food =
6. backseat bench =
7. hold one‘s hand =
8. bubble gum =
9. to commit assault=
10. hardcover books =

63
APPENDIX II : CORPUS
No. English
Collocation Translation
1. break her
promise a-și călca cuvântul
2. vague but
constant fear frică vagă , dar
permanentă
3. press confe rence conferință de presă
4. bank accounts conturi bancare
5. leafless
branches crengi desfrunzite
6. true love iubire adevărată
7. police report raportul
8. fall in love a te îndrăgosti
9. ghost town oraș fantomă
10. light pollution poluare luminoasă
11. blog entr y postare
12. detective skills talent de detectiv
13. executive
assistant asistentă executiv
14. sexual
harassment hărțuire sexuală
15. law firm firmă de avocatură
16. criminal
investigation anchetă penală
17. state officials oficiali de la
nivelul autorității
statulu i
18. home computer calculator de acasă
19. silly game joc nevinovat 20. get a tattoo a se tatua
21. to get back a se întoarce
22. to look up from
her phone a-și ridica ochii din
telefon
23. far away foarte îndepărtată
24. middle finger deget mijlociu
25. blood cells celulele sangvine
26. school board consiliu școlar
27. missing person
report raportul dispariției
28. video game joc video
29. upside down răsturnat
30. tomato sauce sos de roșii
31. parking spot locul de parcare
32. parking lot parcare
33. best friend cea mai bună
prietenă
34. arc of light arc de lumină
35. mental health sănătatea mintală
36. summer
vacation vacanța de vară
37. case study studiu de caz
38. fierce embrace îmbrățișându -se cu
foc
39. fried food prăjeală
40. hardcover books cărți cu coperte
cartonate
41. hold my hand să mă ia de mână
42. to kill time ca să mai treacă

64
timpul
43. to face criminal
charges a fi judecat
44. lawn mower mașina de tuns
iarba
45. soul song cântec soul
46. rounding error greșeală de
rotunjire
47. dirty money bani murdari
48. school day ore
49. biological
evidence dovadă b iologică
50. seat belt
centură de
siguranță
51. golf cart
mașinuță de golf
52. meteor shower ploaie de meteoriți
53. veggie burger burger vegetarian
54. passenger seat scaunul din dreapta
55. night -vision
picture poza de noapte
56. frozen rain ploaie înghețată
57. dinin g room
table masa din sufragerie
58. hospital gown cămașă de spital
59. movie screen ecran
60. backseat bench bancheta din spate
61. dried blood sânge uscat
62. saving accounts conturi de
economii 63. to take care of a avea grijă
64. awkward silence tăcere stingheră
65. put to sleep a adormi
66. go through a trece prin
67. burner phone telefon cu cartelă
68. bedside table noptieră
69. college fund fond de studii
70. hospital bed pat de spital
71. legal guardian tutore
72. to look over a se uita la
73. to reach over –
74. blow a kiss trimi te o sărutare
75. test tube eprubetă
76. to get laid
77. nervous habit obicei căpătat pe
fond nervos
78. lounge chair șezlong
79. bathing suit costum de baie
80. steering wheel volan
81. to look at
someone dead in
the eye a privi țintă
82. bubble gum gumă de mestecat
83. to lose one‘s
mind a-și pierde mințile
84. dry cleaners curățătorie
85. to commit
assault a ataca
86. beam of light fascicule de lumină