You are on page 1of 150

VIETNAM NATIONAL UNIVERSITY HO CHI MINH CITY

UNIVERSITY OF SOCIAL SCIENCES AND HUMANITIES


FACULTY OF ENGLISH LINGUISTICS & LITERATURE

AN ANALYSIS OF ERRORS RELATED TO


ADJECTIVENOUN COLLOCATIONS MADE BY 3rd
YEAR ENGLISH MAJORED STUDENTS AT USSH
AND IMPLICATIONS FOR TEACHING TRANSLATION

A thesis submitted to the

Faculty of English Linguistics & Literature

in partial fulfillment of the Masters degree in TESOL

By

NGUYN L B TNG

Supervised by

L HONG DNG, PhD

HO CHI MINH CITY, FEBRUARY 2017


STATEMENT OF AUTHORSHIP

I hereby certify my authorship of the thesis submitted today entitled:

AN ANALYSIS OF ERRORS RELATED TO ADJECTIVENOUN


COLLOCATIONS MADE BY 3rd YEAR ENGLISH MAJORED
STUDENTS AT USSH AND IMPLICATIONS
FOR TEACHING TRANSLATION

in terms of the statement of Requirements for the Thesis in Masters Program issued
by the Higher Degree Committee. The thesis has not been submitted for the award of
any degree or diploma in any other situation.

Ho Chi Minh City, _____ 2017

Nguyn L B Tng

i
RETENTION AND USE OF THE THESIS

I hereby state that I, Nguyn L B Tng, being the candidate for the degree of Master
in TESOL, accept the requirements of the University relating to the retention and use
of Masters Theses deposited in the Library.

In terms of these conditions, I agree that the original of my thesis deposited in the
Library should be accessible for the purpose of study and research in accordance with
the normal conditions established by the library for the care, loan or reproduction of
the thesis.

Ho Chi Minh City, _____ 2017

Nguyn L B Tng

ii
ACKNOWLEDGEMENTS

First and foremost, I would like to express my gratefulness to my teacher and


supervisor, Dr. L Hong Dng, for all of his invaluable guidance, encouragement and
dedication. Without his wholehearted support and valuable advice, the completion of
this thesis would have been impossible.

I am greatly indebted to all teachers who taught me invaluable knowledge, which was
essential for the fulfilment of this thesis. Specially, I would like to extend my gratitude
to Assoc. Prof. Dr. T Minh Thanh, who inspired me to choose the thesis topic and
helped me form the very first ideas to carry out the study.

My sincerest thanks also go to the Board of Administrators and other members of the
Faculty of English Linguistics and Literature, who directly or indirectly contributed to
the implementation of this thesis.

I am deeply thankful to my colleagues and friends, especially Ms. Vn Di, Mr. Thin
Lc, Mr. Khi Nguyn and Ms. Bo Ngc, whose constant support and valuable
experience have led me through times of difficulty.

Last but not least, my deepest gratitude goes to my family, my loving mother, my
dedicated father, my aunts and uncles, and my beloved one, whose unconditional love
and care have been the very reason for every achievement I have had. Without them,
this thesis would have never been completed.

iii
TABLE OF CONTENTS

STATEMENT OF AUTHORSHIP ....................................................................................................... i


RETENTION AND USE OF THE THESIS ........................................................................................ ii
ACKNOWLEDGEMENTS ................................................................................................................ iii
TABLE OF CONTENTS .................................................................................................................... iv
LIST OF ABBREVIATIONS ............................................................................................................ vii
LIST OF TABLES ............................................................................................................................ viii
LIST OF FIGURES ............................................................................................................................. ix
ABSTRACT ......................................................................................................................................... x
CHAPTER 1 INTRODUCTION .......................................................................................................... 1
1.1. Background to the study ........................................................................................................... 1
1.2. Rationale for the study .............................................................................................................. 3
1.3. Aim of the study........................................................................................................................ 4
1.4. Research questions .................................................................................................................... 4
1.5. Significance of the study ........................................................................................................... 4
1.6. Scope of the study ..................................................................................................................... 5
1.7. Overall structure of the study .................................................................................................... 6
CHAPTER 2 LITERATURE REVIEW ............................................................................................... 7
2.1. Collocation ................................................................................................................................ 7
2.1.1. Definition of collocation ................................................................................................. 7
2.1.2. Categorization of collocations ....................................................................................... 11
2.1.3. Components of collocations .......................................................................................... 13
2.2. Collocations and translation .................................................................................................... 14
2.2.1. The significance of collocations in translation .............................................................. 14
2.2.2. Difficulties in translating collocations .......................................................................... 15
2.2.3. Influences of L1 on learners use of collocations.......................................................... 17
2.2.3.1. English collocations vs. Vietnamese counterparts.............................................. 17
2.2.3.2. Congruent and incongruent collocations ............................................................ 19
2.2.4. Strategies in translating collocations ............................................................................. 20
2.3. Error analysis .......................................................................................................................... 21
2.3.1. Theory of error analysis ................................................................................................ 21
2.3.2. Types of errors related to adjective-noun collocations ................................................. 23
2.3.3. Causes of collocational errors ....................................................................................... 25
2.4. Collocation and language teaching ......................................................................................... 28
2.5. Methods for testing collocational knowledge ......................................................................... 33

iv
2.6. Conceptual framework ............................................................................................................ 35
CHAPTER 3 METHODOLOGY ....................................................................................................... 39
3.1. Research questions .................................................................................................................. 39
3.2. Research design ...................................................................................................................... 39
3.2.1. Context of the study ...................................................................................................... 39
3.2.2. Participants .................................................................................................................... 40
3.2.2.1. Student participants ............................................................................................ 41
3.2.2.2. Teacher participants............................................................................................ 42
3.2.3. COCA ............................................................................................................................ 42
3.2.4. Research instruments .................................................................................................... 44
3.2.4.1. Students midterm exam papers in Basic Translation ........................................ 44
3.2.4.2. Translation test ................................................................................................... 45
3.2.4.3. COLLEX test ...................................................................................................... 48
3.2.4.4. Interview ............................................................................................................. 49
3.2.4.5. Questionnaire ...................................................................................................... 50
3.2.5. Pilot study...................................................................................................................... 53
3.3. Data collection procedure ....................................................................................................... 54
3.3.1. Collection of adjective-noun collocational errors ......................................................... 54
3.3.1.1. Collection of errors from students midterm exam papers ................................. 55
3.3.1.2. Collection of errors from the translation test ...................................................... 57
3.3.2. Collection of additional data ......................................................................................... 58
3.3.2.1. Questionnaire ...................................................................................................... 58
3.3.2.2. COLLEX test ...................................................................................................... 58
3.3.2.3. Interviews to the teachers ................................................................................... 59
3.4. Data analysis procedure .......................................................................................................... 59
3.4.1. Analysis of adjective-noun collocational errors ............................................................ 59
3.4.2. Questionnaire ................................................................................................................ 60
3.4.3. COLLEX test ................................................................................................................ 60
3.4.4. Interview ....................................................................................................................... 61
3.5. Summary ................................................................................................................................. 61
CHAPTER 4 FINDINGS AND DISCUSSION ................................................................................. 62
4.1. Data analysis and discussion ................................................................................................... 62
4.1.1. Analysis of collocational errors from students exam papers ....................................... 62
4.1.1.1. Types of errors .................................................................................................... 64
4.1.1.2. Causes of errors .................................................................................................. 66
4.1.2. Analysis of collocational errors from the translation test .............................................. 73
4.1.2.1. Types of errors .................................................................................................... 75
v
4.1.2.1. Causes of errors .................................................................................................. 77
4.1.3. Summary of the error analyses ...................................................................................... 84
4.1.4. Learners receptive knowledge of collocations ............................................................. 86
4.1.5. The teaching and learning of collocations in Basic Translation.................................... 89
4.1.5.1. The teaching of collocations ............................................................................... 90
4.1.5.2. The learning of collocations ............................................................................... 93
4.2. Major findings ....................................................................................................................... 100
CHAPTER 5 CONCLUSION AND RECOMMENDATIONS ....................................................... 108
5.1. Conclusion ............................................................................................................................ 108
5.2. Suggestions and recommendations ....................................................................................... 110
5.2.1. Suggestions to teachers ............................................................................................... 110
5.2.2. Suggestions to students ............................................................................................... 116
5.2.3. Recommendations for future research......................................................................... 118
REFERENCES ................................................................................................................................. 119
APPENDIX 1 ................................................................................................................................... 126
APPENDIX 2 ................................................................................................................................... 129
APPENDIX 3 ................................................................................................................................... 130
APPENDIX 4 ................................................................................................................................... 131
APPENDIX 5 ................................................................................................................................... 134
APPENDIX 6 ................................................................................................................................... 135

vi
LIST OF ABBREVIATIONS

BNC the British National Corpus

COCA the Corpus of Contemporary American English

EF the Faculty of English Linguistics and Literature

EFL/ESL English as a Second/Foreign Language

FREQ Frequency

LA the Lexical Approach

L1 First Language

L2 Second Language

MI Mutual Information

Ss Students

SV Subject Verb

SVO Subject Verb Object

TESOL Teaching English to Speakers of Other Languages

Ts Teachers

USSH University of Social Sciences and Humanities Ho Chi Minh City

vii
LIST OF TABLES

Table 2.1: Seven types of lexical collocations (Farrokh, 2012) ................................ 12

Table 2.2: The revised six types of lexical collocations ............................................ 12

Table 2.3: Types of collocational errors (Nesselhauf, 2003) ..................................... 23

Table 2.4: Useful collocations to teach students (Hodne, 2009) ............................... 30

Table 3.1: Description of the questionnaire items ..................................................... 52

Table 3.2: Summary of the research instruments....................................................... 52

Table 3.3: Reliability statistics of the translation test ................................................ 53

Table 4.1: Lexical combinations from the midterm-exam papers ............................. 63

Table 4.2: Types of collocational errors from the midterm-exam papers ................. 65

Table 4.3: Causes of collocational errors from the midterm-exam papers ................ 66

Table 4.4: Summary of students responses to the translation test ............................ 74

Table 4.5: Types of collocational errors from the translation test ............................. 75

Table 4.6: Causes of collocational errors from the translation test ........................... 77

Table 4.7: Comparison between productive test and receptive test ........................... 87

Table 4.8: Students performance on each item of the two tests ............................... 88

Table 4.9: Students statement about the teaching of collocations in basic


translation ................................................................................................................... 91

Table 4.10: Strategies used by the teachers to teach collocations ............................. 92

Table 4.11: Sources of students knowledge about collocations ............................... 94

Table 4.12: Students opinions on the roles of collocations in translation ................ 94

Table 4.13: Students strategies to broaden knowledge about collocations .............. 95

Table 4.14: Students strategies to deal with difficulties in translating collocations 97

viii
LIST OF FIGURES

Figure 2.1: Conceptual framework of the study ........................................................ 38

Figure 3.1: The COCA search interface .................................................................... 55

Figure 3.2: The COCA result interface ...................................................................... 56

Figure 4.1: Summary of types of collocational errors ............................................... 85

Figure 4.2: Summary of causes of collocational errors ............................................. 85

Figure 4.3: Students responses to Q6, Q7 and Q8 .................................................... 96

Figure 4.4: Summary of the major findings ............................................................. 106

ix
ABSTRACT
Collocation is an important concept in language teaching. Learning and using
collocations are believed to boost learners communicative competence. However,
collocations also bring many problems to learners due to their arbitrariness as well as
linguistic and cultural differences. Although quite a few studies on collocations have
been conducted, there are still gaps for further research when certain types of
collocations and other influential factors such as translation are taken into account. In
order to fill some of these gaps, the present study was conducted, which focused on
adjective-noun collocational errors in translation. Multiple research methods and
tools, including (1) error analysis, (2) questionnaire and (3) interview, were employed
to thoroughly investigate the errors, their possible causes and to find out appropriate
solutions to the addressed issues.

The study revealed that the student participants, in spite of their high level of
proficiency as English-majored juniors, still made a considerable number of
collocational errors. The majority of the said errors were caused by incorrect choices
of adjectives, which could be explained by the fact that the choices of adjective, as the
collocate, in a collocation is restricted by the noun the node. Regarding the causes
of errors, approximation and L1 transfer were the two major ones. The study also
directed attention to some other causes, which had been usually neglected in previous
studies, such as formal confusion and inappropriate use of dictionaries. Upon further
investigation, it was pointed out that the learners lack of knowledge about collocation
was the major implying factor accounting for most of the observed errors. This was
probably resulted from the fact that the teaching of collocations was mostly integrated
in other tasks. Such implicit instruction, though useful, may not be effective enough
to attract their attention. Based on the findings of the study, recommendations on the
teaching and learning of collocations in translation were presented. Firstly, it is
necessary to raise learners awareness of the concept of collocations. Secondly, the
teaching of collocations in translation should be more explicit to attract their attention.
Finally, it is advised that students should put more effort in improving their knowledge
of collocations.

x
CHAPTER 1
INTRODUCTION

This chapter provides an introduction to the study, including (1) background to the
study, (2) rationale for the study, (3) aim and objectives of the study, (4) research
questions, (5) significance of the study, (6) scope of the study and (7) its overall
structure.

1.1. Background to the study

Among recently developed approaches to language teaching, the Lexical Approach


(LA) is one that lays emphasis on teaching prefabricated chunks instead of separate
words. LA was built from the belief that acquiring and using lexical units as chunks
can effectively improve learners communicative competence, an idea which has been
supported by a number of scholars. Pawley & Syder (1983), for example, have
remarked that a large part of human speech is multiword units which function as
chunks or memorized patterns. Harmer (2001), sharing the same view, wrote that
oral fluency requires a spontaneous process of language and information, which is
marked by the use of a number of common lexical phrases (p. 269). These ideas were,
again, affirmed by Jiang & Nekrasova (2007) who have recently found that formulaic
sequences are processed more quickly and more accurately than non-formulaic
sequences.

One important concept in LA is collocation, which can be understood as the regular


co-occurrence of words in a language. Since it shows how words in a language work
together, or that is to say how formulaic sequences (or chunks) are formed, the use of
the aforementioned approach is hardly possible without thorough understanding of the
concept. Collocation, therefore, is thought to play a central role in learning and in
communication (Richards & Rodgers, 2001, p. 133).

While collocation is important, it is also problematic for EFL/ESL learners due to


several factors. Firstly, while native speakers can produce these prefabricated chunks,
including collocations, quite effortlessly, most learners cannot acquire them without a
great amount of time and effort. As a matter of fact, they are not exposed to L2 as

1
much as native speakers do, and thus the chances for them to practice using these
chunks are quite limited. As a result, they usually resort to rote-learning, which seems
to be an ineffective method to memorize a huge amount of lexical items. Especially
with collocations, the memorizing process is even more challenging due to their
arbitrary combination rules. One cannot explain, for instance, why tea has to go with
strong as in strong tea but not with powerful a synonym of strong. Being confused
by such rules, learners may find it difficult to remember and use collocations correctly.

Another factor making the acquisition of collocations even more difficult is the
influence of learners L1. In many circumstances, learners usually resort to their
mother tongue and try to express their ideas through word-for-word translation from
their L1. Nevertheless, this strategy does not assure the efficiency and accuracy in
communication, since even if two words in a language are collocations of each other,
their equivalents in the other language do not necessarily have the same relationship.
Word-for-word translation may, therefore, lead to the violation of collocational rules
in L2, and thus make the speakers sound awkward, not to mention the negative
influence on their fluency caused by such a process.

Given the difficulties above, it seems inevitable for learners to make errors in using
collocations. Indeed, it is a proven fact that even students of high level of English still
make collocational errors due to large amount of complicated lexical items they have
to memorize and produce, not to mention the arbitrariness of collocations and
differences between languages. However, collocational errors, according to
researchers such as Bahns & Eldaw (1993), Li (2005), Darvishi (2011), can be reduced
by improving learners knowledge of collocations through multiple teaching methods
including identification and correction of collocational errors, introduction of certain
target collocations, etc. Although these methods have proved their effectiveness, it is
an undeniable fact that collocation is still a constant challenge to EFL learners, which
emphasizes the necessity to have more research into the issue.

Along with the increasing popularity of LA, collocation and teaching of collocations
have attracted much attention from scholars and researchers, such as Bahns & Eldaw
(1993), Fan (2009), Farrokh (2012) and Lubis (2013), etc. Most of these studies have
focused either on analyzing learners collocational errors to figure out the causes of

2
and remedies for the errors or on looking for new methods for effectively teaching
collocations to learners. There were also studies that had both foci, in which the latter
was usually done on the basis of the former.

1.2. Rationale for the study

Although a great number of studies into collocation have been conducted, most of
them focused on one type of collocation, namely verb-noun, since this type collocation
occurs very frequently and causes considerable difficulty for learners of English.
However, among other types of collocation, adjective-noun also has high frequency
of occurrence and high level of difficulty, which have been confirmed in several
studies. Trinh (2001) conducted a contrastive analysis between English and
Vietnamese collocations to find out particular lexico-semantic features which are
alien, unfamiliar, or unpredictable, from the point of view of a Vietnamese translator
of English (p. 59). The study revealed that, among the categories, adjective-noun
collocation was the broadest one with about 1700 instances, accounting for 56.6
percent of the collected data. Other researchers, including Channell (1981), Koya
(2005), Shehata (2008) and Kurosaki (2012), have come to the same conclusion that
adjective-noun collocation caused great difficulties for learners.

In the context of Vietnam, the attention towards collocations has been increased with
several studies being conducted in the recent years. Nguyen (2006) studied the
importance of collocational knowledge in language use. In 2008, Mai investigated
learners use of lexical collocation in Vietnamese English translation. More recently,
two studies have been completed by Dang (2014) and Nguyen (2014). The former was
a survey into the teaching of lexical collocations in academic writing, whereas the
latter investigated learners competence in producing and recognizing verb-noun
collocations. However, similar to many foreign studies in the field, none of the
aforementioned works specifically focused on adjective-noun collocation.

Moreover, while most of the studies above aimed to find solutions for problems related
to collocations in language teaching in general, few of them had implication
specifically for teaching translation although translation is a context where
collocation-related difficulties are likely to arise as it requires good command of

3
collocations in both source and target languages. Given the fact that translation by its
nature is complicated and translating collocation is among the biggest challenges to
even professional translators let alone learners of English, it is necessary to have more
studies that specifically deal with collocations in translation.

The aforementioned issues altogether became the research gap that drove the
researcher to conduct the present study that aimed at adjective-noun collocations in
translation and sought relevant implications for teaching translation at University of
Social Sciences and Humanities (USSH) where translation is a subject being taught to
students of the Faculty of English Linguistics and Literature (EF).

1.3. Aim of the study

It is apparent that collocational errors are made by even students with high level of
proficiency. This fact encouraged the researcher to conduct the present study, which
mainly focuses on collocational errors made by 3rd year EF students at USSH.
Precisely, the study aims at analyzing errors related to attributive adjective-noun
collocations in the students translations in order to have a comprehensive and detailed
look into (1) the common types of errors and (2) their possible causes. Also, based on
the result of the error analysis, some useful and applicable solutions are expected to
be found to help the students overcome difficulties in learning and using collocations
and particularly in translating collocations.

1.4. Research questions

In order to accomplish the aforementioned aims, the two research questions were given
as follow:
1. What are the common types of adjective-noun collocational errors made by 3rd
year EF students at USSH in their translations?
2. What are the possible causes of adjective-noun collocational errors made by 3rd
year EF students at USSH in their translations?

1.5. Significance of the study

With those aims accomplished by answering the two research questions, the present
study has its own significance as follow.

4
Firstly, in terms of theoretical values, the study provided a thorough review of
literature on relevant aspects of collocation and collocational errors. Especially, the
review of literature presented a comprehensive list of error types and hypothetical
causes of collocational errors, which was later confirmed through the error analysis.
In addition to them, the analysis revealed some other causes of errors which were
scarcely mentioned in previous studies. Besides, the practical teaching and learning of
collocation in translation classes at EF, USSH were also investigated, which revealed
possible connections to learners deficiency in collocational knowledge.

Secondly, the study had some practical contributions to the teaching and learning of
collocations, especially in translation. With a detailed analysis of types and causes of
errors, the study was expected to provide teachers and learners with valuable
information on how collocational errors are made and thus would help learners avoid
committing such errors. Moreover, basing on the findings, the study suggested some
recommendations for the teaching and learning of collocation in translation at EF,
USSH in order to further improve the students collocational knowledge and help them
overcome difficulties in translating collocations.

1.6. Scope of the study

The present study focused on adjective-noun collocational errors in translation of 3rd


year EF students at USSH. The scope of the study, however, still needs clarification.

First of all, it was mentioned that the present study focused on adjective-noun
collocational errors. This type of collocation includes combinations in which the
adjective may come either before or after the noun. However, in the present study,
adjective-noun collocation strictly refers to combinations in which the adjective
precedes the noun. To put it differently, the study took into account only attributive
adjective-noun collocations and their related errors.

Secondly, the present study was said to deal with collocational errors in translation. It
is commonly known that translation involves two directions: either from learners L1
to L2 or vice versa. To avoid unnecessary complication, the researcher directed his
attention only to Vietnamese English translation since translating from L1 to L2
seems more challenging and thus collocational errors are more likely to occur.

5
To sum up, the present study focused on errors related to attributive adjective-noun
collocations found in Vietnamese English translations of 3rd year EF students at
USSH. Such limited scope might result in the relatively weak generalizability of the
study. However, it is expected to have some contributions to translation teaching at
EF, USSH and provide theoretical and practical background for further research.

1.7. Overall structure of the study

This thesis consists of five main chapters (1) Introduction, (2) Literature Review, (3)
Methodology, (4) Findings and Discussion and (5) Conclusion and Recommendations,
apart from the Acknowledgements, Abstract, References and Appendixes.

The first chapter, namely Introduction, provides the background information of the
study and gives the rationale for carrying out the research. The aims, the significance
and the scope of the study are also presented in this chapter.

The second chapter, Literature Review, can be divided into two major parts. The first
one sets up the theoretical background for the study with a detailed review of theories
and concepts relevant to collocation and collocational errors. The second one is the
review of previous studies in the field. Based on this theoretical background, the
conceptual framework is established as a guideline for the research design and the
collection and analysis of data.

The third chapter, Methodology, describes the methods used to conduct the study. In
this chapter, the research questions are further elaborated. In addition, descriptions of
research designs and procedures for collecting and analyzing data are also presented.

The next chapter, entitled Findings and Discussion, presents the analysis and
discussion of the collected data, as well as the major findings of the studies. This
chapter not only gives the answers to the aforementioned research questions but also
serves as the basis for the recommendations presented in chapter five.

The final chapter entitled Conclusion and Recommendations provides a brief


summary of the study, pedagogical implications for the teaching and learning of
collocations in translation, as well as recommendations for further research.

6
CHAPTER 2
LITERATURE REVIEW

This chapter presents and discusses the major theories and literature necessary for
constructing the theoretical and conceptual framework of the study. Firstly, the
concepts related to collocation, including definition and typology are presented.
Secondly, the relationship between collocations and translation is discussed. Thirdly,
the literature related to error analysis is considered. Fourthly, as the present study aims
to find out some solutions to the learners errors, the literature concerning the teaching
of collocations is reviewed. Subsequently, some common types of collocation testing
method are presented in order to figure out the most appropriate methods for the
present study. Finally, based on that theoretical framework, the conceptual framework
is established as the guideline for the research.

2.1. Collocation

2.1.1. Definition of collocation

The term collocation, according to the Oxford Advanced Learners Dictionary, has
two different meanings. Collocation, as a countable noun, refers to a combination of
words in a language that happens very often and more frequently than would happen
by chance. As an uncountable noun, it refers to the fact of two or more words often
being used together, in a way that happens more frequently than would happen by
chance. Therefore, collocation in this study may be used to address either the concept
of collocation or certain combinations.

The term was made popular by Firth (1957) who used it to refer to the habitual or
characteristic associations of words in texts (cited in Ibrahim, 2003, p. 3). Firth was
considered the pioneer of the frequency-based approach one of the two major schools
of thought in studying collocation. Following this approach, several definitions of
collocation have been provided. For example, Jones & Sinclair (1974) viewed a
collocation as the co-occurrence of two items in a text within a specified
environment (cited in Ibrahim, 2003, p. 4). Later, Cruse (1986) described collocations
as sequences of lexical items which habitually co-occur. More recently, Sinclair

7
defined collocation as the occurrence of two or more words within a short space of
each other in a text (1991, p. 70). In general, most of the definitions based on this
school of thought have taken frequency of occurrence as the criterion to define
collocation.

However, a group of frequently co-occurred words is not necessarily a collocation.


This was proven by Kjellmer (1982) who studied Brown Corpus and found the
frequent occurrences of some adjacent two-word sequences such as of the,
although he, and but too (cited in Kurosaki, 2012). Although these words
frequently co-occurred in the corpus, most native speakers would probably not
consider them as collocations, since they did not make meaningful lexical sequences.
To put it differently, frequency of occurrence alone is not sufficient to define
collocation.

The phraseology approach is another school of thought regarding collocation.


Followers of this approach have made efforts to distinguish collocations from other
types of formulaic sequence and from free combinations. Howarth (1998), for
example, differentiated restricted collocations from free combinations, figurative
idioms, and pure idioms using two criteria (1) the degree of restriction on substitution
and (2) the meaning of words in a combination (literal or figurative). Howarts idea
was based on Cowies (1981), one of the pioneers of the phraseological approach, who
suggested that collocations were found in the fuzzy area on a continuum between
free combinations and idioms. Cowie also proposed two criteria to distinguish them
from each other, which are: combinability and transparency. The former refers not
only to co-occurrence but also to restriction in combination. The degree of restriction
varies from free substitution, which allows one word to combine with an open set of
words, to complete restriction, which allows one word to go with only one specific
word (or very few options) in a specific combination. The latter refers to the meaning
of the words in combinations, whether they have literal meaning [+transparency] or
figurative meaning [-transparency].

The two approaches above have provided important criteria to define collocation.
However, both of them had their own strengths and weaknesses. The frequency-based
approach, with its theory about frequency of occurrence, laid the foundation for

8
defining collocation. However, it failed to provide a comprehensive method to
distinguish collocations from other formulaic sequences, not to mention meaningless
strings of words that frequently co-occurred. The phraseological approach, with the
two concepts combinability and transparency, gave a solution to the mentioned
problem. However, definitions based on this approach would be incomplete without
the criterion set by the other one. Therefore, Kurosaki (2012) combined the
aforementioned ideas and came up with his own definition as follow:

A collocation is a type of word combination in a certain grammatical pattern, and


they refer to an abstract unit of language that occurs frequently. Collocations are
characterized by two criteria: (1) combinability of words within a collocation; and
(2) semantic transparency of word in a collocation. (p. 30).

Kurosakis definition embraced the essence of the two major approaches. Moreover,
by pointing out that collocations follow certain grammatical patterns he also excluded
the cases of adjacent two-word sequences described in Kjellmer (1982) as these
meaningless strings of words do not adhere to these grammatical patterns.

This definition, nevertheless, had certain limitations. Firstly, it did not state the
frequency of occurrence, at which a combination could be considered a collocation.
Secondly, the use of transparency as a criterion proposed by Cowie (1981) was
critiqued by Hodne (2009) for not having a clear-cut boundary between the two
extremes literal and figurative. Taking two combinations constitutional monarchy
and customs agents as examples, she argued that it was hard to determine whether
they are free combinations or restricted collocations, and to tell which element in each
of the combinations is used with its literal sense if they are restricted collocations.
Hodne, therefore, proposed a definition of her own:

Collocation are arbitrarily restricted lexeme combinations that are syntactically


fixed to a certain degree, are included in the collocation dictionary, present an MI
score [Mutual Information] higher than 3.0, and have a raw frequency [or FREQ
for short] of more than three tokens in COCA [the Corpus of Contemporary
American English]. (p. 8)

9
Although the definition above provided reliable criteria for identifying collocations,
the use of both corpora and dictionaries was unnecessarily complicated. In fact, it is
suggested that using corpora alone would suffice as they provide a much larger and
more comprehensive database.

In addition to the aforementioned issue, there is another point in the definition that
need consideration. In fact, Hodne was not the first person to use frequency of
occurrence in a corpus as a criterion to determine collocations. Prior to her, Nesselhauf
(2005) had applied this to his research. Later on, it was adopted in Hong et al. (2011).
Both studies had set the FREQ of at least five tokens in BNC (the British National
Corpus) as the standard threshold for a combination to be considered a collocation.
Given the fact that COCA has a much bigger database than BNC, the FREQ of more
than three tokens set by Hodne (2009) may not be representative enough. Therefore,
the present study will adopt the standard FREQ of at least five tokens proposed by
Nesselhauf (2005) and Hong et al. (2011).

Based on the definition of Hodne (2009) and the aforementioned recommendations,


the operational definition for the present study is formed as follow: Collocations are
arbitrarily restricted lexeme combinations that are syntactically fixed to a certain
degree, present a minimum Mutual Information (MI) score of 3.0, and have a raw
frequency (FREQ) of at least five tokens in COCA.

According to this definition, the FREQ and the MI score of a combination in COCA
are the main criteria to determine whether the combination is a collocation or not.
Detailed descriptions of FREQ and MI will be given later in Chapter 3. However, it
can be briefly explained here that the former shows how frequently the combination
is used by native speakers while the latter indicates how strong the combination is.
With these criteria taken into account, there are four possibilities to consider.

(1) If both criteria are satisfied (FREQ 5; MI 3), meaning that the combination is
frequently used and has a strong combination between the words, it is considered a
collocation.

(2) If neither criterion is satisfied (FREQ < 5; MI < 3), the combination is neither
frequently used nor strong enough and thus it is considered an erroneous one.

10
(3) If only the FREQ meets the standard requirement (FREQ 5; MI < 3), which
means that the combination is frequently used by native speakers but is not strong
enough to be a collocation, it is considered a free combination.

The last possibility (4) only the MI meets the requirement (FREQ < 5; MI 3) is
a little complicated. In her study, Dang (2014) considered these as free combinations.
However, there are several issues that need consideration. First of all, a low frequency
of occurrence means that the combination is not commonly used. In other words, it is
possibly not recognized or accepted by the majority of native speakers regardless how
high its MI score is. Moreover, according to Clear (1995), the MI score is high and
unreliable when the frequency of occurrence is low. Therefore, in the present study,
combinations with only the MI meeting the requirement will be also considered as
erroneous ones.

This operational definition and the four possibilities resulted from it provide important
criteria for the data collection and analysis procedures, which will be discussed later
in Chapter 3.

2.1.2. Categorization of collocations

From different perspectives, collocations can be categorized differently. However,


most scholars agreed with the idea of Benson et al. (1986) who categorized
collocations into grammatical collocations and lexical ones. The former, according to
Fontenelle (1994), involves one element from an open class and an element from a
closed class, typically, but not necessarily, a preposition (p. 4). The latter, on the
contrary, consists of two items belonging to open (non-finite) classes, for instance a
verb and a noun or an adjective and a noun (ibid.)

The present study, however, only focuses on the second type: lexical collocations,
since they are more challenging and demanding to learners in a way that they require
a more selective and precise use of vocabularies. This type of collocation, again, is
classified into different subtypes. Farrokh (2012), basing on Benson et al. (1986),
proposed 7 types of lexical collocation, which are shown in Table 2.1 below.

11
Table 2.1: Seven types of lexical collocations (Farrokh, 2012, p. 59)
Types Examples
1. verb (creation/activation) + noun compose music; make an impression
2. verb (eradication/nullification) + noun revoke a license, demolish a house
3. adjective + noun strong tea, a rough estimate
4. noun + verb bees buzz, bombs explode
5. noun1 (+of) + noun2 pack of dog, a herd of buffalo
6. adjective + adverb/ adverb + adjective sound asleep, hopelessly addictive
7. verb + adverb anchor firmly, argue heatedly

However, according to Benson et al. (1986), in English, nouns are often used as
adjectives. Nouns used attributively may enter into the category of adjective-noun
collocations. To put it differently, the third category may include combinations
consisting of two nouns, the first one of which functions as an adjective modifying the
second one. This category, therefore, should be renamed: adjective/noun + noun.
Moreover, it is noticeable that the first and the second categories are similar in terms
of components: verb + noun. Since verb-noun collocation is not the focus of the
present study, categorizing verbs into different types is unnecessarily complicated.
The two categories, therefore, will be grouped into one, namely verb + noun, resulted
in six types of lexical collocation shown in Table 2.2 below

Table 2.2: The revised six types of lexical collocations


Types Examples
1. verb + noun revoke a license, make an impression
2. adjective/noun + noun strong tea, jet engine, aptitude test
3. noun + verb bees buzz, bombs explode
4. noun1 (+ prep) + noun2 pack of dog, a herd of buffalo
5. adjective + adverb/ adverb + adjective sound asleep, hopelessly addictive
6. verb + adverb anchor firmly, argue heatedly

As mentioned in Chapter 1, the present study only focuses on one type of lexical
collocation: adjective-noun, due to the difficulties this type of collocation causes for
learners and the small number of studies concerning it. Although this type of
collocation involves both adjective + noun and noun + noun combinations, for the sake
of simplicity and consistency, both of them will be addressed as adjective-noun
collocation since the first noun in the combination plays the role of an adjective.

12
2.1.3. Components of collocations

Since the present study involves identifying collocations in texts and checking them
using a corpus (COCA), it is necessary to have some basic knowledge about the
components of a collocation, among which are three important terms that need
consideration: node, collocate and span (Ibrahim, 2003).

Node was defined by Jones & Sinclair (1974) as an item whose total pattern of co-
occurrence with other words is under examination (cited in Ibrahim, 2003, p. 20);
while collocate was defined as any item which appears with the node within a
specified environment (ibid.). Jones & Sinclair (1947) have also made clear that
there is no difference in status between node and collocate; if word A is a node and
word B is one of its collocates, when word B is studied as a node, word A will be one
of its collocates (ibid.).

However, Kurosaki (2012) suggested that for each specific study, node and collocate
should be identified at the beginning to ensure accuracy and consistency of the study.
Stubbs (2001) also stated that a node predicts that a preceding or following word also
occurs (cited in Kurosaki, 2012, p.28). That is to say, in a collocation, the node is the
one that determines the choice of collocates. Since the present study deals with
English adjective-noun collocations which are basically noun phrases, and since in a
noun phrase the head noun is the pivotal element, the noun will be considered as the
node and the adjective will be considered as the collocate.

The last factor to be considered is span. According to Jones & Sinclair (1974), span
is the amount of text within which collocation between items is said to occur (cited
in Ibrahim, 2003, p. 20). Phillips (1985), sharing the same view, remarked that
collocation is recognised within an environment of a number of words preceding
and/or succeeding the node, for example, the five preceding and the five following
words. This environment is termed the span (ibid.). Let us consider the following
examples: He is telling the truth and He is telling only half of the truth. Both
sentences have tell as the node and truth as the collocate of the collocation; however
their spans are obviously different. The span of a collocation, according to Ibrahim,
can be even above phrase level (2003, p. 21). However, with respect to attributive

13
adjective-noun collocations, by nature the adjective usually adjacently precedes the
noun, or sometimes it precedes the noun within a span of three or four words when
multiple adjectives/nouns pre-modify the same head noun.

To sum up, in this study, the node of an adjective-noun collocation is the noun, the
collocate is the adjective (or the noun functioning as an attributive adjective in the
case of noun + noun) and the span being considered is within five words preceding
the node. These components and relevant rules mentioned above play important roles
in the data collection and analysis, which will be discussed in detail in Chapter 3.

2.2. Collocations and translation

2.2.1. The significance of collocations in translation

This section is dedicated to discussing the relationship between collocations and


translation. The former, indeed, has considerable influence upon the latter.

One factor determining a translators performance is his lexical knowledge; the


broader it is, the better his performance is likely to be. A translator, therefore, must
always enrich his vocabulary. However, the acquisition and retention of vocabulary
prove to be a big challenge for most non-native speakers of any language (McCarten,
2007). Moreover, that learners know a word does not mean that they can use them in
communication, let alone in translation. The major cause for this is their inability to
determine in which contexts and with which words a lexical item may occur. To put
it differently, it is their lack of collocational knowledge that accounts for the issue.

Many authors, including Pawley & Syder (1983) and Harmer (2001), have suggested
learning vocabulary as chunks as a solution to the aforementioned problem. Instead
of learning separate words, learners should put them in specific contexts where they
can be learned together with other words that habitually co-occur with them. This
learning strategy has certain advantages. First of all, it helps learners memorize words
more easily since collocational association can act as memory aids (Rahimi &
Sahragard., 2008, p. 15). Moreover, as Firth claimed that one shall know a word by
the company it keeps (cited in Gyllstad, 2007, p.7), collocations provide learners
with better understanding of vocabularies. For example, learners can realize the
differences in meaning between two synonyms (e.g. injury and wound) by looking at

14
the words collocating with them. Finally, remembering words as chunks also reduces
the processing time in language use since learners do not have to spend much time on
choosing words for combinations anymore.

With a large and systematic bank of vocabulary, translators can improve their
performance in other aspects of language use, including listening, reading, writing,
and speaking. In other words, knowledge about collocations indirectly facilitates their
translation competence as Nation (2001) remarked all fluent and appropriate
language requires collocational knowledge (p. 318). McIntosh et al. (2009), sharing
the same view, wrote that a learner with knowledge of collocations will be able to
express himself clearly. Not only does collocational knowledge promote translators
fluency, but it also helps improve the quality of their translations in both naturalness
and accuracy. One of the factors distinguish native speakers from non-native ones is
the naturalness in their language use, marked by the correct use of collocations and
other idiomatic expressions. That is to say, having broad knowledge of collocations
and being able to use it efficiently are among the important factors making non-native
speakers sound native-like. In certain situations, the use of collocations also saves
translators from producing lengthy and unnatural expressions. Furthermore, according
to Newmark (1988), collocations are an important unit of translation that needs special
attention. Some words in certain collocations have meanings that differ from their
usual meanings when they stand alone. Thus having broad knowledge of collocations
will help translator understand and translate texts more accurately and easily.

2.2.2. Difficulties in translating collocations

While collocations play an important role in translation, translating them seems to be


a constant challenge to translators. In order to shed more light on the issue, the
researcher attempts to put the difficulties related to the translation of collocations into
two major categories: intralingual difficulties and interlingual ones. It is important to
note that the types of difficulties are intertwined with each other, and thus it is almost
impossible to have a clear cut between them. The two categories proposed here are
just an ad-hoc solution to discuss the issue more systematically.

15
Intralingual difficulties are those caused by the target language itself. Talking about
collocation, most researchers have agreed on the fact that its arbitrary nature is a major
cause of difficulties for learners and translators. Sughair (2011), for example, wrote
that [while] a native speaker of the language can predict collocations a learner
of the same language finds it hard to collocate the words (p. 5). The difficulties also
come from a universal phenomenon of languages that a word can have different senses
depending on the words in combination with it. For example, a big house, a large
house and a great house have the same meaning, but a great man is different from a
big man or a large man (Mai, 2010). Another good example is the word dry whose
primary meaning denotes the characteristic of not being wet or moist. However, in
expression such as dry cow or dry voice, the meaning has nothing to do with not wet
or moist (Meidasari, 2007, p. 6). The problem is that translators may not know all the
possible senses of a word and even if they do, they may still fail to recognize the exact
sense in a certain combination. Considering this issue, Meidasari (2007) maintained
the importance of context because it is the context in which a word occurs that signals
its secondary meaning. However, a throughout understanding of a context may also
be another challenge for translators, especially when the context involves language-
specific and culture specific lexical items, which are major factors related to
interlingual difficulties.

Interlingual difficulties are those caused by the differences between the two languages.
Differences in terms of culture and language are great obstacles for translators. And
collocations are among the most troublesome language-specific and culture-specific
items. Dweik & Shakra (2010) pointed out that rendering Arabic collocations into
English constitutes a major linguistic and cultural hindrance due to the wide
linguistic and cultural gap between Arabic and English, which consequently led to the
lack of equivalence of specific-culture, bound collocational patterns. Similarly,
Shammas (2013), cited Emery (1991), wrote that collocations are language-specific
and hence unpredictable (p. 108). The gaps between cultures and languages are a
universal problem for translators and language learners; and thus are also observable
in the case of English and Vietnamese. A Vietnamese translator, for example, may
find it hard to translate culture-specific collocations such as continental breakfast or
tow-away zone (Trinh, 2001, p. 68) which do not have equivalents in Vietnamese.

16
Similarly, Vietnamese culture-specific terms such as mt rng (dragons face,
denoting king) or cu/hong tuyn (nine/golden springs, denoting the afterlife) are
difficult to translate into English since translators have to render not only the exact
message but also the effect, the mod, etc. of the SL text. For archaic words such as the
ones above, they need to find appropriate archaic equivalents to convey as closely the
spirit of the source text as possible. Furthermore, when two languages are taken into
consideration, the arbitrariness of collocations is even more obvious. For instance, one
cannot explain why black horses, black dogs and black cats are referred to as nga ,
ch mc and mo mun respectively in Vietnamese. As a result, a native speaker of
English may find it very confusing for having three words which express the same
meaning of black and yet cannot be used interchangeably.

2.2.3. Influences of L1 on learners use of collocations

Differences in terms of culture and language are one of the major causes of difficulties
in translating collocations. To have deeper insight into the issue, this section is devoted
to presenting a brief contrastive analysis between English and Vietnamese collocations
and related factors.

2.2.3.1. English collocations vs. Vietnamese counterparts

Although English and Vietnamese belong to different language groups the former
is a member of Indo-European family, the latter belongs to Austroasiatic family they
still share some similarities. Firstly, they both have some common sentence structures
such as SV or SVO. Secondly, both languages use Latin writing system. Nonetheless,
these similarities are minor compared to the differences. This part, however, only
focuses on the points relevant to collocation. The differences will be categorized into
two groups: one related to the linguistic aspect of language, the other related to the
cultural one.

In terms of linguistics, it is obvious that the structure of adjective-noun collocations


in English does not have an identical equivalent in Vietnamese due to differences in
words order. English grammatical rules allow adjectives to precede head nouns in
noun phrases. On the contrary, in Vietnamese, the adjectives in a noun phrase always
come after the head noun they modify, for example: dense population vs. dn s ng.

17
In other words, the equivalent structure of English adjective-noun collocation in
Vietnamese is noun-adjective (Trinh, 2001). However, in English, nouns can have the
same attributive function as that of adjectives. The similar phenomenon is also
observed in Vietnamese where a noun, like an adjective, can work as a modifier of
another noun. As a result, certain adjective-noun collocations in English are translated
into Noun-Noun combinations in Vietnamese (e.g. golden age and thi vang son). On
the contrary, some noun-adjective combinations in Vietnamese are translated into
Noun-Noun collocations in English (e.g. rng khn and wisdom tooth). In addition to
the differences in structure, the concepts of equivalence between words and
equivalence between collocations do not necessarily coincide between the two
languages. To put it differently, even if two words in a language form a collocation,
their equivalents in another language do not necessarily have the same relationship.
These differences between the two languages are probably a major cause of difficulties
in translating collocations.

With respect to the cultural aspect, language itself is heavily influenced by culture and
collocation, as a part of language, is not an exception. Trinh (2001) conducted a
contrastive analysis between English and Vietnamese collocations, in which he gave
a long list of examples of culture specific collocations. For example, in English, the
term continental breakfast refers to a breakfast that consists of a hot roll or croissant
with some sort of spread or filling, served with coffee used on the wider European
continent, and in most tourist hotels throughout the world (Trinh, 2001, p. 68). Its
word-for-word Vietnamese translation ba sang luc i a, however, does not really
make sense. Conversely, Vietnamese also has culture specific collocations that do not
exist in English. The term ng Xanh, according to Trinh, refers to God, Heaven, or
Creator, and thus a word-for-word English translation such as Mr. Blue is not an
accurate one. These collocations are also a great challenge for translators as they
require not only collocational but also cultural knowledge.

In sum, it is clear that linguistic and cultural gaps between the two languages may
cause great difficulties for translators. These differences also give rise to the concepts
of congruent and incongruent collocations, the factors which possibly affect learners
acquisition and use of collocations, and thus are discussed in detail in the next section.

18
2.2.3.2. Congruent and incongruent collocations

Simply speaking, congruent collocation (Kurosaki, 2012) refers to equivalent


combinations, of which the components are also equivalents of each other, for
example high fever and st cao. In contrast, incongruent collocation (ibid.) refers to
equivalent combinations having inequivalent components, that is to say collocations
that do not have [direct] translation equivalents1 in L1 (Nakata, 2007). For instance,
the correct equivalent of the Vietnamese combination gi r is low price rather than
its translation equivalent cheap price.

The fact that these two types of collocations could confuse learners as they may find
it hard to tell them apart has attracted much attention. Maurer-Stroh (2005), for
instance, distinguished four types of collocations based on two criteria:
predictability and lexical congruence. The former means that two collocations are
semantically predictable, as are high fever and st cao. The latter refers to the
similarity in structure which hardly exists between English adjective-noun
collocations and their Vietnamese equivalents. Among the four types of collocations
proposed by Maurer-Stroh, the fourth type, namely unpredictable and lexically non-
congruent2, according to him, is the one that needs special attention in the
contrastive light (2005, p. 61). These concepts of congruent and incongruent
collocations were also addressed by Nakata (2007), who suggested that collocations,
especially non-congruent [incongruent] items, cannot be acquired easily through mere
exposure and are amenable to form-focused, intentional learning (p. 154).

In general, most researchers, regarding the issue, have agreed that incongruent
collocations tend to cause more difficulties for learners and thus need more attention
in teaching and learning. However, that does not mean congruent collocations should
be neglected. Since learners find it hard to distinguish between these two types of
collocations, it is possible to argue that they may not know when they can resort to
translation equivalents and when they should not. Congruent collocations are,

1
Direct translation equivalent refers to the kind of equivalent attained by translating separately the
components of a combination, or in other words through literal word-for-word translation.
2
The term unpredictable used by Maurer-Stroh is similar in meaning to the term incongruent in the present
study, while lexically non-congruent can be understood as a kind of structural dissimilarity. Therefore, this kind
of collocations proposed by Maurer-Stroh is, in fact, referring to the kind of incongruent collocations being
mentioned in the present study.

19
therefore, not necessarily easy for them. More importantly, this type of collocations
may lead learners who lack collocational knowledge to a false assumption that
collocation equivalence and direct translation equivalence are identical.

2.2.4. Strategies in translating collocations

During the process of translation, translators have to deal with different text types and
purposes. For that reason, a large number of translation methods have been proposed.
Among them, the literal word-for-word translation is one commonly used by
translators due to its simplicity and effectiveness. According to Newmark (1981), in
communicative as in semantic translation, provided that equivalent effect is secured,
the literal word-for-word translation is not only the best, it is the only valid method of
translation (cited in Munday, 2001, p. 44). Toury (1995), sharing the same view with
Newmark, added that the translator has to maximize the efficiency of the cognitive
processes by concentrating energy on especially difficult problems, devoting less
effort to those parts of the text which produce a reasonable translation by the literal
procedure (cited in Munday, 2001, p. 45).

The problem with this translation method is that it is only viable when equivalent
effect is secured (Newmark, ibid.). However, for idiomatic expressions and
particularly collocations, such an effect is not always guaranteed. If the combination
being considered is a congruent collocation, the use of this translation method is
acceptable. However, when it comes to incongruent ones, whose collocation
equivalence and translation equivalence are not identical, the literal translation of such
a combination cannot produce the desired effect. To put it differently, although word-
for-word literal translation is a popular and effective method, inappropriate use of it
in translating collocations only leads to errors.

In addition to the aforementioned method, some other translation strategies also have
the same problem. Dweik & Shakra (2010), for example, found five common
strategies adopted by students in translating collocations in religious texts from Arabic
to English: (1) synonymy, (2) generalization, (3) paraphrasing, (4) deletion, and (5)
literal translation. Similarly, Hussein (2011), studying advanced EFL learners
translations, pointed out several strategies used by the students, including (1) L1

20
transfer/literal translation, (2) substitution/paraphrase, (3) assumed synonymy, (4)
analogy and overgeneralization, (5) formal/semantic association, (6) idiomaticalness,
(7) quasi-morphological similarity, and (8) avoidance of the task. The problem is that
while strategies are usually considered as a means for translators to deal with
difficulties in translation, the use of these strategies in translating collocations, on the
contrary, may lead to faulty translations or errors. Sughair (2011) commented that
one of the big mistakes translators tend to make is adopting several strategies while
translating collocations such as using simplification, reduction, synonymy, and
paraphrasing. The higher the rate of these strategies is, the less effective the
translation is (p. 10). Considering culture gaps in translation, Newmark (1988)
suggested that the use of these strategies, though effective, may lead to serious
problems in case of cultural disparity. These ideas are in accordance with the findings
of Dweik & Shakra (2010) and Hussein (2011). The former pointed out that the five
strategies above usually led students to faulty translations, while the latter considered
the strategies as hypothetical causes of errors (p. 216). Among them, synonymy,
paraphrase and literal word-for-word translation were said to be most frequently used
and also the major causes of errors.

To sum up, wrong choice of translation methods and strategies may be a cause of
errors in translating collocations. This issue will be taken into account again in the
later section which discusses about the hypothetical causes of collocational errors.

2.3. Error analysis

So far, the researcher has presented and discussed basic concepts about collocation.
This part will be devoted to presenting the theoretical background of error analysis
which is a major focus of the present study.

2.3.1. Theory of error analysis

When it comes to error analysis, the first issue to be taken into account is the
distinction between error and mistake. Corder (1967) distinguished two types of
errors: errors of performance and errors of competence. The former, which he
addressed as mistakes, are unsystematic slips of the tongue (or pen) which happen
even with native speakers due to memory lapses, physical states. The latter, which

21
were addressed by the term error, refer to the systematic errors of the learner from
which his knowledge of the language to date can be reconstructed (p. 166 167). Ellis
(2003) maintained that errors reflect gaps in a learners knowledge and occur due
to the learners lack of knowledge, while mistakes refer to occasional lapses in
performance, happen when learners are unable to perform what they have known (p.
17). Regarding mistake and error, most authors have agreed that the latter plays a
significant role in language teaching and learning. According to Corder (1967), errors
are important because they show learners progress in learning, indicate the process
of language learning, and serve as a learning device for learners themselves.
Furthermore, errors cannot be self-corrected until further relevant input has been
provided and converted into intake by the learner (James 1998, p. 83).

Because of their significance, errors have received much attention from researchers,
especially those who have interest in error analysis. Brown (1980) defined error
analysis as the process to observe, analyze, and classify the deviations of the rules of
the second language and then to reveal the system operated by learner (p. 166).
Richards & Schmidt (2002) viewed error analysis as the study and analysis of errors
made by second language learners (p. 184). Error analyses are carried out in order to
(1) find out strategies which learners use in learning a language, (2) identify the causes
of learner errors and (3) obtain information on common difficulties in language
learning, as an aid in teaching or in the preparation of teaching materials. It is believed
that, by analyzing learners errors and identifying sources of errors, one may see the
gaps in the learners knowledge and find appropriate methods to effectively improve
their competence.

The process of error analysis involves different procedures suggested by different


authors. Corder (1974), for instance, proposed a five-step procedure: (1) collection of
learner language, (2) identification of errors, (3) classification of errors, (4)
explanation of errors, and (5) evaluation of errors seriousness. Gass & Selinker (2008)
introduced another procedure with six steps: (1) collecting data, (2) identifying errors,
(3) classifying errors, (4) quantifying errors, (5) analyzing sources, (6) remediation.
The two procedures, in fact, share many similarities. Even the final steps, despite
being addressed by different terms, are similar in nature, as they both focus on the

22
remediation of errors. However, the fourth step (quantification of errors) in Gass &
Selinkers procedure is what the procedure of Corder lacks. This step, according to
Dang (2014) is of great importance because it yields more insights into the learning
process and also specifies what kind of errors need more pedagogical intervention
(p. 37). As a result, the six-step procedure developed by Gass & Selinker (2008) will
be adopted as the framework for conducting the error analysis in the present study.

2.3.2. Types of errors related to adjective-noun collocations

With respect to collocational errors, different authors have different classification


criteria based on their perspectives and the types of collocation being considered.
Among them, there was one commonly known classification proposed by Nesselhauf
(2003), including nine types of collocational errors presented in Table 2.3 below. This
classification was, later, adapted by Hong et al. (2011) and Nguyen (2014), who
omitted several categories (i.e. preposition (noun), structure and preposition (verb))
as they refer to grammatical collocations, which were not the focus of their studies.

Table 2.3: Types of collocational errors (Nesselhauf, 2003, p. 232)


Type of mistake Example
1. Verb: Wrong choice of verb (or non-existent verb) *carry out races (hold races)
2. Noun: Wrong choice of noun (or non-existent noun) *close lacks (close gaps)
3. Usage 1: Combination exists but is not used correctly take notice (to notice)
4. Usage 2: Combination does not exist and cannot be *hold children within bounds
corrected by exchanging single elements (show children where the
boundaries lie)
5. Preposition (verb): Preposition of a prepositional verb *fail in one's exams
missing, present though unacceptable, or wrong (fail one's exams)
6. Preposition (noun): Preposition of a noun missing, *raise the question about
present though unacceptable, or wrong (raise the question of)
7. Determiner: Article or pronoun missing, present *get the permission
though unacceptable, or wrong (get permission)
8. Number: Noun used in singular instead of plural or *pass one's judgments
vice versa (pass judgment)
9. Structure: Syntactic structure wrong *make sb. friends
(make friends with sb)

Although this classification was developed by Nesselhauf for his study on verb-noun
collocation, it can be applied to other types. Kurosaki (2012) conducted a study on
several types of collocation, including adjective-noun, in which he adapted

23
Nesselhaufs classification for all of the collocation categories. The following are the
types of adjective-noun collocational errors in Kurosaki (2012)

1. Adjective: The adjective in a collocation is wrong.


2. Noun: The noun in a collocation is wrong.
3. Determiners: The article or possessive pronoun is missing or added.
4. Structure: Syntactic structure is wrong.
5. Preposition: Preposition is added through unnecessary or wrong choice.
6. Different expression: Translation does not include a collocation and/or
consists of a circumlocution.
7. Number: Noun is used in singular instead of plural or vice versa.

Since the present study focuses on adjective-noun collocations, this way of


classification is really relevant to and appropriate. However, it cannot be applied to
the present study without some modifications. Firstly, for the study focuses on
adjective-noun collocations, errors related to determiners, preposition, number and
structure may not occur frequently. In fact, Kurosaki himself found that learners made
no errors in these categories. Moreover, as grammatical ones, they are not really
relevant to the aims of the study, which only focus on lexical collocational errors.
Therefore, the four categories: determiners, preposition, number and structure should
be omitted. Secondly, although Kurosakis classification of collocational errors is very
comprehensive, the names given to the categories seem not clear enough and thus
need clarification. The types of adjective-noun collocational errors proposed by
Kurosaki (2012), as a result, are revised as follow:

1. Wrong choice of adjective: the adjective in a collocation is incorrectly chosen.


2. Wrong choice of noun: the adjective in a collocation is incorrectly chosen.
3. Wrong combination: the entire combination is incorrect or circumlocutory.

(Adapted from Kurosaki, 2012)

It is necessary to note that wrong combination, the new name for different expression,
includes the two categories usage 1 and usage 2 proposed by Nesselhaulf (2003).
Since these categories refer to erroneous combinations of which both components are
incorrectly chosen, it is unnecessarily complicated to distinguish between them in the

24
present study. Hence, they are put under the same name as mentioned above. The three
aforementioned categories of adjective-noun collocational errors will be used to
categorize the collected errors, which is an important step in data analysis.

2.3.3. Causes of collocational errors

One of the most important purposes of error analysis is to find out the remedy for
learners errors; in order to do so, identifying causes of errors is inevitably essential.
In this section, the researcher will review the relevant literature so as to identify the
hypothetical sources of error applicable to the present study.

The very first reason for learners making of collocational errors, according to
Howarth (1998) and Darvishi (2011), is their unawareness of the concept. This may
lead learners to the faulty assumption that words can co-occur without any restrictions,
although they obviously have come across the phenomenon of restricted combinability
in their mother tongue. This assumption, as a result, makes students unconsciously
produce combination violating collocational rules. This issue may be the result of a
teaching learning process in which collocations are not considered as an important
factor and thus receive little attention from teachers and students. Particularly, in
teaching translation, teachers usually focus more on translation skills and strategies,
while collocations like other aspects such as culture, idioms, etc. are only integrated
in texts for practicing, and sometimes not clearly explained by teachers. The teaching
learning process, therefore, may be a cause of errors. Hence, the research methods
will also include, besides the tests as the primary tools, an interview (with the teachers)
and a questionnaire (to the students) for deeper insight into the actual teaching
situation, so that the research can yield more accurate implication.

The lack of collocational concepts also leads to learners lack of knowledge, which
hinders their performance, especially in translation. In order to compensate for the
deficiency in knowledge, learners usually resort to certain strategies. As previously
discussed above, among translation methods and strategies that may lead to
collocational errors, literal translation, paraphrase and synonymy are common ones.
However, they are not the only causes of collocational errors. Many authors have
discussed other noticeable causes of errors.

25
Several studies, including: Farghal & Obiedat (1995), Yamashita & Jaing (2010), and
Kurosaki (2012) have concluded that interlingual transfer, which often leads to
negative transfer, is known as one major cause of errors. It is said that learners tend to
be linguistically and culturally influenced by their mother tongue due to overusing
word-for-word translation from L1 to L2, which is what most learners do when they
face a difficult expression in the foreign language. In addition to interlingual transfer,
intralingual transfer is also a factor that should not be overlooked, especially when L2
is much different from L1 and thus would possess many features unfamiliar to
learners.

Hong et al. (2011), adapted from Richards (1974) and Tarone (1981), presented a
comprehensive classification of causes of collocational errors that took into account
both interlingual and intralingual transfer as follow:

(1) L1 transliteration/L1 literal translation is included in the concept of negative


transfer mentioned in Li (2005) which is usually caused by the negative effect of
literal word-for-word translations from L1.
(2) Language switch is the use of L1 in the places where L2 should be used.
(3) False concepts hypothesized errors are those derived from faulty comprehension
of distinctions in the target language (cited in Dang, 2014, p. 35).
(4) Overgeneralization refers to the creation of an incorrect structure based on two
structures learners have acquired in L2.
(5) Ignorance of rules restrictions is failure to observe the restrictions of existing
structures, that is, the application of rules to contexts where they do not apply
(ibid.)
(6) Approximation happens when learners use an incorrect lexical item or structure
sharing semantic features similar to those of the desired item.

More recently, Dang (2014) adapted from Li (2005) the classification of causes of
collocational errors. The list included some elements similar to the ones proposed by
Hong et. al (2011) above. However, the two causes overgeneralization and ignorance
of rules restrictions were excluded as they accounted for grammatical errors. The list
thus consisted of the following elements: (1) false concepts hypothesized, (2) the use
of synonyms, (3) negative transfer, (4) word coinage and (5) approximation. The

26
second cause of errors, namely the use of synonyms, refers to the use of synonymous
words/expressions to replace a desired item. The fourth one refers to the act of making
new words to express a desired idea.

Another comprehensive list of hypothetical causes of collocational errors is the one


proposed by Hussein (2011), consisting of eight categories as follow:

(1) L1 transfer/literal translation refers to students tendency to manipulate their L1 in


their L2 translation when they felt defective in authentic linguistic resources.
(2) Substitution/paraphrase is used as a compensatory strategy motivated by a
substitute option based on some sense relationship, or certain semantic properties
(e.g. education job instead of teaching profession).
(3) Assumed synonymity refers to the use a synonym or a near synonym in place of
the appropriate word.
(4) Analogy and overgeneralization is a psychological tendency of students to extend
the meaning of a certain word to other semantic situations where that word does
not reasonably appeal. (e.g. Students know the collocation thick soup and thus
produce a combination like thick tea instead of strong tea)
(5) Formal/semantic association happens when students conceive some sort of formal
or semantic link or affinity between the constituents to be collocated (e.g. He
remembered (reminded) them of that)
(6) Idiomaticalness refers to students tendency to contrive idiomatic forms in English
parallel to those in their L1.
(7) Quasi-morphological similarity occurs due to defective learning, students may feel
that some linguistic forms resound or echo other words and use those forms instead
of the correct ones (e.g. a violation (violent) attack)
(8) Avoidance/abandonment of the task means that students left out the expressions,
which may be the result of ignorance, carelessness, failure to recollect/recall from
memory, and limited time constraints.

Although these classifications are different from each other in components and names,
they in fact share many similarities. First of all, literal translation, the translation
method mentioned by several authors including Hong et al. (2011) as a major cause of
collocational errors, is actually one of the factors accounting for negative interlingual

27
transfer, and thus can be included in the bigger category of L1 Transfer. Assumed
synonymity and use of synonyms are only different names of the same strategy, with
which learners use a lexical item to replace its synonymous word/expression. In
addition, false concepts hypothesized, which refers learners inability to make
distinction between similar words in the target language, may also account for
learners use of synonyms. Similarly, approximation, word coinage, formal/semantic
association and quasi-morphological similarity, according to Hong et al. (2011) and
Dang (2014), are different forms of paraphrase. However, as they involve errors with
different characteristics, it is difficult to put them in one category. Among them,
formal/semantic association and quasi-morphological similarity reflect some sorts of
confusions over forms of words and thus will be grouped into one category, namely
formal confusion, the term which was used by Johansson (2008) to address this type
of error. In her research, Dang (2014) found several errors resulted from the
misplacement of parts of speech and grouped them in the category of approximation.
However, as these errors are also related to form, they will be considered as a type of
formal confusion in the present study. Although errors caused by word coinage are
also related to form, this cause of errors is considered as a separate category due to its
nature, which is different from that of formal confusion.

To sum up, based on the aforementioned classifications, a list of hypothetical causes


of collocational errors is formulated for the present study, which includes: (1)
approximation, (2) use of synonyms, (3) L1 transfer, (4) word coinage, (5) formal
confusion and (6) abandonment of tasks. These hypothetical causes of errors are of
great importance in data collection and analysis as they guide the researcher in
designing the research tools and identifying the causes of the collected errors.

2.4. Collocation and language teaching

The primary purpose of error analysis is to find the remedy for learners errors.
Through the analysis of error, researchers can figure out the appropriate methods for
teaching collocations. This objective can hardly be achieved without basic knowledge
about the teaching of collocations in EFL/ESL. This part, therefore, will be dedicated
to reviewing some relevant literature in the field.

28
The very first questions to be asked when it comes to teaching collocations is should
collocations be taught implicitly or explicitly? What makes the issue controversial is
the assumption that learners can figure out the concept of collocation and acquire
knowledge by themselves through implicit teaching. However, empirical studies have
yielded a contradictory result. Nakata (2007), for example, found that collocation can
hardly be acquired through exposure and thus suggested intentional teaching/learning
of collocations. Dang (2014) also proposed similar ideas. Arguing that collocation is
a universal phenomenon existing in every language, yet learners usually come across
the concept unconsciously without recognizing it, the author suggested that the
concept of collocation should be introduced and explained to learners (p. 25). She
gave the erroneous combination expensive price as an example and explained that to
avoid causing confusion for learner, teachers should draw their awareness to the
notion of collocation rather than giving an arbitrary explanation such as we just do
not usually say so (ibid.)

If collocations should be taught explicitly, the next question should be what kinds of
collocations should be chosen to teach? It is commonly seen that some teachers tend
to focus more on grammatical collocations by correcting learners grammatical errors
such as verb-preposition or preposition-noun compositions, etc. while lexical
collocations and related errors are usually neglected. The issue leads to the suggestion
that more attention should be directed to the latter kind of collocations. In
consideration of lexical collocations, Bahns & Eldaw (1993) have suggested that
EFL teaching should concentrate on those collocations which cannot readily be
paraphrased (p. 101). More recently, Nakata (2007) and Kurosaki (2012) have
remarked that it is incongruent collocations that need special attention. Hodne (2009),
holding a similar point of view, maintained that lexical collocations selected for
teaching should be challenging to but at the same time practical and useful for learners
to acquire. On this basis, she also suggested a comprehensive list of useful
collocations to introduce to students, which is shown in Table 2.4 below. The list can
indeed be used as a guideline for teachers on choosing appropriate collocations to
introduce to students and designing more effective teaching activities.

29
Table 2.4: Useful collocations to teach students (Hodne, 2009, p. 104)
Types of collocation Examples
with words that appear repeatedly in the general election, free election, hold an
exercises election
with words that are already familiar to incredibly busy, trendy restaurants
students but which expand the knowledge of
the known word
related to chapter themes constitutional monarchy (government in
the UK), vocational education (education)
when teaching infrequent words, those grow weary and auspicious start
containing the infrequent word and a known
word
frequent and strong in corpora pledge allegiance, environmentally friendly
Collocations of near synonyms to avoid notable example and important point
using them interchangeably independent of
context
non-congruent with L1 equivalent poor health, dense population
evoking communicative situation hectic schedule, feel frustrated
useful to talk about news and current events high unemployment, cause consternation
revealing various meanings of a word bright future and bright light, cast a vote
and cast doubt
conveying sociocultural aspects (These do affluent suburbs, white prom
not need to be directly taught, but brought to
students attention in order to foment
discussions about cultural issues.)

In addition to the aforementioned issues, methods for teaching collocations are also a
factor that needs consideration. Hill (2000), as cited in Dang (2014), stated that in
order to teach collocation, we have to give it the same kind of status in our
methodology as other aspects of language such as pronunciation, intonation, stress,
and grammar (p. 25). From this perspective, studies on collocation-teaching methods
have been conducted.

Hodne (2009) not only pointed out the useful collocations for teaching but also
introduced a variety of methods for teaching them. She suggested several kinds of
activities including matching, in which learners were asked to match a node with a
collocate/group of collocates or a combination with its equivalent, and filling, which
required learners to complete certain collocations with words found in a text. These
activities, according to her, are helpful in teaching collocations with close synonyms,
introducing new collocations containing familiar words and making use collocations

30
available in course books, etc. Besides, she also highlighted the aid of corpora, with
which activities can be designed to teach culture through collocations or help learners
discover words meanings.

Farrokh (2012) discussed several techniques for teaching collocations. He first


introduced an activity using grids designed by Channell (1981) which was similar to
the one proposed by Hodne (2009) in that learners were required to form potential
collocations based on the given nodes and collocates. While Farrokh did not negate
the usefulness of this activity he argued that such kind of activity may lack
effectiveness as it did not provide a context. Regarding this issue, he introduced a
variety of context-bound approaches which make use of texts, dictionaries as well as
corpora and concordances. Among them, corpora and concordances were believed to
help teachers decide what collocations to teach, provide different contexts in which
collocations appear, and increase learners fluency and naturalness in language use.
Not only did Farrokh focus on input but he also highlighted the importance of output
in teaching collocations. Regarding output as an important factor promoting L2
learning, he remarked that output tasks possibly led to significant acquisition of
collocations.

Pirmoradian & Tabatabaei (2012) examined the effectiveness of a corpus-based


approach, namely Data Driven Learning (DDL), in which Collins Collocation
Dictionary was used as a concordancing tool for learning lexical collocations. It is
believed that the DDL approach, proposed by Johns (1991), can help learners make
their own discoveries about language use, through a concordance and a corpus used
as a database. The study involved two groups: an experimental group and a control
one. During the experiment, the control group was asked to pay attention to
collocations and mis-collocations in some given texts, whereas the treatment group
had access to collocations via the dictionary software installed on computers. The
results showed that the experimental group performed significantly better on lexical
collocations than the control group, which proved the effectiveness of the DDL
approach.

Using corpora in teaching collocations was taken into account again in Jafarpour et
al. (2013) which aimed to test the hypothesis that the traditional method of teaching

31
collocation is not as helpful as the corpus-based approach. This was done by
comparing the effects of the two approaches on learning collocations of near-
synonymous pairs. The result revealed that the experimental group, thanks to the aid
of concordancers from BNC in learning collocations, had better performance and
production of the collocations. The corpus-based approach also proved its superiority
to the traditional one as it provided authentic data helping learners discover
collocational knowledge by themselves.

More recently, Dang (2014) investigated the collocational knowledge of English-


majored students and recommended relevant pedagogical implications. The author
highlighted the importance of explicit instruction as well as using corpora and
concordancers in teaching and learning of collocations. In addition, she combined the
ideas of Woolard (2000), Hill (2000), Lewis (2000) and Boonyasaquan (2009) to
formulate a comprehensive list of classroom procedures for teaching collocations as
follow:

(1) Introducing the term collocation


(2) Convincing learners of the importance of collocations
(3) Turning learners attention to collocations in given texts
(4) Carefully choosing what collocation to teach
(5) Encouraging learners to actively collect and study collocations on their own
(6) Repeating and recording already-met collocations
(7) Identifying and correcting errors
(8) Using corpora and concordancers
(Dang, 2014, p. 29)

Based on this, she developed a question to investigate teachers techniques for


teaching collocations in Academic Writing classes, in response to which the student
participants had to choose from the aforementioned list the techniques used by their
teachers. Thanks to its design, the question required little time to answer, yet it yielded
accurate and comprehensive information for the research. Since the present study also
takes the actual teaching of collocations as an important source of data, such an
effective question is really necessary and thus will be adapted for the questionnaire on
the teaching of collocations in Basic Translation described later in Chapter 3.

32
To sum up, the studies mentioned above have shown the necessity of teaching
collocations explicitly. Moreover, to ensure the effectiveness in teaching, collocations
should be chosen carefully following specific criteria. Methods for teaching
collocations are also an essential factor. Among them, methods that allow teaching
collocations in contexts, as discussed above, seem to be more effective and preferable.

2.5. Methods for testing collocational knowledge

This part is devoted to presenting and discussing methods for testing collocations,
from which the researcher will choose the appropriate methods that can be used in the
present study. The methods to be discussed are divided into three main categories:
corpus-based, translation-based and test-based.

Corpus-based methods. Methods of this type have been the favorite choice of many
researchers studying collocations. Corpora are used as the sources of data in these
methods. Usually, researchers choose at least two corpora/sub-corpora; one of which
is a corpus produced by learners of English (e.g. Taiwanese Learner Corpus of
English, German sub-corpus of ICLE, etc.), the other is a corpus produced by native
speakers of English (e.g. British National Corpus, Corpus of Contemporary American
English, etc.). The data are extracted from the learners corpus and compared with the
native speakers corpus. The strength of these methods is that they provide researchers
with a great amount of authentic data. Since a corpus is usually a large collection of
many types of text at different places and times, the data from a corpus are usually
more representative and more accurate than those collected by other methods. Their
weakness is that an appropriate corpus is not always available for researchers to use;
and it also takes a great amount of time and effort to analyze the corpora.

Translation-based methods. Similar to corpus-based methods, translation-based


ones are also the methods of choice. Researchers, using these methods, usually design
some kinds of translation test which require learners to translate phrases, sentences or
even whole texts from L1 to L2 or vice versa. It allows researchers to investigate the
learners productive knowledge of collocations and stimulate them to elicit their
knowledge of collocations based on their L1 version of collocations (Kurosaki, 2012,
p. 86). Moreover, researchers, using these methods, can modify the tests in the way

33
that helps them collect sufficient amount of data on the right type of collocation.
However, since translation is a complicated process with many factors that can affect
learners performance, researchers need to design tests carefully by choosing
appropriate lexical items, providing contexts, etc. to yield valid and reliable results.

Test-based methods. This category comprises of many types of test such as cloze-
test, multiple choice, COLLEX (collocating lexis), COLLMATCH (collocate
matching), etc. The two last test types (i.e. COLLEX and COLLMATCH) were
developed by Gyllstad (2007) and later adapted by Nguyen (2014). COLLEX has the
format similar to that of multiple choice; while COLLMATCH provides a grid of
words in which learners have to tick in the boxes to choose the correct combinations.
These types of test are commonly used to investigate learners receptive skill. Cloze-
test, specifically, can also be used to examine productive skill. These tests are easy to
score and analyze the results, compared to the two aforementioned methods. However,
as most researchers using these tests have stated, it is difficult to design test items. The
most difficult step, according to them, is choosing the collocations to put into the tests.
Moreover, careful processing is needed to ensure tests validity and reliability.

Since each method has its own strengths and weaknesses, it seems impossible to
choose a perfect one among them. Therefore, as it is suggested in Kurosaki (2012), a
combination of multiple methods is employed in this study. The corpus-based method,
with its advantages, is the first one to be chosen. Since there is no official Vietnamese-
English corpus available, the students translation exam papers are used as the source
of data in the present study. However, the amount of data collected might be
insufficient since the availability of adjective-noun combinations in the students
papers are unpredictable and uncontrollable.

To make up for that limitation, a translation-based method is employed. With a


carefully designed and administrated test, one could stimulate other aspects of the
students' errors which might not be sufficiently reflected in their exam papers. The
problem is that there are many types of test to choose from. Some of them require
learners to translate the whole sentences or texts, the others only require to translate
phrases. For the present study only investigates learners use of collocations,
translating whole sentences or texts might bring unnecessary difficulties to students,

34
and thus negatively affect the reliability of the test. Therefore, the test format proposed
by Kurosaki (2012), which only focuses on translating the target collocations in given
contexts, is chosen.

The two methods above, however, only focus on the students productive knowledge.
In order to have a thorough view of learners collocational knowledge, it is advisable
that their receptive ability should also be investigated. For that reason, the COLLEX
test developed by Gyllstad (2007), among the tests of receptive knowledge mentioned
above, is chosen due to its simplicity in preparation and design.

To sum up, three different methods of testing collocations are used in this study for a
more complete view on the issue. Further descriptions of the methods will be
presented later in Chapter 3.

2.6. Conceptual framework

This section is dedicated to further elaborating on the conceptual framework, which is


formulated based on the major theories presented and discussed above. The framework
significantly serves as a guide for the researcher to carry out the study and achieve the
research goal.

First of all, as the present study lays its emphasis on error analysis, the procedure of
error analysis proposed by Gass & Selinker (2008) is chosen as an important
component in the conceptual framework. With a detailed description of six steps: (1)
collecting data, (2) identifying errors, (3) classifying errors, (4) quantifying errors, (5)
analyzing source, and (6) remediation, the procedure provides a comprehensive
instruction on how to collect, analyze and interpret data.

In addition, for the study deals with collocation and collocational errors, a definition
of collocation is also essential. The definition suggested by Hodne (2009) therefore is
adapted as the operational definition, which defines a collocation as a combination
that appears in COCA (1) with a minimum MI of 3.0 and (2) with at least five-time
frequency. This definition is of great importance as it establishes the criteria to decide
whether a combination is a correct collocation or not, which is essential for the
identification of errors.

35
Categorization of errors is one of the important steps in error analysis. It not only
makes the analysis more systematic but also provides valuable information on the
factors in relation to the error types that need special attention. For that reason, the
typology of errors formulated based on Nesselhauf (2003) and Kurosaki (2012) is
included in the framework. Although it consists of only three categories (1) wrong
choice of adjective, (2) wrong choice of noun and (3) wrong combination, it is believed
to be simple and yet comprehensive enough for categorizing errors.

Another important and also difficult step in error analysis is identifying causes of
errors. Since the making of errors is a complicated mental process, involving different
factors, it is impossible to state with certainty that an error is the result of a specific
factor. It is even more difficult to support the identification of the causes by empirical
evidences, for there are few methods to completely investigate and monitor mental
processes. However, as suggested by most previous studies in the fields, such a goal,
to a certain extent, can be achieved thanks to an appropriate list of hypothetical causes
of errors. As errors caused by the same factor usually share similar characteristics, one
can make inferences about the cause of an error by comparing its characteristics with
those of the given hypothetical list. In a nutshell, a list of hypothetical sources of errors
is of great importance for identifying causes of errors. For that reason, the
comprehensive list consisting of six hypothetical causes of errors: (1) approximation,
(2) the use of synonyms, (3) L1 transfer, (4) word coinage, (5) formal confusion and
(6) abandonment of tasks adapted from multiple authors including Hong et al. (2011)
and Hussein (2011), etc. is employed in the present study.

The aforementioned list not only helps in identifying causes of errors but of also
provides a guideline for choosing the target collocations for the translation and the
COLLEX tests, which will be describe later in Chapter 3. In order for the tests to be
effective the chosen collocations must have the potential to reflect the causes of errors.
For example, since L1 transfer is said to be a possible cause of errors, the researcher
needs to include some incongruent collocations in the test to see if learners are
influenced by it.

The errors for analysis, as previously mentioned, are collected from learners exam
papers and through the translation test. In addition to these instruments, the COLLEX

36
test, which aims to test their receptive knowledge of collocations, is included for a
more complete picture of their collocational knowledge. Besides, the actual teaching
and learning of collocations are also investigated through additional data collected
from the teacher and student participants.

Regarding the teachers, their (1) perception of collocation, (2) practices in teaching
and (3) observation of students are the factors to be considered. Their perception
provides a general view on how they regard collocations, which probably has certain
influences on their teaching. Their practices in teaching show whether and how they
teach collocation, which yield valuable information for the remediation. Last but not
least, their observation of students is also a valuable source of information, since no
one has better understanding of the students than the teacher. The information is
collected through an interview which will be presented in Chapter 3.

With respect to the students, the factors to be taken into account are their (1) perception
of collocation, (2) acquisition and use of collocations, and (3) view on teachers
teaching, which were collected through a questionnaire. The students perception may
affect their learning and thus needs consideration. For their acquisition and use, the
collected information aims to support and strengthen the findings in the analysis of
errors, which is the main method to investigate the students use of collocation.
Finally, their view on teachers teaching provides information about their awareness
of teachers instructions, which may also yield valuable information for the
remediation.

Finally, based on the result of the error analysis and those additional data about their
receptive knowledge and the teaching/learning of collocations, recommendations can
be formulated in order to tackle the addressed issues and remedy the errors.

37
Figure 2.1: Conceptual framework of the study

To sum up, this conceptual framework is of great importance as it provides a thorough


guidance for carrying out the study. It not only helps designing the research tools but
also puts the analysis of error in a manageable procedure with sufficient supporting
theories at each step. With the result of the analysis of error as the main source of
information and additional data from the teachers and the students, it is expected that
the remediation of errors will be more appropriate and sufficient. An illustration of the
framework is given in Figure 2.1 above.

38
CHAPTER 3
METHODOLOGY

This chapter is dedicated to presenting the research methodology employed in the


current study. Firstly, the research design is introduced. Next, detailed descriptions of
context of the study, participants, research materials and research tools are presented.
Finally, at the end of this chapter are the procedures of data collection and analysis.

3.1. Research questions

The present study aimed at answering two main research questions: (1) what are the
common types of adjective-noun collocational errors made by 3rd year EF students at
USSH in their translations? and (2) what are the possible causes of adjective-noun
collocational errors made by 3rd year EF students at USSH in their translations?.
These two questions are expected to be answered through the analysis of collocational
errors collected from the two main research instruments (i.e. the students exam papers
and the translation test).

However, identifying the causes of errors is a complicated process, which needs as


many data to support as possible. Therefore, as discussed in the conceptual framework,
additional information regarding the actual teaching/learning of collocations in
translation classes and learners receptive knowledge of collocation is also collected
through the other instruments (i.e. the COLLEX test, the questionnaire to the student
respondents, and the interview with the teacher participants), which will be presented
later in this chapter. These additional data are expected to yield not only more evidence
supporting the identification of the errors causes but also valuable information that
helps improve the current teaching/learning of collocations if necessary.

3.2. Research design

3.2.1. Context of the study

The current study was conducted in five classes of Basic Translation in the academic
year 2014-2015. The course is among the intermediate courses in the curriculum of
EF, which aim at preparing students for the specialized courses later in the B.A.
program. Precisely, it aims to make students ready for translation and interpretation,

39
which is one of the three specializations they have to select from after finishing the
intermediate courses. With such an important role, the courses objectives are to
enhance students knowledge and at the same time equip them with skills and
techniques necessary for translation, which are quite new to and unfamiliar with them.

Given such objectives, the course is designed with activities involving translation from
Vietnamese to English and vice versa from English to Vietnamese. The translation
tasks are chosen to be simple enough for students to get acquainted with translation
and yet diverse in terms of topic and challenging enough to stimulate their acquisition
of new knowledge. Being able to expose themselves to many types of texts on different
topics can help students considerably broaden their vocabulary, which is crucial to the
work of a translator. Moreover, through these activities they are introduced to new
concepts in translation, one of which is unit of translation. This concept can draw their
attention to the fact that translation is not necessarily at word level as usually seen in
word-for-word translation but also involves phrase, sentence or even paragraph levels
depending on the type of translation unit being considered. As a result, learners will
be more careful in their translation and will be able to choose appropriate methods to
translate effectively and accurately.

With the aforementioned objectives and the fact that collocations play important roles
both in vocabulary acquisition and in translation, collocations teaching/learning
should be included in this course and considered as an important element in the
teaching program. However, it has been seen that teachers of translation tend to lay
more emphasis on practical translation techniques and strategies rather than elements
such as collocations due to their belief that the latter should be taught in grammar
courses. The teaching of collocations, therefore, may receive inadequate attention in
Basic Translation regardless of its importance.

3.2.2. Participants

The current study focused on learners collocational errors in translation. In order to


have a thorough view on the issue and to suggest appropriate remedies, relevant data
were collected not only from learners but also from teachers. That is to say, the study
investigated two groups of participants: one of students and one of teachers.

40
3.2.2.1. Student participants

The student participants in this study were the students of the five Basic Translation
classes mentioned above. They were all juniors at EF, USSH, who had spent 4
semesters polishing their skills through general English courses and other introductory
courses (e.g. Academic Writing, Introduction to Linguistics, etc.). Therefore, they
were supposed to be at a high level of English proficiency and familiar with
collocations, which frequently appear in the course books of these subjects. Moreover,
the students had also been attending the basic translation course, which equipped them
with necessary skills and knowledge of translation. To put it differently, they
completely met the requirements for participating in the study. Given the reason above
and the fact that the seniors, who were also potential candidates, were busy with their
internship and preparation for graduation, the juniors seemed to be the best choice
available.

Among these five classes, one was randomly chosen to carry out the pilot study. The
other four, which participated in the main study, consisted of 136 students in total.
However, as the collection of data was carried out in multiple phases, the numbers of
students participated in these phases were different due to uncontrollable reasons. To
be precise, there were 136 midterm exam papers in Basic Translation collected from
the 136 students because all of them had taken part in the exam. However, the number
of students taking the translation test was only 89 since some of them were absent on
the days the test was conducted. For the same reason, there were 104 students
responding to the questionnaire and the COLLEX test, which were carried out
simultaneously. Ideally, the number of participants should be consistent through all
the phases of data collection. However, due to the constraints of time and the tight
schedule of these classes, it was impossible to arrange a better time when all the
students would be present. Nevertheless, the researcher managed to attain data from
as many students as possible. Moreover, in terms of quantity, the number of
participants in each aforementioned phases seemed also acceptable.

41
3.2.2.2. Teacher participants

In addition to the students, the teachers of the Basic Translation classes also
participated in the study, or precisely in the interview. In total, there were four teachers
in charge of the five mentioned classes. One of them, however, could not take part in
the study due to personal reasons. The other three showed great interest in the study
and were willing to participate. All of them hold a Master degree in TESOL and had
much experience in both teaching and translation. Their gender was not important in
the present study, but it was worth mentioning that all of the three teacher participants
were female. Except for the information above, other personal information about the
teachers was considered private and unnecessary for the study and thus were not to be
presented. Besides, the teacher participants were tagged with the following codes: T1,
T2 and T3 to keep their answers anonymous as a way to protect their personal
identities and stimulate accurate responses.

Three teacher participants might be considered a small number and might not be
representing enough. However, since the current study focused on the Basic
Translation classes and at the time the study was carried out they were the only
teachers teaching the course, they were inevitably the only choice. Moreover, as they
had directly taught the classes, they were undoubtedly the ones having the most
thorough and accurate view on the students and the teaching situation. In other words,
they were not only the only but also the best candidates for the study.

3.2.3. COCA
As mentioned in the operational definition of collocation, the Corpus of
Contemporary American English (COCA) was used in the current study as a reference
source for categorizing the combinations produced by the students. There were
reasons for COCA but not any other corpora or reference sources to be chosen.

Firstly, in previous studies a collocation dictionary can be used instead of a corpus, or


both of them can be employed at the same time. However, since collocation
dictionaries are developed based on corpora and have much smaller database,
replacing the latter with the former or using both of them is inefficient and
unnecessary. For that reason, corpora remain the prime choice in this study.

42
Regarding the corpora of native speakers of English, there are several major ones in
addition to COCA, including the British National Corpus (BNC) and the American
National Corpus (ANC), etc. Compared to them, COCA has certain advantages.
Firstly, with a database of more than 520 million words, recently updated in 2015,
COCA is not only the largest but also the most up-to-date corpus at the moment. Since
the way people use languages changes over time, such a corpus will be more reliable
as it reflects accurately the current, ongoing changes in the language. Its larger
database with 20 million words added each year also means that the corpus is more
representative. Secondly, COCA is one of a few corpora providing easy access
through the Internet and a powerful search engine. Last but not least, according to
Dang (2014), American English is more popular among EF students due to its growing
popularity, especially in movies, reality shows, music, etc. COCA, as a corpus of
American English, is obviously more suitable for the analysis of the language used by
young people. For the reasons above, COCA is inevitably the most appropriate corpus
for the current study.

COCA is accessible at http://corpus.byu.edu/coca/. The user interface is quite simple


to understand and convenient to use. On top of that, the corpus provides a powerful
search engine with many functions necessary for a variety of research purposes. In the
present study, however, the corpus was used for two main purposes. The first one was
to design the research tools, which will be described later in this chapter. The second
one was to check if a combination is a collocation or not using two values provided
by the corpus: raw frequency (FREQ) and mutual information score (MI).

The FREQ of a combination is the times it occurs in the corpus. It shows how
frequently that combination is used by native speakers. In other words, the higher the
FREQ of a combination is, the more commonly it is used. The concept of MI,
however, is more complicated. According to Church & Hanks (1990), it compares
the probability of observing x and y together (the joint probability) with the
probabilities of observing x and y independently (chance) (p.23). Hodne (2009)
suggested that the higher the MI score is the stronger the bond between the elements
in a combination is. The MI score, therefore, can be simply understood as a value
indicating the strength of the bond in a combination. As previously discussed in

43
Chapter 2, for a combination to be considered as a collocation, it must be frequently
used by native speakers and has a strong bond between its components. The FREQ
and the MI score provided by COCA, therefore, were used as the criteria to identify
collocations. The way these criteria were applied in the present study has already been
discussed in the operational definition in Chapter 2.

3.2.4. Research instruments

In order to collect accurate and reliable data to answer the research questions, the
current study employed several research instruments including: (1) the students
midterm exam papers, (2) a translation test, (3) a COLLEX test, (4) a questionnaire
and (5) an interview. Each of them served a specific purpose. The students exam
papers and the translation test reflected the learners productive knowledge of
collocations and provided essential data for the analyses of errors. The COLLEX test,
on the other hand, was used to test their receptive knowledge. Together, these
instruments showed a more complete picture of the students collocational knowledge.
The questionnaire and the interview, additionally, aimed to investigated the teaching
and learning of collocations in Basic Translation from both teachers and learners
perspectives.

3.2.4.1. Students midterm exam papers in Basic Translation

As previously discussed, corpora are valuable and reliable sources of data in error
analysis thanks to the great number of authentic texts and the diversity of topics they
cover. Usually, researchers would make use of official corpora such as the
International Corpus of Learner English (ICLE). However, when there are no
appropriate corpora available, a collection of texts produced naturally and
spontaneously by learners can be used as a substitution because they can accurately
reflect learners competence and thus should be as authentic as ones in a corpus.

In the present study, since no ready-made corpus of Vietnamese learners of English


was available, the students midterm exam papers were chosen as the source of data.
In previous studies, researchers often make use of learners final exam papers, as seen
in Dang (2014) and Nguyen (2014). Nevertheless, at the moment the current study
was carried out, the students did not take the final exam yet. Hence, the researcher had

44
to opt for the midterm papers instead. Moreover, while the final exam would be
organized by the faculty with the same questions for all of the classes, the midterm
exam was decided by teachers and thus the questions were different from one class to
another. This turned out to be an advantage of using the midterm exam papers. Since
they were collected from different classes, the translation tasks in these papers
probably covered different topics and thus the data collected would be more diverse.

There were a total of 136 midterm exam papers collected. All of these papers consisted
of two main parts: one required the students to translate from English to Vietnamese,
the other from Vietnamese to English. Due to the scope of the study, only the
Vietnamese English translation tasks were taken into account. These tasks consisted
texts that focused on several different topics, including: health, society, environment,
literature, etc. The papers, therefore, were expected to provide a large amount of
authentic data for the current study.

3.2.4.2. Translation test

In addition to the exam papers, the translation test was also used to collect
collocational errors from the students. The format of the test was adapted from
Kurosaki (2012); its content, nonetheless, was developed by the researcher to match
the current research context.

Kurosaki (2012) introduced a type of translation test in which the respondents were
asked to translate word combinations from their L1 into English. The combinations to
be translated were put in specific contexts as an aid to their comprehension. The
author, however, argued that the aim of the test was to test learners collocational
knowledge, and thus unnecessary factors that hindered their performance should be
eliminated. For that reason, translations of the contexts were provided beforehand,
leaving only the target combinations for the respondents to translate. Moreover, this
also considerably reduced the amount of time needed to carry out the test, which is
also important in studies where time pressure is a problem. With the aforementioned
advantages, this test format was appropriate for the current study. However, in order
to give the student participants more freedom in their answers, they were allowed to

45
produce their own translations of the contexts if they were not satisfied with the given
one.

Because the study of Kurosaki (2012) focused on French and Japanese learners of
English, the content of the test could not be applied in the present study due to the
difference in participants L1. For that reason, the researcher had to make a new
translation test based on the original format. The development of the test consisted of
three main steps: (1) choosing the target collocations, (2) choosing the contexts and
(3) creating and modifying the test.

First of all, adjective-noun combinations were collected from two sources. The first
one was course books that the students had studied such as Interactions, Mosaic, etc.,
since the lexical items in these books were expected to be familiar with and
appropriate to them. The second source was other studies on collocations such as
Shehata (2008) and Kurosaki (2012), in which lists of common collocations extracted
from corpora were included. These collocations, according to the authors, occurred
frequently in daily use of language and thus were usually introduced to learners. Then,
these collected combinations were checked in COCA to see whether they matched the
criteria set by the operational definition (FREQ 5, MI 3). After that, the
combinations satisfying the aforementioned conditions were checked again using the
Oxford Text Checker, a text difficulty assessing tool based on the Oxford 3000, to
choose the combinations appropriate for the students level of English. According to
the publisher, a typical text at low intermediate level should have about 100% of the
words appearing in the Oxford 3000. As the student participants were 3rd year
English-majored students, their level should be above intermediate. Therefore, the
combinations which passed the test of the Oxford Text Checker could be considered
appropriate for the students level. The others, on the other hand, were excluded.
Finally, from these remaining collocations, the researcher carefully chose the most
suitable ones for the collection and analysis of errors basing on the criteria mentioned
in the conceptual framework (e.g. the selected combinations should potentially reflect
different causes of errors and include both congruent and incongruent collocations,
etc.). The result of the whole process was 42 target adjective-noun collocations, which
were later reduced to 20 after the pilot test.

46
In the second step, the context for each selected collocation was chosen. Again, COCA
was employed as the source. Firstly, the researcher looked up in the corpora with the
chosen collocations as the keywords. With each entry, all the sentences containing the
keyword were shown. For each keyword, one sentence or group of sentences were
chosen. The selected sentences must provide meaningful and complete contexts, in
which the target collocations could be accurately understood. At the same time, they
must also match the students level of English. Therefore, the chosen sentences were
also checked in the Oxford Text Checker. The sentences containing words outside the
Oxford 3000 were replaced or simplified in order to minimize unnecessary
difficulties for the students.

Subsequently, the selected contexts were translated into Vietnamese. The researcher
tried to make the translations as natural as possible, and at the same time to maintain
the equivalent structure of the target collocations, which means the adjective-noun
collocations in English should be translated into noun-adjective (or noun-noun in
some cases) collocations in Vietnamese. The reason for the sentences to be strictly
translated like that was to direct the students attention to the target collocations. The
sentences and its Vietnamese translations, finally, were paired, with the target
collocations being erased from the English ones. The students taking the test had to
fill in the blanks with the correct collocations, using the hints given by the Vietnamese
version. The following is a sample item of the translation test3:

Karlow mt mt chn trong chin tranh v phi i n Philadelphia c


lp mt ci chn gi mi.

Karlow had lost his leg in the war and had gone to Philadelphia to be fitted for a
new ____________.

The respondents had to translate only the target combination. However, they were
allowed to provide their own translations of the whole sentences if they wanted. From
42 chosen combinations, 42 corresponding test items were created initially. After the
pilot test, which will be presented later in this chapter, some of them were opted out.
The translation test used in the current study, subsequently, consisted of 20 items.

3
A sample of the translation test is included in Appendix 1

47
The whole process of choosing the test format, selecting the target collocations and
the relevant contexts, as well as creating and modifying the test described above aimed
to eliminate unnecessary factor hindering the students performance and to focus on
testing their collocational knowledge.

3.2.4.3. COLLEX test

The two instruments above served as the main tools to collect collocational errors from
the students and thus focused only on their productive knowledge. However, the
participants receptive knowledge should also be investigated for a more complete
picture of their collocational knowledge. For that reason, the COLLEX test was
included in the present study.

The format of the test was adapted from Gyllstad (2007) with some modifications. The
original version required respondents to make choices between only two options: one
correct collocation and one distractor. With a 50/50 chance for respondents to answer
correctly even when they do not know the correct option, the test seems not demanding
enough. Moreover, with such a format, learners only need to know the correct option;
the other one is obviously incorrect. However, it was argued that receptive knowledge
involves not only the ability to recognize correct lexical items but also the ability to
point out incorrect ones. For the reasons above, the test format was modified with a
third option both are correct added. This not only increased the difficulty of the test
but also made it possible to check learners ability to identify erroneous collocations.

Similar to the translation test, the COLLEX test consisted of 20 items. Each item had
three options, one of which was the target collocations used in the translation test. The
other one was either an incorrect combination or another correct collocation having
the same meaning with the first one. The incorrect combinations were created by
translating from their Vietnamese equivalents using word-for-word translation or
replacing one component of a correct collocation with its synonym, or taken from the
students answers to the translation test. The correct collocations for this option were
selected from COCA and also from the translation test. The last option was the newly
added one, namely both are correct, mentioned above.

48
With the aforementioned modifications, the COLLEX test could investigate learners
receptive knowledge of collocation more thoroughly. Moreover, by using some of the
students answers to the translation test as options, the researcher expected the
COLLEX test would yield valuable information supporting the error analysis.

3.2.4.4. Interview

The interview was carried out for a thorough view on the actual teaching and learning
of collocations in Basic Translation from the teachers perspective. As the ones
directly teaching these classes, they understood their students and the current situation
more clearly and thoroughly than anyone else. Their opinion, therefore, was expected
to yield accurate and valuable data supporting the error analysis and remediation. The
interview was carefully prepared with eight questions focusing on three main aspects:
(1) the teachers perception of collocation, (2) their teaching practices and (3) their
observation of the students performance.

The teachers perception was taken into account because it was an important factor
influencing their teaching practices. In the present study, the teachers perception of
collocations was investigated through two questions. The first one aimed to explore
their knowledge of collocations. A teacher with profound knowledge about
collocations was likely to have clear perception of the matter. The second question
asked about their opinion on the role of collocations in translation and translation
teaching. If the teachers thought they were useful, they would probably include them
in the teaching program. On the contrary, they would neglect them if they considered
them unimportant.

The teaching of collocations was the next factor to be investigated. Based on the
teachers answers to the filtering question of whether they taught collocations in Basic
Translation, more questions would be asked to elicit further information. If their
answer was yes, the next questions would be about their method of teaching
collocations and corresponding problems they had to face. Conversely, if they denied
teaching collocations in Basic Translation, the questions would be ones that sought for
the reason and further elaboration.

49
Finally, the teachers observation of learners performance on collocations were
explored through five questions, each of which focused on a specific aspect. The first
one aimed to find out if the students had any problems in translating collocations and
what the problems were. The second one focused on the benefits which the teachers
instruction on collocations could bring to the students. Obviously, this question was
relevant only if the teacher affirmed their teaching of collocations in class. In
accordance with the first question, the third one asked about the strategies the students
used to deal with difficulties in translating collocations. The last question asked for
the teachers opinion about the causes of collocational errors in learners translations.
Despite focusing on different aspects, all of the questions above served one chief
purpose: to provide the researcher with an insight, from the teachers perspective, into
the learning of collocations in Basic Translation.

3.2.4.5. Questionnaire

The questionnaire was included in the present study to collect information that cannot
be achieved through either the students exam papers or the tests. Similar to the
interview, the questionnaire also focused on three major aspects: (1) the students
perception of collocations, (2) their acquisition (learning) of collocations and (3) the
teaching of collocations from their perspective. Their perception was taken into
account because, similar to the teachers, it could greatly influence their acquisition of
collocations. Their opinion on the learning and teaching of collocations was collected
for comparison with that of the teachers and also for more insight into the issue.

The items for the questionnaire were adapted from Dang (2014) and Mai (2010). The
questionnaire in Mai (2010) was designed to investigate learners use of collocations
in V-E translation, which was in line with the present study. It consisted of 12 items
which mainly focused on learners perception, acquisition and use of collocations.
Nevertheless, it failed to take into account some important factors such as the teaching
and learning of collocations, which were investigated carefully in Dang (2014).
However, the questionnaire in this study was designed for learners of Academic
Writing. Therefore, the two questionnaires were combined and adapted to match the
purposes of the present study. The questionnaire of the present study, thus, consisted

50
of 11 items, divided into three parts, namely perception of collocations, acquisition
and use of collocations and teaching of collocations.

The first part was comprised of four items adapted from Mai (2010) which asked about
the students perception of collocations. The first and the second questions (Q1, Q2)
investigated their knowledge of the term collocation and the sources of knowledge.
Among them, Q1 functioned as a filter. The student answering No to this question
would be asked to finish the questionnaire there, because without the necessary
knowledge their answers to the other questions would be unreliable. The third and the
fourth questions (Q3, Q4) focused on their opinion about the role of collocations in
translation. Again, Q3 acted as a filter. The students who answered No to Q3 would
be asked to skip Q4 since it was pointless for those who failed to recognize the
importance of collocations in translation to elaborate their ideas on the issue.

The next part consisted of five items which were also adapted from Mai (ibid.). Each
of them focused on one aspect of the students acquisition and use of collocations.
Question 5 (Q5) aimed to find out the methods they frequently used to broaden their
collocational knowledge. The next three questions (Q6, Q7 and Q8) respectively asked
how frequently the students cared about choosing correct collocations in translation,
faced difficulties in doing so and paid attention to their collocational errors. The last
question of this part (Q9) asked about the strategies frequently used by the students to
deal with collocation-related difficulties in translation.

The third part consisted of two items investigating the teaching of collocations in Basic
Translation. The first one (Q10), again, served as a filter, asking the students whether
collocations were taught by the teachers. Those who affirmed the teachers instruction
of collocations in class would be asked to elaborate on the specific teaching strategies
listed in Q11, which was adapted from Dang (2014) as previously mentioned in
Chapter 2.

The questionnaire was expected to provide valuable data related to the teaching and
learning of collocation in Basic Translation from the students perspective which were
unattainable through other instruments. Since the collected data could possibly yield
important information supporting the analysis and remediation of errors, it was

51
considered an important research instrument in the current study. A summary of the
items and their functions were given in Table 3.1 below.

Table 3.1: Description of the questionnaire items


Part Question Content
Q1 Ss knowledge Ss knowledge of the terms collocation
Q2 of collocation Sources of Ss knowledge of the terms
Ss perception Ss opinion on Whether knowledge about collocations
1 Q3 of collocations the role of is important in translation
collocations in How knowledge about collocations
Q4
translation actually helps Ss in their translation
Acquisition of How Ss acquire knowledge about
Q5
collocations collocation
Ss acquisition Ss awareness of choosing correct
Q6
and use of combinations in translation
2
Q7 collocations in Use of How often Ss use wrong combinations
Q8 translation collocations Ss awareness of their own errors
Ss strategies to deal with difficulties in
Q9
translating collocations
Whether collocations are taught in Basic Translation
Q10
Teaching of classes
3
collocations Practical strategies of the teaching of collocations in Basic
Q11
Translation classes

3.2.4.6. Summary of the research instruments

The present study employed a variety of research instruments for a multi-perspective


insight into the problem. Each of them served a specific purpose and had its own
limitations. Nevertheless, when combined, not only did they compensated for each
others weaknesses, but also supported one another to yield reliable data for the study.
A summary of the instruments and their purposes were given in Table 3.2 below.

Table 3.2: Summary of the research instruments


Number of
Instruments Purpose of use
participants
Ss midterm exam papers in Basic 136 students
Translation Data for error analysis
Translation test 89 students
COLLEX test 104 students Ss receptive knowledge
Questionnaire 104 students Information on the teaching/learning of
collocations from Ss perspective
Interview 3 teachers Information on the teaching/learning of
collocations from Ts perspective

52
3.2.5. Pilot study

The pilot study was carried out in order to make sure that the research instruments
would work correctly in the main study. The pilot study involved the translation test
and the questionnaire to the students. The COLLEX test was created later as it needed
some information from the pilot study. Furthermore, its design was not as complicated
as that of the translation. For those reasons, the COLLEX test was not included in the
pilot study.

As mentioned above, the translation test initially consisted of 42 items. The test papers
were delivered to 19 of the students of the Basic Translation class chosen for carrying
out the pilot study. The students were expected to complete the test in 30 minutes.
However, most of them needed about 40 minutes to finish their task. The test papers
later were scored based on the scoring criteria which will be presented later in this
chapter. After that Cronbachs Alpha was calculated to ensure the reliability of the
translation test.

Table 3.3: Reliability statistics of the translation test


Reliability Statistics
Cronbach's Alpha N of Items
.863 42

The Cronbachs Alpha for the translation test was 0.863, higher than the acceptable
coefficient 0.7 in most social science research, which confirmed the internal
consistency of the tests items and its reliability. However, the test with 42 items still
seemed too long for the main study. As seen in the pilot study, most of the participants
needed about 40 minutes to finish the test while the time limit for the collection of data
in class could hardly exceed 30 minutes, not to mention the time needed for other
procedures such as giving instruction, distributing and collecting the test papers, etc.
For the reasons above, the researcher decided to reduce the number of items in the test
to 20 by opting out the items that appeared to be too easy for the participants. As a
result, the translation test used in the main study consisted of only 20 items and was
estimated to be completed in 20 minutes.

53
The questionnaire was given to the rest of the class (20 students). The reason for the
translation test and the questionnaire being distributed to two different groups was to
ensure that the pilot study would not occupy too much time of the class although the
teacher in charge was very supportive and willing to allow more time for it. For the
questionnaire included both multiple choice items and open-ended questions, the
Cronbachs Alpha for it could not be calculated. The pilot study of the questionnaire,
therefore, was to ensure the students clear understanding of the questions and to
identify possibly ambiguous or misleading details. After the pilot study, the items
causing misunderstanding or difficulties for the students were modified or reworded.
The number of items in the questionnaire, however, remained unchanged because all
of them proved to be necessary and the average amount of time needed to finish
answering the questionnaire was only about 10 minutes.

The pilot study not only double-checked the research instruments before the main
study was carried out but also provided essential information for creating the
COLLEX test. Precisely, some of the most frequent incorrect responses to the
translation test were chosen as distractors in the COLLEX. As mentioned above, with
these incorrect combinations included, the COLLEX test was expected to generate
more relevant information for the analysis of errors. It could suggest, for example, an
error was made due to learners lack of knowledge or another reason. The pilot study,
therefore, was of great importance in the present study.

3.3. Data collection procedure

Since the present study employed multiple research instruments, the procedures of
data collection were complicated. This section, therefore, provided a detailed
description of how data were collected with each research instrument.

3.3.1. Collection of adjective-noun collocational errors

As previously mentioned, the adjective-noun collocational errors used in this study


were collected from two sources: (1) the students midterm exam papers in Basic
Translation and (2) their responses to the translation test. The procedures for collecting
data from each source were described below.

54
3.3.1.1. Collection of errors from students midterm exam papers

Firstly, adjective-noun combinations in the students English translations were


carefully identified and extracted from the 136 midterm exam papers in Basic
Translation. The researcher, after that, looked in the original texts for the Vietnamese
equivalents of these combinations. The equivalents came in different forms including
noun-adjective and noun-noun combinations or even long, complicated phrases, which
explained why the identification of the adjective-noun combinations and their
Vietnamese equivalents had to follow the aforementioned order and not the opposite.

In the second step, the identification of collocational errors from the extracted
combinations was carried out as follow. First of all, the node (headword) and the
collocate of each combination were identified. As discussed in Chapter 2, the present
study considered the noun in an adjective-noun combination as the node and the
adjective as the collocate. After these elements were determined, COCA was
employed as the tool to check their collocability. Based on the criteria and the
possibilities set by the operational definition, the FREQ and the MI of a combination
in COCA could tell if it was a collocation (FREQ 5, MI 3), a free combination
(FREQ 5, MI < 3) or an erroneous/uncommon one (FREQ < 5). A detailed
illustration of how COCA was used is given below.

Figure 3.1: The COCA Search Interface

55
In Figure 3.1, the combination common disease was used as an example. The noun
disease was the node and therefore was put in the box WORD(S). The square brackets
told the program to include both the singular and the plural forms of the noun in the
result. The adjective common was the collocate and thus was put in the corresponding
box. The box POST LIST below it determined the part of speech of the collocate. The
two boxes (with numbers) next to the box COLLOCATES indicated the number of
words before (left) and after (right) the headword to be taken into account in each
search entry, which was also understood as the span in which the collocate might be
found. Since the present study focused only on attributive adjective-noun
combinations, the maximum span being considered was five words left to the node.
Each combination was checked with the span starting at 1. If the combination could
not be found in COCA or had the FREQ below 5 within that span, the span would be
increased until the combination appeared and had the FREQ higher than 5. However,
if the result was still unsatisfactory at the span of 5, the combination would be
considered as non-existent or uncommon because it was mostly impossible to find
adjective-noun combinations beyond this span, not to mention that the MI would
decrease as the span increased. The two boxes at the bottom were the criteria for
sorting and limiting the search results. In the present study, the options were set as
above so that both the MI and the FREQ would be shown in the search results as can
be seen in Figure 3.2 below.

Figure 3.2: The COCA Result Interface

In addition to COCA, the Vietnamese equivalents of the combinations were also taken
into account as a criterion to identify errors since it was pointless to produce a correct
collocation if the combination could not successfully convey the message of its
supposed equivalent in the source text. In other word, a combination failed to transfer
the intended message would be considered an error even if it appeared as a perfectly
correct collocation in COCA. The evaluation of the translations was done by the
researcher with the help from the teachers of Basic Translation.

56
To sum up, among the extracted adjective-noun combinations, the ones successfully
conveyed the original message and satisfied the criteria set by the operational
definition would be considered as collocations or free combinations (depending on
which criteria were satisfied). On the contrary, the combinations that did not
accurately transfer the message and/or did not meet the criteria would be recognized
as errors.

3.3.1.2. Collection of errors from the translation test

The translation test was another instrument to collect collocational errors from the
students. After the pilot study, the test was modified to match the time limit, which
resulted in the number of items being reduced to 20. The refined test later was
distributed to the student respondents, except for those who had attended the pilot
study. They were asked to provide their name in the test for research purposes, with
the assurance that their personal information would be held in strict confidentiality and
their performance on the test would by no means affect their academic results.

Since the respondents included students from four different classes, the test had to be
carried out at different times. Nevertheless, the researcher tried to maintain the
consistency of the data collection procedure by giving all of them the same instructions
and the same amount of time (25 minutes) to finish the test. However, the number of
students going to class on the days the test was carried out was still uncontrollable,
which resulted in the inconsistency in number of participants responding to each
research instrument, mentioned earlier in this chapter. Consequently, out of the total
136 students of the four classes, only 89 of them responded to the translation test. Due
to the tight schedule of these classes, it was impossible to give the test again to those
who had missed it. Nevertheless, 89 participants still seemed sufficient for the present
study.

After the test had been given to the participants, the same procedure used to identify
collocational errors in the exam papers was applied to identify errors from the test.
Similarly, the combinations produced by the students were extracted and paired with
their Vietnamese equivalents. This process, however, was simpler than it had been
with the exam papers thanks to the design of the test. Moreover, with such a design,

57
the researcher could easily identify the cases where the respondents left an item
unfinished, which allowed the investigation into a new category of error, namely
abandonment of tasks, which could hardly be approached through the data from the
exam papers. After that, COCA and the aforementioned criteria were again employed
to distinguish erroneous combinations from correct collocations and free combination.

3.3.2. Collection of additional data

As discussed above, in addition to collocational errors, the present study also collected
data related to the current teaching and learning of collocations in Basic Translation
and learners receptive knowledge of collocations in order to support the analysis and
remediation of errors. These data were collected through different instruments
including the COLLEX test, the questionnaire and the interview. The procedure of
data collection with each instrument was presented below.

3.3.2.1. Questionnaire

The questionnaire was to collect data on the current teaching/learning of collocations


from the students perspective. Similar to the translation test, the questionnaire was
also given to the student respondents who did not take part in the pilot study. However,
due to the fact that the researcher had only about 30 minutes for each time he met the
classes, the questionnaire could not be distributed at the same time with the translation
test. Instead, another day was chosen for each class, on which the questionnaire, along
with the COLLEX test was given to them. On the chosen days, the researcher himself
went to the classes at break time and asked the teachers in charge for 30 minutes to
work with the classes. The participants, then, were allotted 10 minutes to fill in the
questionnaire. Similar to the translation test, there were some students absent and thus
did not responded to the questionnaire. Consequently, the total number of students
answering the questionnaire was 104.

3.3.2.2. COLLEX test

The COLLEX test focused on the students receptive knowledge of collocations an


important part of their collocational knowledge which could not be investigated
through the error analysis. Similar to the other instrument, this test was also given to
the student who did not attended the pilot study. Since the COLLEX test was simpler

58
and less time-consuming than the translation test, it was distributed to the participants
on the same day with the questionnaire. With this test, the students were given 15
minutes to complete. Also, they were asked to put their name in the test paper so that
the researcher could later compare the result of this test with that of the translation.
Since the test was distributed on the same day with the questionnaire, there were also
104 test papers collected in total.

3.3.2.3. Interviews to the teachers

Unlike the other instruments, the interview was designed to collect data from the
teachers on the teaching and learning of collocations in Basic Translation. There were
three teachers willingly taking part in the interview, who were also the teachers in
charge of the Basic Translation classes being studied. As they had their own schedules
and in order to keep their information confidential, the interviews were carried out
individually at the most convenient time for each of them.

At the beginning, the purposes of the interview were clearly explained to the teachers.
They were assured that the information they were going to provide would be used for
the sake of research only, and their identity and personal information would be kept
strictly confidential. During the interviews, extensive notes were taken carefully. The
interviews were also audiotaped, with the teachers permission, in case any piece of
information was missed during note-taking.

3.4. Data analysis procedure

3.4.1. Analysis of adjective-noun collocational errors

After the collocational errors from the Basic Translation midterm exam papers and the
translation test had been collected, they were carefully analyzed. The procedures for
analyzing the errors from the two sources were mostly the same except for the
difference in the number of categories. The main tools for calculation used in the
analysis were the programs Microsoft Excel 2013 and SPSS 20.

Firstly, the numbers and percentages of correct collocations, free combinations and
errors collected were calculated. Then, the errors were put into the three categories
mentioned in the conceptual framework (i.e. wrong choice of noun, wrong choice of

59
adjective and wrong combination). However, for the errors from the translation test, a
category called no/incomplete answer was added to deal with the cases when the
students left certain items unfinished, which, according to Hussein (2011), was also a
type of errors. The number and percentage of errors in each category, after that, was
computed. In the next step, the possible causes of the errors were identified based on
the list of hypothetical causes of errors presented in Chapter 2. However, the category
abandonment of tasks was not taken into account in the analysis of errors from the
exam papers, since it was difficult to extract such errors from this source. Besides,
during this process a new category, namely others, was added to categorize the errors
that did not originate from the hypothetical causes. Their frequencies and percentages
were also calculated. Finally, based on the errors and the causes, corresponding
corrections were suggested.

3.4.2. Questionnaire

Prior to the analysis, the questionnaires were checked carefully so that those with
missing information would be excluded. The number of valid questionnaire,
fortunately, remained unchanged. After that, the questionnaires were coded and
inputted into the SPSS 20. During this process, the data were grouped into three main
categories regarding the participants perception of collocations, their acquisition and
use of collocations, and their opinion on the teaching of collocations in Basic
Translation. Simple descriptive statistics of the data including frequency and
percentage of the participants responses, then, were calculated. These data from the
students perspective, finally, were compared with those obtained from the teachers
and also were used to support the error analysis.

3.4.3. COLLEX test

The students performance on the COLLEX test would be compared with theirs on the
translation test for a more complete picture of their collocational knowledge. For that
reason, only the test papers from the students who attended both the COLLEX test and
the translation test were taken into account. The chosen COLLEX test papers, then,
were scored with each correct answer counted as 1 point, resulting in the maximum
score being 20. The chosen translation test papers were also scored again. However,

60
this time, free combinations were excluded, leaving only correct collocations as
acceptable answers, since the researcher wanted to compare the students ability to
recognize collocations with their ability to produce collocations. A paired sample t-
test with the test scores as variables, then, was computed using SPSS 20 for a general
comparison between the participants receptive knowledge and their productive
knowledge. After that, their performance on each item in the COLLEX test was
compared with its counterpart in the translation for a more detailed analysis of the
students receptive and productive competence in each case, which would probably
yield more data on their knowledge of collocations.

3.4.4. Interview

The notes of the interviews were carefully checked with the audio recordings to ensure
that no information was missing. After that, from the notes and the recordings, data on
the teachers perception of collocations and their opinions on the teaching and learning
of collocations in Basic Translation were extracted and put in corresponding
categories. The data, then, were analyzed qualitatively. Particularly, noticeable
opinions and comments from the teachers were highlighted and grouped by relevant
topics. The data, finally, were compared with those from the students and were also
used in the analysis and remediation of errors.

3.5. Summary

The chapter provided a complete description of the methodology used in the present
study. Based on the conceptual framework, five research instruments, each of which
served specific purposes, were designed and presented. In relation to these
instruments, the procedure of data collection and the way in which the data were
analyzed were also clearly described.

61
CHAPTER 4
FINDINGS AND DISCUSSION

This chapter presents and discusses the results of the study based on the analysis of
the data collected from the research instruments. The data were carefully analyzed in
order to find the answer to the research questions mentioned in the previous chapters.

4.1. Data analysis and discussion

In order to have a comprehensive insight into the participants making of collocational


errors, a great amount of data was collected. The participants, as mentioned in Chapter
3, consisted of four classes of 3rd year English majored students studying Basic
Translation and the three instructors teaching the subject. The data were then analyzed
following the conceptual framework presented in Chapter 2. As the main aim of the
study, the analyses of collocational errors are presented first. The analyses of data from
the other sources, including the multiple choice test, the questionnaire and the
interview, are discussed later.

4.1.1. Analysis of collocational errors from students exam papers

As mentioned in the previous chapters, the collocational errors to be analyzed were


collected from two primary sources, i.e. the students midterm exam papers in Basic
Translation and the translation test designed by the researcher. Both sources were
important to the study. While the former, as a corpus-based source, provided authentic
information about the participants use of collocations, the latter allowed the research
to investigate the issue more thoroughly from another perspective, with the kinds of
errors that may not be seen in the exam papers. The collected errors were, then,
analyzed based on the framework of error analysis adopted from Gass & Selinker
(2008). In this section, the results of the error analysis, including the types and the
hypothetical causes of errors, will be presented for each source of data respectively,
starting with the errors from the students midterm-exam papers.

From 136 midterm exam papers in Basic Translation, a totality of 1427 adjective-noun
combinations was identified. Based on the criteria set by the operational definition
presented in Chapter 2, these combinations were then classified into three categories,

62
namely correct collocations, free combinations and erroneous combinations, as
shown in Table 4.1 below.

Table 4.1: Lexical combinations from the midterm-exam papers


Acceptable Combinations Errors
Total
Correct Collocations Free Combinations Erroneous Combinations
Count 731 285 411 1427
Percent 51.2% 20.0% 28.8% 100%

One noticeable point from the result is the fact that the percentage of free combinations
is much lower than the number observed in several recent studies, especially in Dang
(2014) a study involved the analysis of collocational errors, which also used the same
criteria to categories lexical combinations. In the aforementioned study, there were
5403 free combinations out of 6285 identified combinations, accounting for 85.97%
of the totality. The explanation given to such a high number was that free
combinations were very popular and used very frequently and that they could appear
several times in an essay with every of their single incidence was recorded and
counted (ibid., p. 83). However, while such an explanation was reasonable and
convincing, it does not mean that the low percentage of free combinations in the
present study is inaccurate. In fact, there is a factor which may explain the difference:
the targets of the two studies. In Dang (ibid.), the author investigated multiple types
of lexical combinations, including verb-noun, adjective-noun, adverb-adjective, etc.
Among them, there might be certain types of combination, such as verb-noun, that
allowed freer combinations of words and thus led to the high frequency of free
combinations observed. In contrast, the present study only focuses on one type of
combination, i.e. adjective-noun. The low proportion of free combination, therefore,
may indicate that the combination between adjectives and nouns is more restricted,
which limits the availability of choices in combining adjectives and nouns, making
accurate production of adjective-noun collocations more challenging.

The fact that more than a half of all the combinations (51.2%) are correct collocation
seems to be encouraging. It signifies that the students had knowledge about collocation
and made use of it in their translations. However, it is an ineligible fact that they did
also produce erroneous combinations in the process. Thus, no conclusion or

63
assumption should be carelessly made before the significance of the erroneous
combinations is thoroughly investigated.

Regarding the number of erroneous combinations produced by the students, there are
411 of such combinations, accounting for 28.8% of all the combinations acquired.
Although the percentage of erroneous combinations is lower than that of correct
collocations (50.1%), this number still has its own significance due to some reasons.
Firstly, 28.8% is, in fact, a considerable number compared to other similar studies.
The proportion of incorrect combinations in Dang (2014), for example, was only
6.72%. In Nguyen (2014) it was 31.9%, which is just slightly higher than 28.8%.
Secondly, it is necessary to note that many of the combinations in the midterm-exams,
according to the teachers, were not really unfamiliar to the students, as they had been
introduced to them during the course. Moreover, the students also had access to
monolingual dictionaries during the exam, which is also an advantage for them. With
the given conditions, it is expected that the students performed better in their test. To
put it differently, the percentage of erroneous combinations (28.8%), though may not
be a discouraging number, is significant enough to draw attention towards the
students knowledge about collocation.

To sum up, the figures indicate that although the students possessed knowledge about
collocation and used it in their translations with relatively high degree of accuracy, a
considerable number of incorrect combinations were inevitably made, which might be
partly attributed to the possibly high restriction in adjective-noun combinations,
shown by the low percentage of free combinations. That factor, however, is not enough
to explain all the difficulties the participants had to face in using collocation correctly.
For a deeper insight into the problem, the types and the causes of errors need to be
analyzed.

4.1.1.1. Types of errors

The erroneous combinations were classified into three categories, namely wrong
choice of adjective, wrong choice of noun and wrong combination, which reflect the
nature of the errors. The first one, as the name suggested, refers to combinations which
are made incorrect because of the adjectives and can be improved by replacing them

64
with other adjectives. Regarding the second category, it is the nouns that need
replacing. The last one, however, is more complicated than the other two, since
combinations of this type cannot be improved by simply changing either of their
components. The purpose of this categorization is to find out which tends to be more
problematic for students: choosing the collocate (adjective), choosing the node (noun)
or producing the combination itself. For that purpose, the incorrect combinations were
categorized and presented in Table 4.2 below.

Table 4.2: Types of collocational errors from the midterm-exam papers


Types of Error Count Percent Examples
underground (coral) reef, destroying
Wrong Choice of Adjective 213 51.8%
(fishing) method, fatal rate
small piece (of money), modern
Wrong Choice of Noun 62 15.1%
chemical
environment protecting, environmental
Wrong Combination 136 33.1%
living, medical pharma
Total 411 100%

The figures show that wrong choice of adjective is the dominant category, with 213
instances accounting for 51.8% of the collected errors. Some typical examples of this
type are combinations such as underground (coral) reef (di ngm san h),
destroying (fishing) method, fatal rate (t l t vong) which should be respectively
translated as underwater (coral) reef, destructive (fishing) method and death rate. At
the second position is the category wrong combination, which accounts for 33.1% of
the totality, with incorrect combinations like environment protecting (vic bo v mi
trng environmental protection), environmental living (mi trng sng living
environment), medical pharma (cng ty dc pharmaceutical company), etc. Wrong
choice of noun, as a result, is the category with the smallest number of incidences. In
total, there are only 62 combinations (15.1%) in which the nouns were incorrectly
chosen. This category includes the instances such as small piece (of money) (mt mn
tin nh small amount of money), modern chemical (dc phm hin i modern
medicine) etc.

The statistics above, in other words, signify that choosing correct adjectives for
adjective-noun combinations seems to be the most common challenge for the
participants. Nevertheless, it does not mean that choosing nouns is easy for them,

65
because when the two categories wrong choice of noun and wrong combination being
added together, the total number of combinations having the nouns incorrectly chosen
were 198, accounting for nearly half of all the errors. Besides, as previously discussed,
in an adjective-noun combination, it is the noun that determines the adjective, and thus
incorrect choice of the former usually leads to incorrect choice of the latter, which
explains why the number of errors with both adjectives and nouns incorrectly used is
even higher than that of errors with incorrect nouns alone. To put it differently, it is
possible to conclude that choosing nouns for adjective-noun combinations is also quite
a problem for the participants. The categorization of error types has shed some light
on the issue. In the next section, the causes of errors are identified for deeper insights.

4.1.1.2. Causes of errors

One essential step in error analysis is identifying the causes of errors, by which
solutions can be found to avoid making similar errors in the future. However, it is not
an easy task because the making of errors is a complicated mental process and in most
circumstances, there is little evidence strong enough for researchers to confirm with
100 percent of certainty that an error is caused by a certain factor. For that reason, this
process of identifying the causes of errors needs to be done carefully to reduce the
probability of making false assumptions. In the present study, in order to achieve a
result with as high degree of accuracy as possible, a comprehensive list of hypothetical
causes of collocational errors was employed. Based on this list, the underlying causes
of all 411 errors were painstakingly analyzed and presented in Table 4.3 below.

Table 4.3: Causes of collocational errors from the midterm-exam papers4


Causes of Errors Count Percent Examples
Approximation 120 25.4% fish-catching/hunting methods, ozone damaging
Use of Synonyms 73 15.4% cardiovascular troubles, submarine (coral) reef
L1 Transfer 128 27.1% attractive (street) food, underground (coral) reef
Word Coinage 19 4.0% uneffective medicine, ozone destroyment
Formal Confusion 112 23.7% Europe countries, tasteful food
Others 21 4.4% coral field/ plain, diligent web, coral oust ranges
Total 473 100.0%

4
Refer to Appendix 6 for more examples of students' collocational errors in context

66
Along with the five hypothetical causes in the list, a new category, called others, was
added to address the cases that did not belong to the other five or needed special
explanation. Moreover, it is important to note that some errors might have more than
one cause; the total number of causes, therefore, is higher than that of the errors as
shown in Table 4.3 above.

Approximation
According to Hong et al. (2011), approximation is the act of using an incorrect lexical
item in place of the expected one to express an intended meaning due to some
similarity in the semantic properties of the two items. Approximation, as previously
mentioned, is a form of paraphrasing, which usually happens when learners do not
know or cannot find the necessary word and opt to use another lexical item instead.
This strategy, as observed in other studies, is a common cause of error. Hong et al.
(2011), for example, have found that 21.19% of the errors collected in their study had
been caused by approximation. Similarly, Dang (2014) revealed that about 27.18% of
the collected errors resulting from the use of this strategy.

In the present study, the result seems to conform with that of the aforementioned ones.
With 120 observed cases, making up 24.5% of the totality, approximation is the second
most common cause of errors. It seems that the students, unable to produce the desired
combination, usually resorted to this strategy to produce combinations using the words
which they thought could convey the intended messages. For example, when
translating the phrase cc phng thc nh c, some students failed to produe the
collocation fishing methods, and thus opted to use some incorrect/uncommon
combinations such as fish-catching methods, fish-hunting methods or fishing
measures. There were some similarities in terms of semantic properties between fish-
catching/fish-hunting and fishing or between measure and method, which might
explain the students making of errors. Similarly, ozone damaging, an incorrect
translation of vic ph hy tng ozone, was also a result of approximation when they
were unable to produce correct expression such as ozone destruction/ depletion.

Use of synonyms
It is commonly known that distinguishing between synonyms is a big challenge to
learners of English. Moreover, many of them usually resort to bilingual dictionaries,

67
which often provide the same translation/explanation for synonyms. As a result, the
learners tend to make a false assumption that words which are synonyms of each other
have the same meaning and thus can be used interchangeably. Such an assumption,
along with the lack of knowledge about restriction rules of collocations, usually leads
to the learners making of collocational errors by using a synonym in place of the
desired word.

The statistics show that the use of synonyms is also a common cause of error, with 73
cases, making up 15.4 % of the total number. Some typical examples for this cause of
errors are combinations such as popular diseases (nhng bnh ph bin),
cardiovascular troubles (vn v tim mch) and submarine (coral) reef (di ngm
san h) which should be translated as common diseases, cardiovascular problems, and
underwater (coral) reef respectively. It is possible that the synonymous meanings
between popular and common, trouble and problem, submarine and underwater are
the reason making the learners choose the former ones to substitute the latter, without
knowing that the uses of them are inappropriate in these combinations. The result
signifies that there was possibly a lack of knowledge about restriction rules of
collocations among the participants, and they also, as Kim (2009) suggested, had
difficulties in choosing the correct word among the synonymous ones in the target
language.

L1 transfer
The influence of L1 on learners use of L2 is commonly seen, especially when they
have to deal with collocations and translation. It has been confirmed by several authors
including Kim (2009), Mai (2010), and Kurosaki (2012) that learners use of English
collocations might be heavily influenced by their L1. Therefore, in the present study
L1 transfer was expected to be a major cause of collocational errors. It seems that the
statistical evidence also supports the aforementioned assumption. There were 128
errors resulting from L1 transfer, accounting for 27.1% of all the observed cases,
making it the primary cause of errors. The influence of L1 on the participants making
of collocational errors usually took place in two circumstances.

The first one is when two or more words in English share the same equivalent in
learners L1. When this happens, learners may possibly use the one they know in

68
English for every single incidence of its assumed L1 equivalent appearing in the
original text, unaware that the two words are not always the correct equivalent of each
other. This seems similar to the case of synonyms and thus making it confusing to
distinguish between them. However, while synonymous words can be used
interchangeably in certain cases, the words sharing the same L1 equivalent described
here can hardly replace each other in most situations. This phenomenon accounted for
a number of errors caused by L1 transfer, with some typical examples such as
attractive (street) food, underground (coral) reef, nutrition system which are incorrect
substitutions of delicious (street) food (mn n ng ph hp dn), underwater coral
reef (bi ngm san h) and nutritious diet (ch dinh dng) respectively.

The second circumstance for L1 transfer to take place is when the learners resorted to
literal word-for-word translation, which usually ended up in an erroneous translation
violating the rules of collocation or betraying the implying message of the source text.
Upon investigating the collected errors, a relative large number of incorrect
combinations caused by literal word-for-word translation were identified. For
example, the phrase bi ton kh was figuratively used in one of the tests, which
actually meant a big problem or a challenge. However, some of the students, unable
to realize the figurative meaning of the phrase, literally translated it as mathematic
problems or a tricky math. Another example of errors caused by word-for-word
translation is the case of bnh ic (deafness or hearing loss) which was translated by
some students as deaf disease. While Vietnamese people tend to use the word bnh
(disease) for most abnormalities of the body, the native speakers of English consider
deafness as a handicap rather than a disease. This difference in cultural perspective
may be the cause of the aforementioned translation. Similarly, the phrases mu bnh
m con con (a tiny piece of bread) and mt mn tin nh (a small amount of money)
were translated as a small bread and a small money respectively, which also violated
the rules of word-combination.

It seems that collocations and translation all together created big challenges for the
students. Being unable to produce a desired combination, they resorted to their
knowledge in L1 and literal word-for-word translation, which led to the large number
of errors observed. The result is in line with that of Dang (2014), which also focused

69
on Vietnamese learners of English. In her study, the author also pointed out that L1
transfer was the primary cause of collocational errors. However, some other authors,
who have conducted studies on participants with different L1 other than Vietnamese,
have come to quite different results. Farghal & Obiedat (1995) and Hong et al. (2011),
for example, have pointed out that L1 transfer only accounted for 12.6% and 11.92%
of the errors made by their participants respectively, although those figures still made
L1 transfer a common cause of errors in their studies. One possible explanation for
this is the difference in the participants native languages. Besides, the theoretical
framework chosen by the authors might also have affected the results. Regardless of
the differences in statistics, all of the authors have agreed that L1 transfer is one of the
major causes of collocational errors, which supports the result of the present study.

Word coinage
This hypothetical cause of errors refers to the act of creating new vocabulary items,
which usually results in malformed words or spelling mistakes. This cause of errors
has been neglected by several authors due to its low frequency of occurrence.
However, there were still several incidences of word coinage observed in the study.
Some of the errors, such as radioactive emition, nutrious regime, pharmatical
products, etc. might result from the learners uncertainty of the desired words (i.e.
emission, nutritious and pharmaceutical respectively). It is possible that the students
had seen those words before, but due to certain reasons they failed to use them
correctly and thus ended up with those spelling mistakes. In some other cases, the
errors were made by the students misuse of affixes. The students might know the
roots of the desired words but were uncertain about the words themselves. As a result,
they tried to form the words using the affixes they know, which led to some malformed
words such as uneffective medicine, ozone destroyment, etc. In total, there were 19
errors caused by word coinage, accounting for only 4% of all the cases. Although the
number is small, the errors described above are the proof of the students lack of
knowledge about both collocation and vocabulary in general.

Formal confusion
The term formal confusion used in this study refers to the cases in which learners use
another part of speech of the desired word, or another English word that looks or

70
sounds like the desired one in place of the word itself. This seems similar to word
coinage; however, while word coinage results in incorrect vocabulary items, the
misused words resulting from formal confusion are perfectly correct ones in English.

Regarding the cases of misplaced parts of speech, a large number of errors of this kind
were identified. For examples, (powerful) Europe countries (cc cng quc chu
u), literature works (cc tc phm vn chng), power countries (cc cng quc)
and densely fog (sng m dy c) were the incorrect versions of European
countries, literary works, powerful countries and dense fog respectively. The point
here is that, most of the combinations above were expected to be familiar with the
students. Indeed, as third year English majored students, it is quite unusual for them
not to know those combinations. It is, therefore, possible to think that the students
carelessness and their anxiety while taking the tests were the main reasons behind the
making of these errors.

With respect to the errors in which the incorrectly-used words were similar to the
desired ones in spelling or pronunciation, several cases were identified, including
combinations such as coal reefs, tasteful food, small load (of bread), etc. It is possible
that they mistook the words underlined with the desired ones, which are coral, tasty,
loaf respectively. However, it is also possible that these were just mere spelling-
mistakes. Either way, the errors, similar to those mentioned above, reflect a certain
degree of the students carelessness, as well as their lack of vocabulary knowledge in
general.

In total, there were 112 cases belonging to this category, accounting for 23.7% of all
the cases observed, making formal confusion the third most common cause of errors.
One more noticeable point is that some combinations, which were supposed to be
familiar with the students, were produced incorrectly quite frequently. For example,
the phrase vic bo v mi trng (environmental protection) was translated into
environment protection 16 times. Similarly, the combination powerful Europe
countries mentioned above occurred 8 times. All these figures together signify how
mental factors (e.g. carelessness, anxiety, etc.) in addition to knowledge might
possibly affect the students linguistic performance in general and their use of
collocations in particular.

71
Others
This category was added in order to discuss the cases that did not belong to the
aforementioned causes of errors or could not be explained by a single one of them.
There were only 21 incidences in this category, making up 4.4% of the totality.
However, the information these errors provided was quite interesting.

Firstly, there were some strange combinations such as diligent web, coral oust ranges
which were the translations of chuyn trang (du lch) (travel column/magazine) and
di ngm san h (underwater coral reef) respectively. In fact, the reason why the
words like diligent and oust had been used in these combinations was inexplicable,
since there was not any connection in form, meaning or pronunciation between these
words and the correct ones. However, upon checking some popularly used e-
dictionaries such as Vdict by accident, it was found out that oust was among the
answers for the entry ngm (the meaning of ngm here, however, is similar to
stab in the back), and diligent was one of the answers for the entry chuyn (which
means hard working in this case). Such coincidence brought about a possibility that
the words used in the strange combinations mentioned above might came from such
sources like this dictionary. Given the fact that e-dictionaries are being used quite
commonly today due to their availability on smartphones and online sources, this
assumption seems plausible. Upon further investigation, such coincidence was found
in combinations previously listed in the other categories. Some combinations like
coral field, coral plain (bi san h coral reef), contrast relation (mi quan h tng
phn contradictory relationship), and attractive street food (mn n ng ph hp
dn delicious street food) are typical examples, in which field and plain were found
in the entry bi while contrast and attractive were seen in the entries tng phn and
hp dn respectively. Although it is not confirmed with 100 percent of certainty, the
existence of these combinations and such coincidence might signify possible influence
of the way the students used dictionaries, especially Vietnamese-English e-
dictionaries, on their making of collocational errors. They also suggest the pressure of
time the students might have to face during the test, under which they could not
consider the word they found carefully. This idea is, in fact, in line with the opinions
of Farrokh (2012) who suggested that when students face a difficult collocation, they

72
may seek help from a dictionary. It is the kind of dictionary that decides if their
problem can be successfully solved or not.

In addition to the combinations discussed above there were others whose causes could
not be identified in any way. They included cases like dim name, bad view (s phn
hm hiu unfortunate fate), ancient people (ng cha ta our ancestors/fathers),
current days (nhng nm gn y recent years), etc. It was inexplicable how these
errors were made. Maybe, the students, under the pressure of time, just used any
vocabulary items they could think of or were familiar with to compensate for their lack
of knowledge. The cause of these errors is closest to approximation, but the sense
relation between the misused words and the desired ones was not strong enough to put
them into that category.

To sum up, upon investigating the causes of the errors related to adjective-noun
collocations collected from the participants midterm-exam papers, based on the
criteria set by the hypothetical causes of errors, some conclusions are possibly made.
Firstly, L1 transfer and approximation, conforming with Dang (2014), were the most
common causes of errors. Moreover, the results revealed that the students lack of
vocabulary knowledge in general and their uncertainty about words spelling or their
parts of speech in particular also contributed to the making of errors. Last but not least,
it was also suggested that careless use of dictionaries and time pressure might possibly
lead to collocational errors.

4.1.2. Analysis of collocational errors from the translation test

The purpose of the translation test, as stated above, is to draw more information about
the learners errors, which the analysis of their exam papers may not have revealed.
The test consisted of 20 items and was taken by 89 students. From these 89 test papers,
only 1594 responses were satisfactory. The other 186 cases were either not answered
or incomplete, which were also considered as a type of error, and will be discussed
later. Among these satisfactory responses, there were 539 erroneous combinations.
These combinations together with the 186 instances of no/incomplete answers
mentioned above made a total of 725 errors. The other 1055 responses were acceptable
answers, consisting of 903 correct collocations, 84 free combinations and 68 cases of

73
single words. The occurrence of responses in the form of single words was, in fact,
unexpected since it had been highlighted in the test that the translations should be
adjective-noun combinations. However, as these single-worded responses were
lexically and grammatically correct, they were also considered as acceptable answers.
A summary of the result is presented in Table 4.4 below.

Table 4.4: Summary of students responses to the translation test


Count Total Percent Total
No/Incomplete Answers 186 10.4%
Errors 725 40.7%
Erroneous Combinations 539 30.3%
Single Words 68 3.8%
Acceptable
Free Combinations 84 1055 4.8% 59.3%
Answers
Correct Collocations 903 50.7%
Total 1780 1780 100% 100%

The result above has several noticeable differences from that of the error analysis
conducted on the exam papers. Firstly, the frequency of free combinations is very low.
With 84 free combinations observed, accounting for only 4.8% of the totality, it is
much lower than the frequency of free combinations collected from the midterm-exam
papers, which was 20%. This low frequency of occurrence possibly resulted from the
nature of the translation test, which aimed to test the participants production of
adjective-noun collocations in English. Because of that, most of the target Vietnamese
combinations in the test needed specific English adjective-noun collocations as their
equivalents, which limited the choice of vocabulary items and the number of
acceptable combinations.

Secondly, with 725 observed cases, accounting for 40.7% of the totality, the
proportion of errors from the translation test is higher than the proportion of errors
from the midterm-exam papers, which is only 28.8%. The difference of 11.9% is quite
a noticeable number, which might possibly result from two factors. Firstly, thanks to
the design of the test, the cases where no/incomplete answers were given can be easily
identified, which allows the investigation into a new category of error. Secondly, the
level of difficulty of the test might also be a reason for the high frequency of errors.
Given the facts that many of the target combinations were expected to be unfamiliar
to the students, it is possible to assume that the translation test was more demanding

74
than the midterm-exam in terms of collocation, which resulted in a greater number of
errors being made.

It can be seen, through the brief overview above, that the translation test can
undoubtedly provide some new information which the analysis of the students exam
papers could not. In order to fully exploit this valuable source of data, a thorough
analysis of the errors was conducted, which also involve two major steps: categorizing
the errors according to their types and identifying their possible causes.

4.1.2.1. Types of errors

Similar to the errors from the students exam papers above, the incorrect combinations
collected from the translation test were also put into three categories wrong choice of
adjective, wrong choice of noun and wrong combination, following the framework
mentioned in Chapter 2. Regarding the cases where no answer or incomplete answers
were given, which have been also considered as a type of error, a new category, namely
No/Incomplete Answer, was added to address them since it is impossible to put these
errors into any of the three other categories. According to these four categories the
errors were categorized and calculated, which resulted in the figures presented in Table
4.5 below.

Table 4.5: Types of collocational errors from the translation test


Types of Errors Count Percent Examples
light tea, condense milk, purple eyes,
Wrong Choice of Adjective 462 63.7%
dead sentence

Wrong Choice of Noun 27 3.7% bad healthy, great proud/pleasure

sweet water, executive charge,


Wrong Combination 50 6.9%
unhealthy condition
No/Incomplete Answers 186 25.7%
Total 725 100.0%

Regarding the three types of erroneous combinations, the errors analyses of the
translation test and of the students exam papers show a similar tendency, with wrong
choice of adjective being the broadest category, following by wrong combination and
wrong choice of noun respectively. As the number indicated, there were 462
combinations incorrectly produced due to wrong choices of adjectives, accounting for

75
63.7% of all the errors. Combinations such as light tea (tr long weak tea),
condense milk (sa c condensed milk), purple eyes (mt bm black eyes) and
dead sentence (n t hnh death sentence) are examples of this type. Considering
wrong combination, there were 50 instances in this category, including some
combinations such as sweet water (nc ngt soft drink), executive charge (n t
hnh), unhealthy condition (sc khe km poor health). Wrong choice of noun, as
the category with the lowest frequency of occurrence, had only 27 instances,
accounting for 3.7% of the totality. Some examples of this error type are bad healthy
(sc khe km), great proud and great pleasure (vinh d ln lao great honor).
Compared to the figures from the error analysis of the students exam papers, the
statistics here show some differences. Regardless of those statistical dissimilarities,
the results confirm the aforementioned finding, which suggests that choosing correct
adjectives is the commonest challenge to the students when it comes to producing
adjective-noun collocations in English.

With the new category of error (i.e. no/incomplete answer) taken into account, the
analysis shows some valuable information which was unobservable in the previous
one. Interestingly, a large number of test items were left with either no answers or
incomplete ones. In total, there were 186 incidences, accounting for 25.7% of the
totality, which was a considerable number compared to the percentages of the two
categories wrong choice of noun and wrong combination (3.7% and 6.9%
respectively). This kind of error may result from several factors, including the
expected high level of difficulty of the test, the pressure of time, or the fact that the
test was not considered important by the students as it did not influence their academic
result. That is to say, the high frequency of occurrence of this error type may signify
the participants noticeable lack of either collocational knowledge or motivation in
completing the test.

To sum up, the results above confirm the findings of the analysis of errors from the
students midterm-exam papers, and provide valuable information about the new
category of error type. The next section will be dedicated to identifying the causes of
the collected errors.

76
4.1.2.1. Causes of errors

Again, the list of hypothetical causes of collocational errors was employed as the
framework for the conduct of this step. This list, as mentioned above, consists of six
categories including (1) approximation, (2) use of synonyms, (3) L1 transfer, (4) word
coinage, (5) formal confusion and (6) other. Regarding the instances where
no/incomplete answers were given, however, these errors were not incorrect
combinations, and thus could not be grouped into any of the hypothetical causes above.
In order to address this kind of error, a new category, namely abandonment of tasks
was added. As a result, the list of hypothetical causes of error used here included seven
categories in total. Based on them, the possible causes of the collected errors were
identified and presented in Table 4.6 below.

Table 4.6: Causes of collocational errors from the translation test


Causes of Errors Count Percent Examples
imitate leg, weakened health,
Approximation 147 19.3%
carbodinate drink, plastic/ceramic teeth

Use of Synonyms 38 5.0% manmade leg, thick population

black fortune/destiny, enough condition,


L1 Transfer 186 24.4%
condensed soup

Word Coinage 15 2.0% hurted eyes, eated oil, contensy milk

tense population, poor qualification,


Formal Confusion 59 7.7%
bad healthy, value experience
stiff/ solid/ tough/ tight soup, heavy
Others 132 17.3%
milk/ soup, light tea/ soup
Abandonment of Tasks 186 24.4%
Total 763 100.0%

Approximation
The figures show that approximation is a major cause of errors in the translation test,
with 147 instances, making 19.3% of the totality. This conforms with the result from
the analysis of the exam papers above and also with the studies of Hong et al. (2011)
and Dang (2014), which have pointed out that approximation is one of the most
common causes of collocational errors.

77
In addition, the analysis also revealed some new information regarding how some of
the errors had possibly been made in relation to the use of approximation. As
mentioned above, approximation is the use of one vocabulary item in place of another
due to certain semantic similarities between them. This kind of approximation, which
had been commonly seen in the exam papers, was also observable in the translation
test, with some examples such as imitate leg (chn gi artificial leg), unlucky thing
(vn en bad luck), weakened health (sc khe km poor health), etc.

However, there is another kind of approximation which was also commonly used by
the students in this test. With this kind of approximation, the participants did not base
on sematic similarities between vocabulary items as described above, but resorted to
their knowledge about the subjects which the target combinations refer to and tried to
create combinations that describe the subjects characteristics. For example, to
translate the phrase nc ngt (soft drink), some students created several combinations
such as gas drink, carbodinate drink, carbondized water, which might result from their
knowledge that soft drinks usually contain carbon dioxide a kind of gas. Similarly,
cork/plastic leg, plastic/ceramic teeth, which are incorrect translations of chn gi
(artificial leg) and rng gi (false teeth) respectively, might be the result of the
students assumption that artificial legs and false teeth are made from those materials.
Some other examples are purple eyes, black eyes, bluish eyes, blue eyes which possibly
originated from the students knowledge about the possible colors of a bruised eye.
These translations may also embrace the influence of the participants culture, in
which colors such as purple or blue are usually used to describe the color of bruises.

While the reason why this way of approximation was commonly seen in the translation
test and not in the exam papers is hard to explain, it is an undeniable fact that the
discovery of this strategy as a cause of collocational errors is a valuable finding,
shedding more light on how collocational errors were made and signifying the
students lack of vocabulary knowledge in general.

Use of synonyms
The use of synonyms here is also a cause of adjective-noun collocational errors, but
not a major one. In total, there were 38 errors resulting from the use of synonyms,
making up only 5% of the totality, which is noticeably smaller than the number found

78
in the analysis of the exam papers (15%). A viable explanation for this is the difference
in content between the two tests. It is possible that the target combinations in the exam
papers stimulated the use of synonyms more than those in the translation test.

While the number is small, it still has its own significance. Collocational errors caused
by the use of synonyms occur when learners use a synonymous word in place of the
desired one, unaware of the restrictions in word combinations. For example, some
students translated the phrases chn gi (artificial leg) and dn s ng (dense
population) as manmade leg and thick population respectively, without knowing that
these synonymous words (i.e. artificial manmade; dense thick) could not be used
interchangeably in these contexts. The errors, in other words, indicate the participants
lack of collocational knowledge, which strengthens the result from the analysis of the
midterm-exam papers.

L1 transfer
It has been suggested by many studies and also by the error analysis of the exam papers
above that L1 transfer is a major cause of lexical collocational errors. This idea is once
again confirmed by the statistics presented in Table 4.6 above. With 186 instances,
making up 24.4% of the totality, L1 transfer was the primary cause of errors in the
translation test. Similar to those in the exam papers, most of the errors caused by L1
transfer here took place when the students either resorted to word-for-word translation
or failed to make a distinction between lexical equivalence and collocational
equivalence, which leads to a false assumption that two words being equivalent in one
context will have the same relationship in others.

Regarding the first circumstance, several errors caused by the use of word-for-word
translation could be identified, including black fortune/destiny (vn en bad luck),
enough condition (iu kin sufficient condition). In the first example, it seems
that the student failed to understand the figurative meaning of the word en, which
means bad here, and thus ended up with such an incorrect combination. Considering
the second one, the students used word-for-word translation to compensate for their
lack of vocabulary knowledge, unaware that enough is an adverb and thus could not
be used in such a way.

79
Compared to the use of word-for-word translation, the second circumstance seems to
take place more frequently with a large number of combinations incorrectly made due
to the participants inability to distinguish between lexical equivalence and
collocational equivalence. Although this phenomenon was mentioned during the error
analysis of the students midterm-exam papers, more evidence for it could be found
here thanks to the design of the translation test. In order to see how the participants
would respond to the situation when a word in Vietnamese has different English
equivalents depending on the combinations in which it appears, some target
combinations such as sp c (thick soup), sa c (condensed milk), sp long (thin
soup) and tr long (weak tea), etc. were specially included in the test. Upon
investigating the students responses to these combinations, the researcher found that
the students tended to use the same English equivalent for one Vietnamese word,
regardless of the combination in which it appeared. For example, there were eleven
students translating the word c in both sa c and sp c as condensed/condense
unaware that the words c in the two combinations had different meanings and thus
the latter one could not be translated into condensed. Similarly, thirteen of the students
used the word artificial as the equivalent for the word gi in both chn gi (artificial
leg) and rng gi (false teeth), without knowing that the use of this word was
inappropriate in the latter combination. In addition to the two examples above, there
were many other instances that can be identified, such as crowded population,
precious experience which were incorrect translations of dn s ng (dense
population) and kinh nghim qu bu (valuable experience) respectively.

The discussions above, in general, support the result of the error analysis conducted
on the students exam papers regarding the influence of L1 on their making of
collocational errors and provided relatively dependable evidence of their inability to
realize the mismatch between lexical equivalence and collocational equivalence.

Word coinage
Among the minor causes of collocational errors, word coinage is one that signifies
learners deficiency of vocabulary knowledge. It can be seen from the figures above
that word coinage occurred quite infrequently in the translation test, with only 15
instances (2%). Some of them are malformed words resulting from incorrect use of

80
affixes, including hurted eyes, eated oil, etc. The others are words that do not exist in
English such as carbodinate drink, contensy milk, licid eyes, etc. which might result
from the students efforts in creating a vocabulary item basing on their vague memory.

Formal confusion
The figures show that formal confusion was the cause of 59 incorrect combinations
found in the translation test, accounting for 7.7% of all the observed incidences. As
mentioned above, formal confusion involves incorrect uses of vocabulary items in
place of desired ones due to formal associations between them. For example, tense
population and poor qualification, which were incorrect translations of dn s ng
(dense population) and cht lng km (poor quality) respectively, might result from
the formal similarities between tense and dense, qualification and quality. Similarly,
gracious/precise experience were possibly caused by formal confusion, due to their
formal associations with the word precious. However, it is noted that precious
experience is also an incorrect translation of kinh nghim qu bu (valuable
experience), caused by L1 transfer, which means that those errors above might result
from two causes (i.e. formal confusion and L1 transfer). Formal confusion also refers
to the cases where an inappropriate part of speech of a desired word was used, with
several instances identified in the translation test, including died/dead sentence, bad
healthy, value experience and bruise eyes. Those combinations were incorrect
translations of n t hnh (death sentence), sc khe km (poor health), kinh nghim
qu bu (valuable experience) and mt bm (bruised eyes) respectively. As previously
mentioned, formal confusion may signify the students deficiency in vocabulary
knowledge and/or the influence of mental factors (e.g. anxiety, motivation, etc.) on
them.

The figures also indicate that the occurrence of formal confusion in the translation test
is noticeably less frequent than that in the students exam papers (with 112 instances
observed, accounting for 23.7% of all the totality). However, this does not necessarily
mean that the participant performed better in the translation test. Usually, formal
confusion takes place when learners have certain ideas about the desired vocabulary
items but cannot produce them correctly and thus end up with incorrect substitutions
that formally associate with the intended ones. However, with unfamiliar word

81
combinations, it is unlikely that they could have even the slightest idea about the
desired words in mind and thus formal confusion, probably, would not occur in such
cases. Given the fact that the test was expected to be more demanding than the midterm
exam in terms of collocational knowledge, the idea above seems to be a viable
explanation for the low frequency of formal confusion found in the translation test.
Besides, the relatively large number of responses identified as no/incomplete answers
in the translation test also supported the aforementioned proposition. The participants,
without any idea about the desired combinations, were not likely to make errors related
to formal confusion. Instead, they opted to other strategies or even gave up on the
tasks, which resulted in the large number of no/incomplete answers mentioned above.
In sum, the low frequency of occurrence of formal confusion seen in the translation
test is not necessarily an encouraging sign. It may, instead, signify the participants
unfamiliarity to the given target combinations.

Others
This category, similar to the one in the analysis conducted on the exam papers, was
included to address other causes of errors besides the five hypothetical ones mentioned
above. As the figures indicated, there were 132 incidences in this category, making up
17.3% of the totality. These errors, as the analysis revealed, might result from the
following causes.

Firstly, there were combinations such as light tea (tr long weak tea), light soup
(sp long thin soup), heavy milk (sa c condensed milk) and heavy soup (sp
c thick soup), some of which had quite noticeably high frequency of occurrence,
for example, light tea appeared 33 times and light soup 14 times. At first, the causes
of these errors and their high frequency of occurrence seemed to be inexplicable.
However, upon close look, a suggestive detail was identified. That is one of the items
in the test, which is as follow:

Tr long l loi tr c mu v v u nht.


"__________________" means the tea that is light in color and taste.

In the given English translation, the adjective light was used to describe the taste and
color of weak tea, and quite coincidentally, this word also appeared 33 times in the
incorrect combination light tea mentioned above. It is, thus, possible that the students,

82
unable to find the appropriate adjective, opted to use the word they found right in the
test due to the fact that it was used to describe the attributes of the tea. If this is the
cause of the error, then it also explains for the making of the other three. Regarding
the case of light soup, the students translating tr long as light tea above might
probably apply the adjective light to the translation of sp long due to the appearance
of the word long in both combinations, which resulted in such a deviant combination.
This proposition seems to be supported by statistics. Out of the sixteen students who
produced the combination light soup, fourteen of them also translated tr long as
light tea, making it a viable explanation for the cause of this error. With respect to the
cases of heavy milk and heavy soup, it is possible that the students used the adjective
heavy in these combinations as it is an antonym of light in certain contexts, given the
fact that c and long also have the same relationship in Vietnamese. This inference
was also backed by the statistics, which showed that all of the five cases of heavy soup
and five out of the six instances of heavy milk came from the students who also
produced the combinations light soup and light tea. With the suggestive details and
statistics above, it could be suggested that the students use of the aforementioned
strategy, in which they relied on certain appealing vocabulary items or details in the
content of test to compensate for their lack of knowledge, was the possible causes of
these errors. This strategy, to some extent, seems like a kind of approximation.
However, it was included in this category due to its distinctive origin and the
complicated discussion needed to prove its existence.

There were other errors, which might originate from the students use of dictionary
previously mentioned in the error analysis of the exam papers above. They include
deviant combinations such as stiff soup, solid soup, tough soup, tight soup which were
all incorrect translations of the phrase sp c (thick soup). While there was not a vivid
connection between the desired adjective and the incorrect ones which could account
for the use of these substitutions, it was found out that all these adjectives (i.e. stiff,
solid, tough and tight) appeared in the entry c of several Vietnamese-English e-
dictionaries. Such a coincidence suggested that the students, in an effort to seek for
necessary vocabulary items and under the pressure of time, had checked dictionaries
and carelessly picked these adjectives.

83
Aside from those errors, there were other cases in this category which have
unidentifiable causes. They include combinations such as good/necessary/prior
condition, life sentence, great pleasure, etc. which were incorrect translations of iu
kin (sufficient condition), n t hnh (capital punishment) and vinh d ln lao
(great honor) respectively. As suggested in the error analysis of the exam papers, it
seemed that the students, unable to produce the desired translations, opted to use any
vocabulary items they could think of or were familiar with, which resulted in such
erroneous combinations.

Abandonment of tasks
Abandonment of tasks refers to learners act of giving up on a task without completing
it, which might result from their lack of knowledge, motivation or shortage of time.
The figures show that this category, along with L1 transfer, was the primary cause of
errors in the translation test, with 186 instances of no/incomplete answers collected,
accounting for 24.4% of the totality. Given the fact that the students were provided
with a sufficient amount of time to complete the test, this high frequency of occurrence
signifies their lack of motivation. It also possibly indicates their unfamiliarity with
the target combinations in the test, which conforms with the explanation for the low
frequency of occurrence of formal confusion discussed above.

This cause of error, which was not investigated in the analysis of the exam papers,
shed more light on the students making of collocational errors. It suggested that
motivation is a very important factor, without which learners, facing a challenging
combination, are likely to abandon it rather than trying other strategies to overcome
the difficulty.

4.1.3. Summary of the error analyses

The two error analyses above provide extremely valuable information for the present
study. Although there were differences in their sources of data (i.e. the students
midterm-exam papers and the translation test), they yielded several similar results,
which supported each other. As can be seen in Figure 4.1 below, both analyses arrived
at the result that wrong choice of adjective was the broadest category among the three

84
types of errors, which suggests that choosing correct adjectives for adjective-noun
combinations in English is a constant challenge to the students.

70.00%
60.00%
50.00%
40.00%
30.00%
20.00%
10.00%
0.00%
Wrong choice of Wrong choice of Wrong combination No/Incomplete
adjective noun answer
From the midterm exam papers From the translation test

Figure 4.1: Summary of types of collocational errors

With respect to the causes of errors which were summarized in Figure 4.2 below, both
analyses revealed that L1 transfer and approximation were the first and the second
primary causes of errors respectively. Use of synonyms and word coinage were also
among the causes of errors but occurred with lower frequencies. The results above
suggested that the students making of collocational errors was heavily influenced by
their L1. They also indicated that the students lack of collocational and vocabulary
knowledge, as well as their carelessness and other mental factors when taking the tests
might lead to collocational errors.

30.00%
25.00%
20.00%
15.00%
10.00%
5.00%
0.00%

From the midterm exam papers From the translation test

Figure 4.2: Summary of causes of collocational errors

85
However, there were also notable differences between them. Firstly, the frequencies
of occurrence of formal confusion, one of the causes of error, were noticeably different
between the two analyses. This dissimilarity, as previously discussed, suggested that
formal confusion, which signifies the students carelessness and lack of vocabulary
knowledge, occurs more frequently when learners are familiar with target
combinations than when they are not. Regarding the category called others, the error
analysis conducted on the exam paper proposed that incorrect use of Vietnamese-
English dictionaries, especially e-dictionaries, possibly accounted for some of the
collected errors, which was later confirmed by the error analysis of the translation test.
The latter analysis, moreover, suggested the use of a strategy in which the students
use a vocabulary item in the content of the test, which looked suggestive to them, to
compensate for their lack of knowledge. This strategy also led to collocational errors.
Finally, the error analysis conducted on the translation test allowed the investigation
into another cause of error, namely abandonment of tasks. The result of the analysis
showed a high frequency of occurrence of this phenomenon, which might result from
the students lack of motivation and unfamiliarity with the target combinations.

4.1.4. Learners receptive knowledge of collocations

It has been suggested from the analyses above that a large number of the errors
collected had certain connection with the participants lack of knowledge about
collocation. Regarding collocational knowledge, most authors have distinguished
between receptive and productive ones and stated that learners receptive knowledge
of collocation had influence over their productive knowledge. While the error analyses
above provided comprehensive view into the students productive ability, they gave
little information about their receptive knowledge. Therefore, in order to thoroughly
investigate the participants knowledge of collocation, a receptive test in the form of
a multiple choice one was conducted.

The result of this test was compared with that of the translation test to see how the
students receptive knowledge of collocations is in comparison to their performance
in the productive task (i.e. the translation test). To ensure the accuracy of the
comparison, the student having done only one of the tests were excluded. As a result,
only 61 of the participants, who had attended both of them, were taken into account.

86
More importantly, as mentioned in Chapter 2 a new scoring criterion was applied to
the translation test, by which only responses in the form of correct adjective-noun
collocations were considered acceptable to ensure to relevance between the two tests.
With the results of the tests as the variables, a paired-samples t-test was carried out,
the result of which is presented in Table 4.7.

Table 4.7: Comparison between productive test and receptive test


Paired samples statistics
Mean N Std. Deviation Std. Error Mean
Pair 1 Productive test 10.03 61 2.989 .383
Receptive test 8.92 61 2.610 .334

Paired samples test


Pair 1
Productive test Receptive test
Paired Mean 1.115
Differences Std. Deviation 3.852
Std. Error Mean .493
95% Confidence Interval of Lower .128
the Difference Upper 2.101
t 2.260
df 60
Sig. (2-tailed) .027

The paired-sample t-test was conducted at 95% confidence interval of the difference,
with the zero hypothesis that there is no statistical difference between the results of
the receptive and the productive tests. The sig. number (0.027 < 0.05) indicates that
the zero hypothesis is rejected. That is to say, the results of the two tests are statistically
significantly different. Moreover, the mean value of the receptive test scores is 8.92,
which is lower than that of the productive test scores (10.03). The numbers above
signify that not only did the students performed differently in the two tests but they
also performed slightly worse in the receptive one. This is quite unexpected compared
to the results of previous studies which have suggested that learners tend to do better
in receptive tests than in productive ones.

The students poor performance on the receptive test might come from the fact that
the test, with three options, required the ability to recognize not only correct
collocations but also incorrect ones. With limited knowledge about collocations, they

87
might have difficulties in telling whether an unfamiliar combination was a correct
collocation or not. This is observable by comparing the results of the two tests together,
item by item. (Before being compared, the items in the receptive test were rearranged
to match the order of those in the translation test.)

Table 4.8: Students performance on each item of the two tests


Translation Test Receptive Test
Correct Column Correct Column
Answer Total N % Answer Total N %
Item 1 46 75.4% 13 21.3%
Item 2 15 24.6% 14 23.0%
Item 3 18 29.5% 30 49.2%
Item 4 9 14.8% 33 54.1%
Item 5 12 19.7% 23 37.7%
Item 6 22 36.1% 47 77.0%
Item 7 21 34.4% 13 21.3%
Item 8 47 77.0% 52 85.2%
Item 9 50 82.0% 6 9.8%
Item 10 21 34.4% 21 34.4%
Item 11 44 72.1% 31 50.8%
Item 12 43 70.5% 29 47.5%
Item 13 54 88.5% 11 18.0%
Item 14 35 57.4% 23 37.7%
Item 15 37 60.7% 32 52.5%
Item 16 16 26.2% 43 70.5%
Item 17 25 41.0% 23 37.7%
Item 18 24 39.3% 39 63.9%
Item 19 33 54.1% 16 26.2%
Item 20 40 65.6% 45 73.8%

It can be seen, in Table 4.8, that the students performances in the two tests were
inconsistent. Firstly, there were cases, such as items 4, 6 and 16, where they performed
significantly better in the receptive test than in the other. Most of them were able to
recognize the correct collocations included in these items (i.e. thin soup, condensed
milk and sufficient condition), regardless of the distractors. However, when being
asked to translate their Vietnamese equivalents (i.e. sp long, sa c and iu kin
) into English, only a few could successfully produce these collocations as
translations. This is in line the proposition of Alsakran (2011) that learners productive
knowledge of collocations may lag behind their receptive knowledge. In other words,

88
that they are able to recognize a correct collocation does not ensure that they can
successfully produce it when needed.

On the contrary, there were also incidences where the students productive
performance surpassed their receptive one. For example, most of them were able to
produce the target collocations as correct answers to items 1, 9 and 13 in the translation
test, but not many could answer these items in the receptive one correctly. This result
is quite surprising as it seems contradictory to Alsakrans idea above. However, it is
explicable with the designs of the two tests being taken into account. To be precise,
item 1, 9 and 13 in the translation test required the students to provide the English
equivalents of the phrases chn gi, du n and gi ln respectively. Among the
acceptable answers to these items were common English collocations such as artificial
leg, cooking oil and strong wind which most of the students were able to produce.
Nevertheless, in the receptive test, these collocations were paired with other correct
but less commonly used collocations: fake leg, edible oil and powerful wind. The
correct answer to these items in the receptive test, therefore, was the option both are
correct. However, while most of the students could correctly produce the former
collocations, only a few of them were able to recognize that the other ones were also
correct ones, which explained the results.

To sum up the receptive test suggested gaps in the students knowledge about
collocations. Due to the lack of knowledge, they failed to recognize collocations which
were unfamiliar to them. Even with ones they could recognize, there were chances that
they were unable to produce them correctly, which directly hindered their productive
performance.

4.1.5. The teaching and learning of collocations in Basic Translation

The error analyses and the receptive test of collocation above have suggested a lack of
knowledge about collocations among the student respondents, which possibly has
certain connection with the teaching and learning process. In order to have a
comprehensive insight into the students making of errors and find possible remedies,
the actual teaching and learning of collocation in the three Basic Translation classes
were investigated based on the data collected from the interview with the teachers and
the questionnaire to the student respondents.

89
4.1.5.1. The teaching of collocations

The teaching of collocations and related concepts in Basic Translation were


investigated through data collected from both the teachers and the students.

From teachers perspective


Firstly, it is important to know the teachers perception, because it is an important
factor deciding their actual teaching practices. In the present study, the teachers
perception of collocation was investigated through questions 1 and 2 in the interview.
Regarding the first question, which aimed to check their knowledge of collocation, all
three teachers demonstrated their comprehension of the term collocation. However,
only T1 was able to further elaborate on the types and uses. In response to question 2,
all of them agreed on the idea that knowledge about collocations plays an important
role in translation. Translation is not a simple process of translating word-for-word.
It requires a thorough understanding of a texts message in order to transfer it to a
target language. Knowledge about collocations not only helps in discovering the
meaning of a text but also provides efficient ways to translate it, T2 explained.
Commenting on the significance of collocational knowledge in translation, T1 added
that using collocation makes students language more natural and understandable with
alternative and richer ways of expressing their ideas. In general, all of the teachers
being interviewed recognized the importance of collocational knowledge in
translation, which had had certain influence on their teaching practices.

With respect to the teachers actual instruction on collocations in Basic Translation


classes, the relevant data were collected through question 3 in the interview. All of the
teachers being asked affirmed their teaching of collocations in Basic Translation
classes. However, upon further questioning they revealed that they had not often
provided explicit instruction on collocations. Only T1 claimed to use activities such
as gap-filling and paraphrasing to teach collocations. The other two, most of the time,
had employed integrated teaching method by drawing students attention to
collocations (if they existed) in translation texts. While this method is efficient as it
allows teachers to teach translation and collocations at the same time, its effectiveness
heavily relies on the texts. If the texts contain few or inappropriate collocations, this
method does not seem applicable. In addition to those methods, the teachers had also

90
encouraged the students self-studying to broaden their knowledge about collocation
by reading relevant material, learning vocabulary in chunks or blocks, and making
their own glossary, etc. However, they did not mention having any method to monitor
the students self-learning of collocations. Without their monitoring, it is possible that
some students did not have enough motivation to study seriously by themselves.

From students perspective


The teaching of collocations in Basic Translation classes was investigated not only
from the teachers perspective but also from the students through questions 10 and 11
in the questionnaire. Among 104 student respondents, only 72 answered Q10 as the
other ones were already filtered by Q1 as described in Chapter 3.

Table 4.9: Students statement about the teaching of collocations


in Basic Translation
Frequency Percent Valid Percent Cumulative Percent
Yes 60 57.7 83.3 83.3
Valid No 12 11.5 16.7 100.0
Total 72 69.2 100.0
Missing System 32 30.8
Total 104 100.0

The figures show that the students had different opinions on the teaching of
collocations in Basic Translation. More than half of the respondents (57.7%) affirmed
their teachers instruction on collocations in class while 11.5% of them denied it, not
to mention a large percentage of the participants (30.8%) did not answer this question
due to their answer to the filtering question that they did not understand the notion of
collocation. The figures above indicate a mismatch between the teachers and the
students opinions. While all of the teachers claimed that they had had some teaching
of collocations in Basic Translation classes, only more than half of the students had
the same idea. One possible explanation is that the teachers had not usually given
explicit instruction but employed the integrated teaching method, and thus the
students, not paying enough attention or not having enough relevant knowledge, had
been unaware of their integrated teaching of collocations.

91
Among the 60 students answering yes to Q10, only 59 of them actually responded to
Q11, which asked about the strategies used by the teachers to teach collocations. Their
answers are presented in Table 4.10 below.

Table 4.10: Strategies used by the teachers to teach collocations


Strategies Count Column N %
Introducing the term collocation 17 28.8%
Introducing the role of collocation 8 13.6%
Directing students attention to collocations in tasks/texts (1) 29 49.2%
Teaching some collocations in class 26 44.1%
Encouraging self-study (2) 31 52.5%
Reviewing learned collocations 17 28.8%
Using collocation dictionaries 5 8.5%
Using linguistic corpora 1 1.7%
Identifying errors, asking students to look for corrections (3) 24 40.7%
Identifying errors and giving corrections (4) 30 50.8%
Others 0 0.0%
Total 59 100.0%

It can be seen that the students opinion was quite in line with the teachers. Strategies
(1), (3) and (4) above belong to the integrated teaching of collocation mentioned by
the teachers as their commonly used method. By directing students attention to
collocations in texts and identifying collocation-related errors, the teachers could make
use of translation tasks to increase students awareness towards collocations as well as
their collocational knowledge. Giving corrections to their errors was also useful but
not always doable, possibly due to the heavy workload the teachers had to cope with.
Regarding strategy (2), it was also frequently used by the teachers as previously
presented in the interview with them. However, the uses of linguistic corpora and
collocation dictionaries, despite their usefulness, were affirmed by few student
respondents, which indicates that they are among the infrequently used strategies.

To sum up, the data suggested that collocations were taught in Basic Translation
classes. However, the teachers usually provided instructions about collocations
through integrated-teaching and encouraged students self-study, while other useful
strategies such as using collocation dictionaries and linguistic corpora were
infrequently used. Although integrated teaching of collocations is convenient and
appropriate in the situation that collocation is not a part of the Basic Translation

92
curriculum as the teachers explained, its effectiveness heavily relies on whether the
texts have necessary collocations to teach or not. Given the remarks of two teacher
respondents that the texts used to teach Basic Translation might contain few
collocations and some of them were outdated and no longer appropriate for teaching,
it is possible to doubt the effectiveness of the aforementioned method. Similarly, self-
study requires students ability to work on their own, effort and responsibility. With
one third of the student respondents did not understand the concept of collocation and
another 11.5% negated their teachers teaching of collocations in class, this method
may not be as effective as expected.

4.1.5.2. The learning of collocations

The learning of collocations was also investigated from students and teachers
perspectives. For the former one, the relevant data were collected through questions 1
to 8 in the questionnaire. For the latter, the data were obtained through questions from
4 to 7 in the interview.

From students perspective


Firstly, similar to the teachers, the students perception of collocations, as one of the
factors determining their learning activities, was investigated thanks to questions 1 to
4 in the questionnaire. Q1, as previously presented was a filtering question, used to
opt out the student respondents who were unaware of collocation. Among 104 students
responding to this question, nearly one-third (30.8%) did not understand the concept
of collocation. This is a noticeable number, provided that their Basic Translation
teachers had given instruction about collocation in class, not to mention the fact that
they, as 3rd year English majored students, should have learnt about it in previous
courses. The other 72 students, who knew about collocation, were questioned about
their source of knowledge (Q2). Their answers to this question are presented in Table
4.11 below, which shows that their knowledge came from various sources. The
majority of them (76.4%) obtained the knowledge from teachers. Only 17 of them
(23.6%) learnt from self-study by reading reference materials. The figures indicate that
the students were quite passive in acquiring the knowledge, with about two-thirds of
them relied on teachers instruction.

93
Table 4.11: Sources of students knowledge about collocations
Valid Cumulative
Frequency Percent
Percent Percent
Valid Studied in compulsory 38.9
28 26.9 38.9
curriculum
Introduced by teachers out 76.4
27 26.0 37.5
of compulsory curriculum
Reading reference materials 17 16.3 23.6 100.0
Other 0 0 0 100.0
Total 72 69.2 100.0
Missing System 32 30.8
Total 104 100.0

The students opinion on the role of collocations in translation was also investigated.
In response to Q3, which is a filtering question, all the students who had knowledge
about collocations affirmed their importance. Regarding Q4, asking about specific
roles of collocations in translation, they gave different opinions.

Table 4.12: Students opinions on the roles of collocations in translation


Roles of collocations in translation Count Column N %
Increase writing fluency 39 54.1%
Providing alternative and richer ways of expression 44 61.1%
Producing more precise and more natural language 39 54.1%
Maintaining and strengthening cohesion of the text 25 34.7%
Others 0 0.0%
Total 72 100.0%

This question, according to Mai (2010), aimed to find out to what extent the students
realized the benefits of using collocations in translation. Given the figures in Table
4.12 with only one option chosen by about 61% of the respondents, and two others
chosen by slightly more than 50% of them, it can be seen that the students were aware
of the importance of collocations but did not fully understand it.

With reference to the students acquisition and use of collocations, necessary data were
collected through several questions. Q5 aimed to investigate their acquisition of
collocations, or, precisely, the strategies used to broaden their knowledge. Q6, Q7 and
Q8, in general, dealt with the participants use of collocations. However, they only
focused on the aspects which could not be approached through the analyses of errors,
including the students awareness of choosing correct collocations in translation, the

94
likeliness to face difficulties in translating collocations and their attention towards the
collocational errors they made. Last but not least, Q9 was designed to investigate how
the respondents deal with collocation-related difficulties.

Table 4.13: Students strategies to broaden knowledge of collocations


Strategies Count Column N %
Reading reference materials 26 36.1%
Paying attention to collocations used in texts 53 73.6%
Memorizing words in groups or chunks 34 47.2%
Looking up for collocations in dictionaries 35 48.6%
Using linguistic corpora 6 8.3%
Others 1 1.4%
Total 72 100.0%

Table 4.13 shows different strategies the respondents used to broaden their knowledge
about collocation. Among them, learning by focusing on collocations found in texts
was most frequently used. This conforms with the result found in the investigation into
the teaching of collocation, which showed that the main method of teaching
collocations employed by the Basic Translation teachers was integrated teaching, or
precisely, drawing learners attention towards collocations appearing in texts. This
strategy is useful for the students as it allows them to practice other skills while
broadening their knowledge about collocation at the same time. The contexts in the
texts also help with their comprehension and acquisition. However, as mentioned
above, the effectiveness of this strategy relies heavily on whether the texts contain
enough collocations for learning and also whether the students really know what they
need to focus on. In addition, other useful strategies such as learning words in chunks
or reading reference materials were also used but by fewer students. Among them, the
former one has been suggested by follower of LA as an effective way to learn
vocabulary and collocation. However, only nearly half of the respondents made use of
this strategy. Collocation dictionaries, despite not being mentioned by the teachers in
the teaching of collocations, was used quite commonly by the students as a way to
broaden their knowledge. However, in contrast to collocation dictionaries, linguistics
corpora were mostly unused. It seems that corpora were still an unfamiliar concept to
the students, and thus without instructions they did not know what corpora were or
how to use them effectively.

95
Regarding the students use of collocation, their responses to Q6, Q7 and Q8 are
presented and compared in Figure 4.1 below.

60.0%
52.8%
51.4%
50.0%
38.9%
38.9% 38.9% 37.5%
40.0%

30.0%

19.4%
20.0%
9.7%
10.0% 8.3%
1.4% 2.8%
0.0% 0.0% 0.0% 0.0%
0.0%
Never Rarely Sometimes Often Very often

Being aware of choosing correct collocations in translation


Having collocation-related difficulties
Paying attention to collocational errors

Figure 4.3: Students responses to Q6, Q7 and Q8

The figure shows that the majority of the students had high awareness of choosing
collocations in translation, with 51.4% of the totality (72 students) often cared about
choosing correct collocations for their translation and 37.5% of them very often.
Despite that, producing correct collocations seems to be a constant challenge to them
as the figures show that more than half of the respondents (52.8%) often have
difficulties in translating collocations, another 38.9% of them face difficulties very
often. It seems, however, that they did not pay as much attention to their collocational
errors. The majority of them claimed that they often (38.9%) or sometimes (38.9%)
pay attention to the errors they made. Only 19.4% did so very often. To sum up, the
students paid more attention to choosing correct collocations but less attention to their
collocational errors although they admitted that they had many difficulties in
translating collocations. One possible explanation is the fact that the teachers did not
always provide corrections for erroneous collocations found in students translation.
The student, without teachers corrections, had no idea about what caused the errors
or where to look for corrections, and thus might neglect them.

96
With respect to the strategies the respondents used to deal with difficulties in
translating collocations, the relevant data were presented below.

Table 4.14: Students strategies to deal with difficulties in translating collocations


Strategies Count Column N %
Using word-for-word translation 15 20.8%
Paraphrasing or finding other words to express the idea 46 63.9%
Using a synonym of the correct word 25 34.7%
Avoiding translating the difficult collocations if possible 12 16.7%
Checking dictionaries 50 69.4%
Others 2 2.8%
Total 72 100.0%

Table 4.14 shows that paraphrasing is among the frequently used strategies, with 46
of the respondents (63.9%) affirming their use of it. This conforms with the result of
the error analyses above, which suggested that approximation (paraphrasing) is among
the common causes of collocational errors. Word-for-word translation, on the
contrary, was said to be used much less frequently. While word-for-word translation
is considered a form of L1 transfer in this study, the use of it does not completely
reflect the influence of L1 transfer on learners, because there is a kind of unconscious
L1 transfer which cannot be investigated using the questionnaire. As seen in the error
analyses, L1 transfer was the primary cause of collocational errors; however, word-
for-word translation only accounted for a small part of the errors. Using synonyms
was a quite common strategies affirmed by one third of the respondents (34.7%). This
is also in line with the error analyses, which showed that the use of synonyms resulted
in a relatively large number of collocational errors. Only a few of the respondents
admitted abandoning their tasks when they could not produce the desired
combinations, which is quite contrary to what observed in the translation test above.
The most noticeable strategy, however, is the use of dictionaries, which was affirmed
by the majority of the respondents (69.4%). It has been suggested during the analyses
of errors that learners incorrect use of dictionaries might lead to collocational errors.
The fact that checking dictionaries was the students favorite strategy to deal with
collocation-related difficulties seems to support this proposition.

97
From teachers perspective
The teachers opinions about students learning and use of collocations were
investigated through questions from 4 to 7, which focused on four major issues: (1)
whether the students had difficulties in translating collocations, (2) how the teachers
teaching of collocations helped them in dealing with these difficulties, (3) what
strategies the students used when facing difficulties in translating collocations and (4)
what were the main causes of collocational errors in their translations.

All of the teachers being asked affirmed that their students, more or less, had
difficulties in translating collocations, which mainly resulted from their lack of
knowledge and experience. Due to their lack of knowledge and experience, students
tend to understand word combinations based on the literal meanings of its
components, T2 explained, for example they interpreted the word body in the
phrase a body of something (an amount of something) as the whole physical
structure of a human/ an animal, without realizing that in this combination the word
did not have its literal meaning. Unable to understand the correct meaning of a
combination, they could hardly provide accurate translations. Lack of knowledge and
experience also amplified negative influences of linguistic and cultural differences, T3
added.

However, the teachers believed that their teaching of collocations had positive effects
on the students, helping them discover a texts meaning more accurately and translate
it more easily. According to them, the teaching of collocation helps students have a
wide selection of words in different situations, and make effective use of synonyms
and lexical items to make their translation more natural and understandable.
Nevertheless, T1 remarked that the effectiveness of their teaching relied heavily on
the students hardworking and effort. Without these characteristics, they could hardly
improve themselves regardless of the teachers help.

Considering the strategies used by the students to deal with collocation-related


difficulties, the teachers suggested three major ones. The first strategy, according to
T2, was checking dictionary or similar sources. When facing a difficult combination,
students usually look for translations in dictionaries or on the Internet, she explained.
The problem is, however, that they tend to hastily use a translation they find, without

98
considering about the appropriateness of using it in a specific context. The second
strategy, as being suggested by the three teachers, was using word-for-word
translation. Some students tend to understand a combination based on the meaning of
each separate word, which usually leads to the use of this strategy. Last but not least,
in order to compensate for their lack of knowledge, some students also provide
random translations, with the hope that such translations could, by chance,
accurately convey the necessary messages.

With reference to the causes of collocational errors, all of the teachers agreed that
students lack of knowledge about collocation, culture and translation is among the
major reasons leading to collocational errors. Without necessary knowledge, students
could neither accurately discover a texts meaning nor produce appropriate
translations. Besides, according to the teachers, the use of the strategies above,
especially word-for-word translation, could also result in collocational errors. In
Vietnamese English translation, it is common for students to look for the English
equivalent of each individual word in a Vietnamese sentence before translating it,
regardless of the fact that not all sentences can be successfully translated that way,
T3 remarked. All of the teachers opinions above are in line with the results of the
error analyses, which pointed out that the aforementioned factors more or less
contributed to learners making of collocational errors.

In conclusion, the data collected from the teachers and the students perspectives
provide valuable information about the learning of collocation in Basic Translation
classes. Firstly, it can be seen that some of student respondents had relatively clear
perception of collocation and its role in translation; nevertheless, there were a
noticeable number of them who were not aware of collocation or did not fully
understand its importance. Moreover, the data suggested the student had been quite
passive in learning collocations. To be precise, they had gained their knowledge
mostly from teachers instructions and collocations in translation texts, rather than
actively searching other sources or doing extensive reading. On top of that, it can be
inferred that the students had rarely paid much attention to their collocational errors,
although learning from errors can be very effective. Such attitude and learning
methods may be the implying factors affecting their actual performance. Secondly,

99
the data seem to conform with the results from the analyses of errors, which suggested
that Approximation and L1 Transfer were the major causes of errors. Last but not least,
a large number of the student respondents affirmed that they had usually checked
dictionaries when facing difficult combinations, which, according to the teachers,
might lead to errors if the students were careless. This information is important as it
supports the aforementioned suggestion that a number of the collected errors possibly
resulted from the students incorrect use of dictionaries.

4.2. Major findings

The present study focused on analyzing errors related to adjective-noun collocations


made by the 3rd year English majored students in order to identify the underlying
reasons behind the errors and find appropriates remedies. The study, therefore, aimed
at answering two main research questions as follow:

1. What are the common errors related to attributive adjective-noun collocations


made by 3rd year English majored students at USSH?
2. What are the possible causes of these errors?

In order to answer these questions, different sources of data were collected, analyzed
and discussed, which altogether provided comprehensive insights into the issues. This
section is dedicated to recapitulating those major findings of the study.

Regarding the first research question, both of the error analyses showed that, among
the categories of errors, wrong choice of adjective was the broadest one, followed by
wrong combination and wrong choice of noun respectively, which indicated that the
student respondents tended to have difficulties in choosing the correct adjectives for
adjective-noun collocations. The explanation for this lies in the fact that in collocations
the choice of word for one component is usually restricted by the other. In the case of
adjective-noun collocations, it is the noun that restricts the choice of the adjective
because language users commonly produce a noun phrase by selecting the noun first.
Moreover, the sense of the adjective in an adjective-noun collocation is also
determined by the noun, and some words may have uncommon senses that only occur
in specific combinations. For example, the word green in green leaves and green
fingers has two different meanings simply because of the words it precedes, and while

100
green in the former combination has its literal sense, green in the latter one has an
idiomatic sense which occurs restrictively in that combination. Learners, in order to
produce an adjective-noun collocation, have to go through multiple process of
choosing words. The noun can be quite easily chosen if they know the words that
express the desired sense. However, after a noun is chosen, the choice of adjectives,
as explained above, will be restricted by that noun. Learners will need not only lexical
knowledge but also collocational knowledge to choose the accurate adjective. To put
it differently, even though they know the adjectives with the desired sense, they may
fail to pick the correct one if they do not know which of them can co-occur with the
chosen noun, not to mention the cases where the adjective has an uncommon sense.

The argument above explains why choosing correct adjectives for adjective-noun
collocation is the major challenge to the participants. However, it does not mean that
attention should be paid solely on adjectives. On the contrary, it implies the
significance of choosing nouns in producing adjective-noun collocations because,
after all, the noun is the factor determining the choice of adjectives. It can be inferred
that wrong choice of nouns usually leads to wrong choice of adjectives. Evidently, the
error analyses showed that combinations with both nouns and adjectives incorrectly
used occurred even more frequently than those with only nouns incorrectly chosen.

The purpose of identifying the common types of error in error analyses is to find out
the factors, in relation to the error types, which need attention. In the present study, it
can be seen from the arguments above that in order to produce desired adjective-noun
collocations, learners need to make accurate choices of nouns and adjectives, which
requires sufficient knowledge of both vocabulary and collocation. More importantly,
the fact that the participants tended to face difficulties in choosing correct adjectives,
as previously discussed, suggests the necessity to focus on broadening their
collocational knowledge, since it is commonly known that learning/teaching of
collocations usually receives less attention in comparison to general learning/teaching
of vocabulary. To sum up, by identifying the common types of error, the study points
out the aspects in learners knowledge that need focusing on. However, to find specific
ways to effectively broaden their knowledge about vocabulary and collocation
requires deeper insights into the causes of errors and other factors such as the actual

101
teaching and learning of collocations, which were addressed by the second research
question.

Regarding the causes of errors, the analyses yield the following major findings. Firstly,
L1 transfer was identified as the primary cause of error, which was not only in line
with previous studies but was also supported by the teachers opinion. L1 transfer, as
previously discussed, usually occurred when learners either resorted to word-for-word
translation or failed to distinguish between collocational equivalence and lexical
equivalence, which led to a false assumption that two words being equivalent in one
context would have the same relationship in others. The former case reflects learners
inexperience in translation, which was mentioned by the teachers as one of the factor
causing difficulties for the students. The latter case, which was more frequently
observed in the error analyses, on the other hand, probably signifies learners lack of
knowledge of collocation. Without the necessary knowledge to produce a desired
collocation, learners, as the teachers explained, tend to rely on their L1 knowledge and
try to attain collocation equivalence by using lexical equivalence. Especially when
translation is involved, as in the present study, the influence of L1 seems to be
amplified. Moreover, that L1 transfer can take place unconsciously makes it even more
problematic for learners as they cannot avoid what they are not aware of.

After L1 transfer, approximation was another major cause of collocational errors. It


is, in fact, a strategy that learners use to compensate for their lack of knowledge by
substituting the lexical items they are unable to produce with expressions that they
thought could convey the desired message. There are different types of approximation
depending on the way substitutions being made, which were observed in the present
study. The first was the type of approximation described by Dang (2014) as the act of
using a lexical item in place of the desired one, with which it shares some semantic
properties. This is the type of approximation mentioned in most previous studies. In
addition, the error analyses suggested another type of approximation in which
substitutions were made not basing on semantic properties that share between words
but on learners knowledge of the real world. To put it differently, they tried to express
the intended message by describing certain characteristics of the subject being referred
to. That approximation was a major cause of collocational errors, as the analyses

102
suggested, was not only in line with previous studies but was also supported by the
questionnaire to the student respondents, which showed that the majority of them had
been using approximation as a strategy to deal with collocation-related difficulties in
their translations. Besides, the second type of approximation discussed above can be
considered a new finding in this study as it has hardly been mentioned in any previous
one.

Along with L1 transfer and approximation, use of synonyms is a cause of errors that
has been frequently discussed in other studies. In the present study, it also accounted
for a considerable number of errors although it was not as prominent as the previous
two. The use of synonyms takes place due to learners false assumption that synonyms
are interchangeable in word combinations, which signifies their lack of knowledge
about collocation. If they had known that word combinations adhere to collocational
rules and thus synonyms cannot always replace each other, they would probably have
avoid making this kind of error.

Unlike in previous studies where it was usually neglected or included in other


categories, formal confusion was identified as a major cause of errors in the present
one. It involved the cases in which misused words were confused with the desired ones
due to formal similarities between them and the misplacements of parts of speech of
the desired words. As discussed in the error analyses, it indicates learners lack of
vocabulary knowledge. However, given the fact that it occurred more frequently when
learners had to deal with familiar combinations than with unfamiliar ones, formal
confusion may also result from deficiency in learners memories or pure carelessness.
Whatever the reason is, this cause of error is worth consideration as it accounted for
many errors that the student respondents, at their level, were supposed not to make.

Another cause of errors that also connected to form of words was word coinage.
However, unlike formal confusion word coinage involved the cases of deviant lexical
items which usually resulted from either misuse of affixes or misspelling. In the
present study, word coinage only accounted for a small number of errors; however,
similar to formal confusion, it also signifies the students lack of vocabulary
knowledge and deficiency in their memory.

103
All the aforementioned causes of errors belong to the list of hypothetical causes of
errors, which have been more or less mentioned in previous studies. Apart from them,
the present study also identified some other causes that have been rarely observed or
usually neglected.

The first one to be mentioned is the students incorrect use of dictionaries. In fact, the
idea that learners resort to dictionaries when dealing with difficult collocations is not
new as it was suggested by several authors, including Farrokh (2012). However, not
many studies mentioned it as a cause of collocational errors. In the present study,
several errors were identified with suggestive details that could be considered as
evidence for the students use of dictionaries. On top of that, the investigation into the
teaching and learning of collocations also supported this finding. More than half of the
student respondents affirmed their use of dictionary when translating collocations,
while the teachers remarked that some of their students had made mistakes due to their
careless use of dictionaries. Moreover, it is necessary to note that the identified errors
only reflected one aspect of the problem. There were probably other errors having
connection to the students use of dictionaries, which are not provable due to the limit
of the study. For example, upon checking bilingual dictionaries, learners found several
words that were equivalents of the target one. Unable to distinguish one from another,
they chose an equivalent which happened to be a synonym of the desired one and
created an erroneous combination. Obviously, the use of synonym would be identified
as the cause of this error and by which the use of dictionaries was neglected. Given
the fact that dictionaries are commonly used in translation, and since the present study
focuses on collocational errors in translation, learners use of dictionary, as discussed
above, is a factor that needs proper consideration.

The second one to be mentioned was a strategy observed through the translation test.
As previously described in the error analysis, some of the student respondents, unable
to produce the desired translations, seek solutions from seemingly suggestive details
in the test itself. This strategy involves the act of substituting a desired lexical item
with another one and thus can be considered a type of approximation. However, since
it was only an ad-hoc solution used by some of the respondents and was rarely seen
elsewhere, the researcher decided to mention it separately. Although this strategy

104
accounted for a limited number of errors, it provided valuable information regarding
what learners could do in compensation for their lack of knowledge.

In addition to the errors resulted from the aforementioned causes, there were other
errors whose causes were unidentifiable. It seems that these errors were the results of
the students effort to finish their tasks under the pressure of time and their lack of
knowledge. Unable to produce a desired combination, they might think of a
combination they were familiar with as substitutes, following teachers advice that no
blanks should be left in a translation test. However, when the students lacked
motivation they might also give up on their tasks as it was observed through the
translation test, which had no influence on their academic result, that a large number
of instances were left with no or incomplete answers. Therefore, motivation is also a
factor affecting their performance.

To sum up, through the error analyses, the present study has identified the causes of
collocational errors and also the underlying factors beneath them such as learners lack
of knowledge (both vocabulary and collocation), mental factors (carelessness, anxiety,
motivation, etc.), and time pressure, etc. While the causes of errors show how the
errors were made, those underlying factors give suggestions on how learners
performance can be improved. Some of them, such as anxiety or time pressure, seem
to be difficult to approach. Other mental factors (carelessness, motivation) are possibly
adjustable. Most importantly, learners knowledge, which is one of the determining
factors, is improvable through teaching and learning. However, in order to do so, a
thorough understanding of the current teaching/learning situation in the target classes
is essential.

The investigation into the teaching and learning of collocations in the Basic
Translation classes showed that the main methods the teachers had been using for
teaching collocations were integrated teaching and encouraging self-study. The first
method, as mentioned above, had its effectiveness determined by the availability of
collocations in translation tests and also by learners awareness of the activity.
Similarly, the second method heavily relied on the students awareness of collocation
and motivation. However, as the data suggested, only a part of the student respondents
recognized the teachers teaching of collocations. In addition, the learners were quite

105
passive in broadening their knowledge of collocation, and they did not fully
understand its importance either, not to mention a considerable number of them did
not even understand the concept of collocation. Those issues above may hinder the
effectiveness of the aforementioned teaching methods and thus negatively influence
the students knowledge. Regarding this, the error analyses suggested many flaws in
their productive ability. Moreover, the receptive test of collocations also showed the
learners limited receptive ability, which reflects their insufficient passive knowledge
of collocation. The insufficiency in passive knowledge obviously resulted in limited
active knowledge, which directly determined their productive performance.

Figure 4.4: Summary of the major findings

106
In brief, the study has pointed out the causes leading to learners making of
collocational errors. Most of them, according to authors including Dweik & Shakra
(2010) and Hussein (2011), are strategies employed by learners to deal with difficulties
in translating collocations. These difficulties are believed to result from several
underlying factors including insufficiency of knowledge, mental conditions or other
objective factors from the environment, among which the lack of knowledge is one
major factor that can be addressed by making changes to the teaching/learning process,
and thus corresponding recommendations on the teaching and learning of collocations
will be made in Chapter 5 for the improvement in students knowledge and the
remediation of errors. These major findings of the study are recapitulated in Figure 4.4
above.

107
CHAPTER 5
CONCLUSION AND RECOMMENDATIONS

This chapter serves as the concluding part of the current research by recapitulating the
major findings that give answers to the research questions. Moreover, based on these
findings, recommendations are made to teachers and students for improving the
teaching and learning of collocations in Basic Translation in particular and in
translation courses in general. Finally, given the limitation in the design of the study,
suggestions for future research are also presented in this chapter.

5.1. Conclusion

The researcher conducted the study which aimed at analyzing adjective-noun


collocational errors made by 3rd year students of EF, USSH in their translations, in
order to have a thorough insight into the nature of the errors and their causes. In
addition, thanks to the implementation of multiple research instruments, other factors
including the students receptive knowledge and the actual teaching and learning of
collocations in Basic Translation were also carefully investigated for a better
understanding of their collocational knowledge, and the underlying reasons hindering
their performance. In general, the analysis of errors showed that the students, despite
their expected high level of English competence, made quite a large number of errors,
which possibly resulted from several major causes. The study also revealed a
deficiency in the students collocational knowledge which was probably the result of
the teaching and learning process.

The data for error analysis in the present study were collected from two sources the
students midterm exam papers and the translation test. Their performances on both
the exam and the test were, in general, quite unsatisfactory. Precisely, nearly one third
of the adjective-noun combinations collected from the exam papers were errors. With
the translation test, the errors accounted for roughly 40 percent of the extracted
combinations. Among the types of error, wrong choice of adjective was the dominant
category, which was followed by wrong combination and wrong choice of noun
respectively. The result suggested that the choice of adjectives tends to cause more
difficulties for learners and thus needs special attention; however, as discussed above,

108
the choice of nouns is also very important as it is the noun that determines the choice
of adjectives. The causes of the errors were also identified, among which L1 transfer,
approximation and formal confusion were dominant ones, accounting for a large
portion of the errors. The study also revealed several emerging causes of errors besides
the hypothetical ones, among which the most noticeable was incorrect use of
dictionary. The errors and their causes signified several issues including the limitations
in the students productive knowledge of collocations and their lack of concentration
and/or motivation in translating/producing collocations.

Not only was the students productive knowledge of collocation insufficient but their
receptive knowledge, as reflected through the COLLEX test, was also limited, which
signified their lack of knowledge about collocations in general. The deficiency in
knowledge was probably the result of the teaching and learning process that lacked
focus on collocations. Although collocations were said to be taught in Basic
Translation, the teachers seemed to devote inadequate attention toward the actual
teaching of collocations. Particularly, instructions on collocations were mostly given
through integrated teaching, of which the effectiveness was influenced by the nature
of the tasks themselves. In addition, although the teachers did encourage the students
self-learning of collocations, they hardly provided any instruction for self-study,
neither did they monitor the process. Also, they did point out collocational errors in
students translation but sometimes did not give corrections, which resulted in the
students frequent neglect of these errors.

The teaching practices obviously influenced the students learning of collocations.


Particularly, some of the student respondents negated their teachers teaching of
collocations in Basic Translation, not to mention the ones who even did not understand
the term collocation. Without sufficient background knowledge and awareness of the
concept, the students could neither follow teachers instructions nor broaden their
knowledge about collocations through self-study. Besides, the research also revealed
some other factors, to which the students deficiency in collocational knowledge could
be partly attributed, such as the lack of motivation in studying and limitations in
vocabulary knowledge.

109
To sum up, the research suggested that although the students were supposed to be at
relatively high level of English proficiency, there were inevitably flaws in their
collocational knowledge which were attributed to several reasons including the
teaching and learning process. In the next sections, recommendations are given to
teachers and students in order to address the aforementioned issues.

5.2. Suggestions and recommendations

Error analyses were carried out to identify nature of errors and their causes, based on
which appropriate solutions could be found for the remediation of errors. In the present
study, the adjective-noun collocational errors might be attributed to several causes.
Most of them, however, originated from the same underlying factor: the teaching and
learning process. Therefore, in this section, recommendations on the teaching and
learning of collocations in Basic Translation in particular and translation courses in
general are formulated and given to both teachers and students.

5.2.1. Suggestions to teachers

Although all of the teachers being questioned affirmed their teaching of collocations
in Basic Translation, their teaching practices as the research revealed were not always
effective. The researcher, therefore, basing on the theoretical framework and the
authentic data, tentatively suggested some adjustments to the teaching of collocations
as follow.

Firstly, it is suggested that the teaching of collocations should be more explicit. While
the teachers did include collocations in their teaching program, most of their
instructions were integrated in other activities. Despite their explanation that
collocations were not a focus of courses like Basic Translation and thus could hardly
be taught explicitly, there has been an undeniable fact that collocations are an
important unit of translation and usually cause difficulties to translators, which was
also affirmed in the study. Besides, the teachers implicit instructions on collocations
proved to be insufficient to catch students attention as some of them negated such
instructions or were even unaware of the concept of collocation. Moreover, it has been
suggested in previous studies that explicit teaching of collocations was beneficial to
students. Given the reasons above, it is not only appropriate but also necessary for

110
collocations to be taught explicitly in Basic Translation and also in translations courses
in general.

Secondly, for the teaching of collocations to be effective, it is also important to raise


students awareness. As mentioned above, the study revealed a large portion of the
student respondents who were not aware of either the concept of collocation or
teachers relevant instructions. Consequently, they could hardly obtain any knowledge
about collocation from the teachers and neither could they broaden their knowledge
through self-study. Developing students awareness, therefore, is necessarily the very
first step in the teaching process. To achieve such a goal, it is advisable that teachers
first introduce the term collocation to the students, explain the concept and highlight
its role. According to Dang (2014), they can also use students native language as a
means to introduce the concept as it is a universal phenomenon existing in every
language. Explicit teaching of collocations also plays an important role in raising
students awareness. By giving explicit and formal instructions on collocations,
teachers can attract students attention and thus foster their awareness more easily.

In addition, it is necessary to introduce to students the fundamental concepts related


to collocations such as congruence and collocational equivalence, etc. as it is common
for students without knowledge about these concepts to form false assumptions that
easily lead to errors. For example, they may develop a tendency to seek collocation
equivalence through direct word-for-word translation or to use synonymous words as
substitutes for ones they do not remember. This issue was affirmed in the current study
as the analysis of errors showed that L1 transfer, substitutions (approximation), use of
synonyms, etc., which obviously resulted from the students lack of the
aforementioned knowledge, accounted for a large number of errors. Teachers,
therefore, need to draw students attention towards these issues in order to reduce their
probability of making errors. For example, they can use contrastive analysis to
highlight the differences between native and foreign languages, and explain the
concept of congruence/incongruence. Collocations in translation tasks can also be used
as examples to draw their attention to the fact that not all collocations are able to be
translated using word-for-word translation. Similarly, it is advisable to explain to them
that synonyms may not be used interchangeably in collocations as combination rules

111
are sometimes arbitrary and the meaning of one word may be restricted by the other
in a combination.

As discussed above, explicit instructions on collocations and related concepts in


translation would bring great benefits to students. However, according to the teachers
of Basic Translation, it was impossible to implement such formal teaching due to the
constraints of time. Within that limited time of the course, they had to deliver a great
amount of knowledge about translation to students, which allowed them to focus only
on certain issues that they found important. Given the difficulty above, it is necessary
to have a method that can make instructions on collocations more explicit and effective
and, at the same time, maintain the sufficiency of knowledge about translation
delivered to students within the time limit. Such a method incorporates some essential
steps including choosing of collocations and contexts, choosing suitable teaching
activities and giving appropriate corrections.

Choosing target collocations and contexts


The first step is choosing target collocations for teaching. Due to the limit of time,
teachers should pick collocations that are useful and necessary to teach to the students.
However, it is quite problematic since the degree of usefulness and necessity depends
on many factors including students level, their needs, courses objectives, etc.
Fortunately, Hodne (2009) provided a good solution to the issue by proposing criteria,
based on which teachers can decide which collocations are useful to teach to their
students. These criteria, which were already presented in Table 2.5, take into account
several essential factors including students needs, frequency of occurrence of the
collocations, probability for them to be used in communication, etc. To put it
differently, they should focus on collocations frequently used by native speakers,
collocations relevant to their purposes or future jobs (economy, tourism, journalist,
etc.) and those frequently occurring in communication. Problematic ones such as
incongruent collocations, collocations with synonymous words, collocations
involving cultural gaps, etc., as Bahns & Eldaw (1993), Hodne (2009) and Kurosaki
(2012) have suggested, should also be taught to students.

It is advisable to teach collocations in contexts as the contexts would help students in


discovering their meaning and usage. Moreover, teachers can also use the contexts to

112
organize other activities. The chosen contexts should be suitable for the students level
and contain the target collocations. Choosing such a text seems to be a challenge to
teachers. However, with the development of technology, it is much easier than before.
Teachers now have access to valuable reference sources, including corpora such as
COCA or BNC, where they can find out the collocations frequently used by native
speakers (i.e. those with high FREQ) and also contexts in which they occur. These
contexts can be chosen for translation tasks or other activities. Especially, with courses
as Basic Translation where most translation tasks are given in the form of sentences
or groups of sentences, corpora are even more useful as they can provide hundreds of
meaningful sentences containing the target collocations for teachers to choose from.
They can also find texts from other sources through search engines such as Google.

Activities for teaching collocations


With aforementioned requirements, the activities used to teach collocations in
translation need to give students chances to practice translation and at the same time
broaden their collocational knowledge. The integrated teaching of collocations
through translation tasks that the teachers have been using, in fact, is a very good
method as students can learn new collocations while translating the texts. The problem
of this method, however, is that its effectiveness, as previously discussed, relies on
whether the texts contain enough useful collocations for them to learn. This problem,
nevertheless, can be solved by careful selection of target collocations and contexts.
Moreover, teachers need to give clear instructions to ensure students awareness of the
target collocations in the texts.

In addition to the integrated teaching method mentioned above, teachers can also
design activities that are more collocation-oriented but also incorporate teaching
translation. Take the translation task introduced by Kurosaki (2012), which was
adapted as one of the research instruments in this study, as an example, although it
mainly focuses on collocations, it can also be used in teaching translation. By asking
students to translate target collocations in given contexts, the task not only provides
them with knowledge about the target collocations but also draws their attention to the
fact that sometimes instead of translating words separately they need to translate
phrases or group of words as a whole unit of translation. If teachers want the task to

113
be more challenging, they can ask students to translate the whole sentences rather than
just the target collocations. However, they will need to draw students attention to the
target collocations when giving correction.

A similar activity was introduced by Lewis (2000), in which students are asked to
identify collocations in texts following teachers instructions and later to translate
these collocations into their L1. This activity, according to Lewis (2000), will help
students become more aware of collocation and less inclined to translating word-for-
word (p. 103). It can also be adapted as a method for self-study. Teachers can assign
reading texts to groups of students and ask them to identify in the texts collocations of
a certain type (e.g. adjective-noun, verb-noun, etc.) and provide their own translations
of these collocations. Teachers may need to instruct students in using collocation
dictionaries or corpora to help them with their assignment. The groups collections of
collocations, later, will be corrected by teachers and used as a glossary for the students.
By organizing self-study in such a way, teachers of translation can improve their
students knowledge of collocations and also save time for other important in-class
activities.

Regarding the concepts related to collocations, they can be introduced to students


through the following exercises suggested by Lewis (1997), in which students are
asked either to find the antonyms of certain combinations or to produce correct
collocation from a given list of words (see Appendix 5). These exercises can be used
to show students that the meaning of a word depends on the word it combines with.
Teachers can also ask them to translate the collocations into Vietnamese to highlight
the fact that a word in English may have different equivalents in Vietnamese and thus
they need to be careful in choosing the correct equivalent or using word-for-word
translation. These exercises can also be adapted to illustrate the risk of using synonyms
in translating collocations by asking students to form correct collocations from several
nouns and a group of synonymous adjectives and translate them to their native
language. They will realize that synonyms cannot be used interchangeably in certain
collocation and that their synonymous relationship also depends on the combination
in which they occur.

114
In addition, the translation task proposed by Kurosaki (2012) mentioned above can be
employed for the same purpose. By purposely choosing collocations that stimulate the
use of synonyms (e.g. gi mnh with mnh means either strong or powerful in English)
or ones that have the same component but will be translated differently into the other
language (e.g. tr long and sp long with long being translated as weak and thin
respectively), etc., teachers can also raise students awareness of the aforementioned
issues. The strength of this exercise is that it provides contexts for the target
collocations and can be used with both L1 L2 translation and vice versa.

Giving corrections
One important step in teaching is giving corrections. By doing so teachers can help
students recognize their weakness and learn from their errors. As a matter of fact,
identifying and correcting collocational errors is also an effective way to teach
collocations. The study showed that although the teacher participants pointed out the
collocational errors in their students translations, they did not always give corrections.
Moreover, it is commonly seen that teachers tend to focus more on grammatical
collocational errors than on lexical ones, due to the fact that the former is more easily
identified. Therefore, it is suggested that teachers pay more attention to identifying
and correcting lexical collocational errors in students translations. To reduce the
amount of time needed for errors correction, teachers can choose to focus only on
typical errors. They can also assign groups of students to identify and correct errors
by themselves. To sum up, correction of collocational errors, especially lexical ones,
in students translation should be promoted as it is also a way to make the teaching of
collocations in translation more explicitly and raise students awareness of the issue.

In addition to the adjustments to the teaching practices, there are some other issues
that also need teachers attention as they might possibly affect students performance.
The first one to be mentioned is students use of dictionaries. It is an undeniable fact
that dictionaries are essential in translation. However, incorrect use of such a useful
tool may lead to serious errors. As suggested by the analysis of errors, the students
careless use of bilingual dictionaries resulted in several errors that should not have
been committed if they had been more careful, and possibly accounted for errors in
other categories. Although the number of errors was not large, it emphasized the need

115
to pay more attention to their use of dictionaries. Particularly, teachers should remind
them that bilingual dictionaries are useful but should be used cautiously because a
word in English may have multiple Vietnamese equivalents and vice versa. The use of
monolingual dictionaries, therefore, needs to be highlighted, especially when students
are confused by bilingual ones. More importantly teachers should encourage the use
of collocation dictionaries as it will bring more benefits to students. Linguistic corpora
are also a good alternative to the dictionaries as they provide free access to much larger
database and many more functions that students can make use of to broaden their
knowledge. However, as corpora are quite complicated, teachers may need to give
some instructions as to how to use them.

The last issue is that teachers themselves also need to frequently update their
knowledge about collocation as languages are changing continuously and their
knowledge might be soon outdated. Again, linguistic corpora with frequently updated
database such as COCA seem to be a more reliable and comprehensive source than
collocation dictionaries for teachers to improve their own knowledge and also to
prepare lessons on collocations.

5.2.2. Suggestions to students

Based on the findings of the study, it is suggested that the students should put more
effort in improving their knowledge of collocations. Firstly, they need to reinforce
their understanding of the term collocation and related concepts, and their roles in
communication as well as translation. The knowledge would raise their awareness of
collocations and possibly reduce their probability of committing errors related to them.
However, understanding the term and its related concepts is not enough. It is advisable
for students to have profound knowledge of how words combine together, which
would help them in recognizing and producing collocations, especially lexical ones,
accurately. These kinds of knowledge are obtainable from different sources and
activities including teachers instructions and self-study. Therefore, students should
pay more attention to and follow their teachers instructions. More importantly,
however, they should think seriously about self-study. As it is obvious that teacher can
never introduce everything about collocations to them, they need to broaden their
knowledge by themselves basing on teachers instructions. For example, they can read

116
material on collocations. Extensive reading is also useful, through which they are able
to improve their reading comprehension skill, and learn new vocabulary and
collocations included in the texts. Besides, the use of collocation dictionaries and
corpora in translation is recommended. Students can also base on these sources to
compile their own glossaries of collocations.

The second recommendation is on students learning of vocabulary, which is believed


to have positive effects on collocations acquisition and vice versa. On one hand, it is
obvious that one without a wide vocabulary would have many difficulties in
recognizing and producing collocations. On the other hand, effective learning of
vocabulary requires collocational knowledge. Learning words as chunks, according to
LA, would enhance their acquisition and retention. Moreover, as Firth remarked that
one shall know a word by the company it keeps (cited in Gyllstad, 2007, p.7),
learning words in combination with others would also help students discover their
meaning and usage more easily and accurately. Learning vocabulary in such a way, in
return, can improve their knowledge about collocations. While learning chunks of
words they may consciously or unconsciously memorize the ways the words combine,
which will eventually be helpful for them when they need to produce certain
collocations. To sum up, students should broaden their vocabulary by learning words
as chunks as it is an effective way to improve both their vocabulary and collocational
knowledge.

The last recommendation aims at their translation skills. Having a broad vocabulary
and profound knowledge of collocations is necessary but not sufficient to ensure a
successful translation since the knowledge, after all, will be meaningless if students
do not have the essential skills to put it into practice. Therefore, it is also recommended
that students improve their skills and knowledge in translation. Particularly they
should be aware that there are different translation methods other than word-for-word
translation, which they need to practice using more frequently. Moreover, knowledge
about vocabulary and collocations will not become active knowledge until they
actually use it, such as in translation. In sum, students are recommended to broaden
their knowledge and put it into practice with different translation methods. Also, it is
a good way to strengthen their memory.

117
5.2.3. Recommendations for future research

The present study focused on collocational errors in translation. However, due to the
limits in time and scope of the study, it only took into account one type of lexical
collocation: adjective-noun. Therefore, a study on larger scale that includes all other
types of lexical collocations can be conducted for a more thorough insight into the
issue. Also, this study should be carried out with a larger size of sample for stronger
generalizability.

One other point that needs further research is the influence of learners use of
dictionaries on their making of errors. The present study revealed that inappropriate
use of dictionaries accounted for some errors, which was also mentioned in several
previous studies. However, due to the lack of evidence, it was impossible to further
elaborate on the issue. Therefore, future studies can focus more on learners use of
dictionaries in translations in order to investigate this issue more thoroughly.

Last but not least, the present study made some recommendations on how collocations
could be taught and learnt in translation. However, these recommendations were
produced mostly on theoretical basis and thus still need further empirical studies to
confirm their effectiveness and convince teachers of the necessity of teaching
collocations in translation.

118
REFERENCES

Alsakran, R. A. (2011). The productive and receptive knowledge of collocations by


advanced Arabic-speaking ESL/EFL learners. Fort Collins: Colorado State
University.

Bahns, J., & Eldaw, M. (1993). Should we teach EFL students collocations? System,
21(1), 101-114. Retrieved May 15, 2015 from http://sci-hub.cc/10.1016/0346-
251X(93)90010-E.

Benson, M., Benson, E., & Ilson, R. (1986). The BBI combinatory dictionary of
English: A guide to word combinations. Amsterdam: John Benjamins.

Brown, H.D. (1980). Principles of language learning and teaching. New Jersey:
Prentice-Hall Inc.

Channell, J. (1981). Applying semantic theory to vocabulary teaching. ELT Journal,


35, 115-122. Oxford: Oxford University Press. Retrieved May 16, 2015 from
http://sci-hub.cc/10.1093/elt/XXXV.2.115.

Church, K., & Hanks, P. (1990). Word association norms, mutual information, and
lexicography. Computational Linguistics, 16(1), 22-29. Retrieved July 15,
2015 from http://www.aclweb.org/anthology/J90-1003.

Clear, J. (1995). T-Score and mutual information score from Birmingham Corpus
website. Retrieved August 5, 2015 from
http://lingua.mtsu.edu/chinesecomputing/docs/tscore.html.

Corder, S. P. (1967). The significance of learners errors. International Review of


Applied Linguistics in Language Teaching, 3(4), 161-170. Retrieved May 4,
2015 from: http://libgen.org/scimag/get.php?doi=10.1515%2Firal.1967.5.1-
4.161.

Corder, S. P. (1974). Error analysis. In J. P. Allen, & S. P. Corder (Eds), The


Edinburgh Course in Applied Linguistics Volume 3: Techniques in applied
linguistics. Oxford: Oxford University Press.

119
Cowie, A. P. (1981). The treatment of collocations and idioms in learners
dictionaries. Applied Linguistics, 2(3), 223-235. Retrieved July 15, 2015 from
http://sci-hub.cc/10.1093/applin/ii.3.223

Cruse, A. (1986). Lexical semantics. Cambridge: Cambridge University Press.

Dang, Thi Van Di. (2014). A survey into the teaching of lexical collocations in
Academic Writing at Faculty of English Linguistics and Literature
University of Social Sciences and Humanities Ho Chi Minh City. M.A.
Thesis. Ho Chi Minh: University of Social Sciences and Humanities.

Darvishi, S. (2011). The investigation of collocational errors in university students


writing majoring in English. IPEDR, (18). Singapore: IACSIT Press.

Dweik, B. S., & Shakra, M. M. A. (2010). Strategies in translating collocations in


religious texts from Arabic into English. Retrieved July 15, 2015 from
http://eric.ed.gov/?id=ED510365

Ellis, R. (2003). Second language acquisition (8th Ed.). Oxford: Oxford University
Press

Fan, M. (2009). An exploratory study of collocational use by ESL students A task-


based approach. System, 37(1), 110-123. Retrieved January 22, 2015 from
http://davies-linguistics.byu.edu/ling485/for_class/teaching/fan.pdf.

Farghal, M., & Obiedat, H. (1995). Collocations: A neglected variable in EFL. IRAL,
33(4), 315-333.

Farrokh, P. (2012). Raising awareness of collocation in ESL/EFL classrooms.


Journal of Studies in Education, 2(3), 55-74.

Firt, J. R. (1957). Paper in Linguistics 1934-1951. London: Oxford University Press.

Fontenelle, T. (1994). What on Earth are collocations? An assessment of the ways in


which certain words co-occur and others do not. English Today, 40, 10(4).
Retrieved December 14, 2014 from http://www.ugr.es/~inped/
exploringnewpaths/collocations/What_on_Earth_are_Collocations.pdf.

120
Gass, S. M., & Selinker, L. (2008). Second language acquisition: An introductory
course (3rd Ed.). New York: Routledge.

Gyllstad, H. (2007). Testing English collocations: Developing receptive tests for use
with advanced Swedish learners. Lund University. Retrieved April 28, 2015
from http://lup.lub.lu.se/luur/download?func=downloadFile&recordOId
=599011&fileOId=2172422.

Harmer, J. (2001). The practice of English language teaching (3rd ed.). London:
Longman.

Hodne, S. L. (2009). Collocations and teaching: Investigating word combinations in


two English textbooks for Norwegian upper secondary school students.
Doctoral Dissertation. Toronto: Toronto University

Hong, A. L., Rahim, H. A., Hua, T. K., & Salehuddin, K. (2011). Collocations in
Malaysian English learners writing: A corpus-based error analysis. 3L: The
Southeast Asian Journal of English Studies, 17 (Special Issue), 33-44.

Howarth, P. (1998). The phraseology of learners' academic writing. In Cowie, P.


(Ed.). Phraseology: Theory, Analysis and Applications, Oxford: Oxford
University Press.

Hussein, A. F. (2011). Translatability of collocations: A constant challenge to EFL


learners. Jordan Journal of Educational Sciences, 7(2), 209-219.

Ibarhim, Y. (2003). The translation of collocation into Arabic: Problems and


solutions. Retrieved March 26, 2015 from http://etheses.whiterose.ac.uk
/4402/1/uk_bl_ethos_569582.pdf.

Jafarpour, A. A., Hashemian, M., & Alipour, S. (2013). A corpus-based approach


toward teaching collocation of synonyms. Theory and Practice in Language
Studies, 3(1), 51-60.

James, C. (1998). Errors in language learning and use: Exploring error analysis.
London: Longman.

121
Jiang, N., & Nekrasova, T. (2007) The processing of formulaic sequences by second
language speakers. The Modern Language Journal, 91(3), 433-445.

Johansson, S. (2008). Contrastive analysis and learner language: A corpus-based


approach. Retrieved November 11, 2014 from http://www.hf.uio.no/ilos/
forskning/grupper/Corpus_Linguistics_Group/papers/contrastive-analysis-
and-learner-language_learner-language-part.pdf.

Kim, D. H. (2009). A study on the use of lexical collocations of Korean Heritage


learners: Identifying the Sources of Errors. University of Southern
California. Retrieved December 17, 2015 from http://digitallibrary.usc.edu/
cdm/ref/collection/p15799coll127/id/206828.

Koya, T. (2005). The acquisition of basic collocations by Japanese learners of


English. Doctoral Dissertation, Waseda University. Retrieved October 11,
2014 from http://www.cs.waikato.ac.nz/~shaoqun/tmp/Honbun-4160.pdf.

Kurosaki, S. (2012). An analysis of the knowledge and use of English collocations by


French and Japanese learners. Retrieved September 2, 2014 from
http://digirep.rhul.ac.uk/file/d65f4c4b-8bac-d738-3735-83b5e5e031a7/7/
Shino_K._Thesis_20121015.pdf.

Lewis, M. (1997). Implementing the Lexical Approach: Putting theory into practice.
United Kingdom: Heinle ELT.

Lewis, M. (2000). Further developments in the Lexical Approach. Hove: Language


Teaching Publications.

Li, C. C. (2005). A study of collocational error types in ESL/EFL college learners


writing. M.A. Thesis. Taiwan: Ming Chuan University.

Lubis, S. (2013). Collocation as source of translation unacceptabilty: Indonesian


students experiences. International Journal of English Linguistics, 3(5), 20-
28. Retrieved September 2, 2014 from http://www.ccsenet.org/journal/
index.php/ijel/article/view/30640.

Luu, Trong Tuan. (2012). An introduction into translation theories. Ho Chi Minh:
National University Publishing House.

122
Mai, Lan. (2010). Insight into students use of lexical collocation in Vietnamese-
English translation. Hanoi: Hanoi University.

Maurer-Stroh, P. (2005). "House-high favourites?" A contrastive analysis of


adjective-noun collocations in German and English. Elope, 2, 57-64.
Retrieved October 15, 2014 from http://www.sdas.edus.si/Elope/PDF/
ElopeVol2MaurerStroh.pdf.

McCarten, J. (2007). Teaching vocabulary: Lessons from the corpus, Lessons for the
classroom. New York: Cambridge University Press.

McIntosh, C., Francis, B., & Poole, R. (Eds). (2009). Oxford collocations dictionary
for students of English (2nd Ed.). Shanghai: Oxford University Press.

Meidasari, V. E. (2007). Strategies of Translating Collocation. Majalah Faktor,


September-October, 4-13.

Munday, J. (2001). Introducing translation studies: Theories and applications. New


York, NY: Routledge.

Nakata, T. (2007). English collocation learning through meaning-focused and form-


focused activities: Interactions of Activity Type and L1-L2 Congruence.
Retrieved December 20, 2014 from http://paaljapan.org/resources/
proceedings/PAAL11/pdfs/13.pdf.

Nation, I.S.P. (2001). Learning vocabulary in another language. New York:


Cambridge University Press.

Nesselhauf, N. (2003). The use of collocations by advanced learners of English and


some implications for teaching. Applied Linguistics, 24(2), 223-242. Oxford
University Press. Retrieved September 2, 2015 from
http://www.corpus4u.org/forum/upload/forum/2005062000435073.pdf.

Nesselhauf, N. (2005). Collocations in a learner corpus (Vol. Studies in Corpus


Linguistics). Amsterdam: John Benjamins Publishing.

Newmark, P. (1988). A textbook of translation. New York: Prentice-Hall


International

123
Nguyen, Hien Luong. (2006). Importance of collocational knowledge. Hanoi:
University of Foreign Studies.

Nguyen, Le Kim Phung. (2014). The collocational competence of advanced learners


in recognising and producing verb-noun combinations: A Study on 3rd year
honor students. B.A. Thesis. Ho Chi Minh: University of Social Sciences and
Humanities.

Pawley, A., & Syder, F. H. (1983). Two puzzles for linguistic theory: Nativelike
selection and nativelike fluency. In J.C. Richards, & R.W. Schmidt (Eds.),
Language and Communication (pp.191-226). London: Longman.

Pirmoradian, M., & Tabatabaei, O. (2012). The enhancement of lexical collocation


learning through concordancing: A case of Iranian EFL learners. MJAL, 4(4).

Rahimi, A., & Sahragrad, R. (2008). Vocabulary learning can be fun. The Asian EFL
Journal, 14(2), 63-89. Retrieved July 12, 2015 from
https://www.researchgate.net/publication/273775801_vocabulary_learning_c
an_be_fun.

Richards, J. C. (1974). A non-contrastive approach to error analysis. In J. C.


Richards (Ed.), Error analysis: Perspectives on second language acquisition
(pp. 172-188). London: Longman.

Richards, J. C., & Rodgers, T. S. (2001). Approaches and methods in language


teaching. Cambridge: Cambridge University Press.

Richards, J. C., & Schmidt, R. W. (2002). Longman dictionary of language teaching


and applied linguistics. New York: Longman.

Shammas, N. A. (2013). Collocation in English: Comprehension and Use by MA


Students at Arab Universities. International Journal of Humanities and Social
Science, 3(9), 107-122.

Shehata, K. A. (2008). L1 influence on the reception and production of collocations


by advanced ESL/EFL Arabic learners of English. M.A. Thesis. Ohio:
University of Ohio.

124
Sinclair, J. (1991). Corpus, concordance, collocation. Oxford: Oxford University
Press.

Sughair, Y. A. (2011). The translation of lexical collocations in literary texts. M.A.


Thesis. UAE: American University of Sharjah.

Tarone, E. (1981). Some thoughts on the notion of communication strategy. TESOL


Quarterly. Vol. 15, 285-295.

Trinh, N. F. (2001). English and Vietnamese collocations: A contrastive analysis.


Doctoral Dissertation. Australia: Macquarie University.

Yamashita, J. & Jiang, N. (2010). L1 Influence on the Acquisition of L2


Collocations: Japanese ESL Users and EFL Learners Acquiring English
Collocations. TESOL Quarterly, 44, No.4, 647-668.

125
APPENDIX 1
TRANSLATION TEST

Name: ______________________________________________________________

Class: ______________________________________________________________

INSTRUCTION: Fill in the blanks by translating the underlined words in the Vietnamese
sentences into English. For each blank, you will need two words (one Adjective and one
Noun). If you do not like to fill in the blanks, you can provide complete translations of your
own.

1. Karlow mt mt chn trong chin tranh v c lp mt ci chn gi mi.


Karlow lost his leg in the war and was fitted for a new __________________.

.........................................................................................................................................................

2. ng ta mt ton b rng v phi mang rng gi mi c th n c.


He became totally toothless and had to wear __________________ to eat.

.........................................................................................................................................................

3. Nu mun sp c, hy thm bt vo ni sp.


If you want __________________, add some flour to the stock.

.........................................................................................................................................................

4. Cho sp ra mt ci t ln, v thm nc vo nu ba n mun c sp long.


Transfer soup to a large bowl, and add more water if you want __________________.

.........................................................................................................................................................

5. Tr long l loi tr c mu v v u nht.


"__________________" means the tea that is light in color and taste.

.........................................................................................................................................................

6. Sa c l sa b c loi b nc.
__________________ is cow's milk from which water has been removed.

.........................................................................................................................................................

126
7. Ngi n ng c tm thy trong tin
h tra ng mui gay, m t b m, va nhi u v t thng
u.
The man was found with a broken nose, __________________, and wounds on his head.

.........................................................................................................................................................

8. U ng qua nhi u nc ngo t, ch ng ha n nh Coca Cola, khng t t cho sc khoe cua ba n.


Drinking too much __________________, such as Coca Cola, is not good for your health.

.........................................................................................................................................................

9. Neptune l mt nhn hiu du n ni ting.


Neptune is a famous brand of __________________.

.........................................................................................................................................................

10. Dn s ng va thi u t la nhng nguyn nhn ti m n gy nn xung t.


__________________ and scarce land resources were the underlying sources of conflict.

.........................................................................................................................................................

11. C nang mua m t t m bua may m n t ng kh v n en cua min


h i.
She bought a lucky charm in order to get rid of her __________________.

.........................................................................................................................................................

12. V co cht lng km, nn chic xe c gi rt r.


Because of its __________________, the car is very cheap.

.........................................................................................................................................................

13. Gi ln (ln n 60 dm mt gi) co kha nng gy thit ha i v tai san.


__________________ (up to 60mph) have the potential to cause property damage.

.........................................................................................................................................................

14. Gh in tr thnh biu tng cho n t hnh M.


The electric chair has become a symbol of the __________________ in the United States.

.........................................................................................................................................................

15. Mt trng bin mt, v c mt cn gi nh ht nh hi th ca ngi ang hp hi.


The moon disappeared, and there was a(n) __________________, like the breath of a dying man.

.........................................................................................................................................................

127
16. Ht th l iu kin cn nhng khng phi l iu kin bn tn ti.
Breathing is a necessary but not a(n) __________________ of your remaining alive.

.........................................................................................................................................................

17. Lise khng phi l mt u bp c o to qua trng lp, nhng c c nhiu kinh
nghim qu bu bt ngun t vic gip m trong mt khch sn nh Copenhagen.
Lise is not a trained cook, but has a wealth of __________________ which began with helping
her mother in a small hotel in Copenhagen.

.........................................................................................................................................................

.........................................................................................................................................................

18. D vn cn tr, ng y vn phi ngh hu v sc khe km.


Although he was still young, he had to retired due to __________________.

.........................................................................................................................................................

19. Chnh sch mi mang li nhng iu kin thun li cho pht trin du lch.
The new policy provides __________________ for the development of tourism.

.........................................................................................................................................................

20. c mi vo Nh Trng v gp tng thng M qu l mt vinh d ln lao.


It was a __________________ to be invited to the White House and to meet the president of the
United States.

.........................................................................................................................................................

.........................................................................................................................................................

Thank you for your cooperation!

128
APPENDIX 2
COLLEX TEST
Name: ______________________________________________________________

Class: ______________________________________________________________

INSTRUCTION: For each pair of word combinations below, choose the one that, in your
opinion, is the correct collocation (frequently used by native speakers). In some cases, both
combinations given are correct.

1. A. Thick soup B. Dense soup C. Both are correct


2. A. Thin soup B. Weak soup C. Both are correct
3. A. False teeth B. Artificial teeth C. Both are correct
4. A. Artificial limb B. Fake limb C. Both are correct
5. A. Weak tea B. Washy tea C. Both are correct
6. A. Thick milk B. Condensed milk C. Both are correct
7. A. Cooking oil B. Edible oil C. Both are correct
8. A. Dense population B. Crowded population C. Both are correct
9. A. Soft drink B. Sugared drink C. Both are correct
10. A. Black eyes B. Bruised eyes C. Both are correct
11. A. Low quality B. Bad quality C. Both are correct
12. A. Powerful wind B. Strong wind B. Both are correct
13. A. Bad fortune B. Black fortune C. Both are correct
14. A. Capital punishment B. Dead punishment C. Both are correct
15. A. Weak wind B. Light wind C. Both are correct
16. A. Valuable experience B. Precious experience C. Both are correct
17. A. Poor health B. Weak health C. Both are correct
18. A. Ample condition B. Sufficient condition C. Both are correct
19. A. Advantageous condition B. Favorable condition C. Both are correct
20. A. Big honor B. Great honor C. Both are correct

Thank you for your cooperation!

129
APPENDIX 3
QUESTIONS FOR INTERVIEWING TEACHERS

A. Perception of collocation
1. Do you know about collocation? (types, uses, etc.)
2. What do you think about the role of collocation in translation and translation teaching?
B. Teaching of collocation
3. Do you teach (lexical) collocations in your class of Basic Translation?
If the answer is Yes
a. How do you teach collocations?
b. Do you have any problems in teaching collocations?
If the answer is No
a. Why do you not teach collocations in your class of Basic Translation?
b. Do you think that the teaching of collocations should be required in translation
classes?
C. Observation of Ss performance
4. Do the students have any difficulties/problems in translating collocations? (What are
they?)
5. How does your teaching of collocations help the students in dealing with the
difficulties/problems? (Only ask if the answer to Q3 is Yes)
6. What are the strategies the students usually use when facing problems in translating
collocations?
7. In your opinion, what are the main causes of collocational errors in the students
translations?

130
APPENDIX 4
QUESTIONNAIRE

Name: _________________________________________________________________

Class: __________________________________________________________________

INSTRUCTION: For each response, please circle or specify the choice(s) that best
corresponds to your view.
A. Perception of collocations
1. Do you know what a collocation is?
a. Yes
b. No

If your answer is No please stop the questionnaire here. Thank you for your time and
cooperation!

2. How do you know the concept of collocation?


a. Studied in universitys compulsory curriculum
b. Introduced by teachers out of compulsory curriculum
c. Reading reference materials on this issue
d. Others: (please specify) _______________________________________________
__________________________________________________________________
3. Do you think that knowledge about collocation is important in translation?
a. Yes
b. No (Please move to Question 5)

4. What does knowledge about collocation help you in translation process? (You can
choose more than one option)
a. Increase written fluency, which allows more time to discover the source texts
meanings
b. Providing alternative and richer ways of expression
c. Producing more precise and more natural language
d. Maintaining and strengthening cohesion of the text
e. Others: (please specify) ___________________________________________
_______________________________________________________________

131
B. Acquisition and use of collocations
5. What do you often do to broaden your knowledge about collocations? (You can
choose more than one option)
a. Reading reference materials
b. Paying attention to collocations used in texts (i.e. reading texts, translation texts,
etc.)
c. Memorizing groups of words that often go together when learning vocabulary
d. Looking up for collocations in dictionaries
e. Using linguistic corpora
f. Others (please specify) ______________________________________________
________________________________________________________________

6. How often do you care about choosing right word combinations during your
translations?
a. Very often
b. Often
c. Sometimes
d. Rarely
e. Never

7. How often do you face the problem of selecting the right collocations during
your translation?
a. Very often
b. Often
c. Sometimes
d. Rarely
e. Never

8. How often are you aware of erroneous combinations in your translations?


a. Very often
b. Often
c. Sometimes
d. Rarely
e. Never

132
9. What do you often do when facing difficulties in translating collocations? (You
can choose more than one option)
a. Using word-for-word translation
b. Trying to paraphrase or find other words to express the idea
c. Using a synonym of the correct word which you do not remember
d. Avoiding translating the difficult collocations if possible
e. Checking dictionaries
f. Other (please specify) ____________________________________________
______________________________________________________________

C. Teaching of collocations
10. Did your teacher of Basic Translation teach collocations in class?
a. Yes
b. No (Please finish the questionnaire here. Thank for your time and cooperation!)

11. How did (s)he teach collocations? (You can choose more than one option.)
a. Introducing the term collocation
b. Introducing the role of collocation
c. Directing your attention to collocations in translation tasks/texts
d. Choosing some collocations to teach/introduce in class
e. Encouraging you to collect and study collocation on your own
f. Mentioning already-taught collocations for a few times
g. Using collocation dictionaries
h. Using linguistic corpora
i. Identifying your errors and ask you to look for corrections
j. Identifying your errors and give corrections in your translations.
k. Others: (please specify) ____________________________________________
_______________________________________________________________

Thank you for your cooperation!

133
APPENDIX 5
ACTIVITESS FOR TEACHING COLLOCATIONS
(Lewis, 2000)

Type 1:

Fill in column 2 with an adjective which is opposite in meaning to the adjective in


Column 1 and also makes a correct collocation with the word in Column 3.
Column 1 Column 2 Column 3
helpful __________________ suggestion
efficient __________________ system
careful __________________ piece of work
safe __________________ choice

light __________________ green


light __________________ suitcase
light __________________ rain
light __________________ work

Type 2:

One word in each group does not make a strong word partnership with the word in
capitals. Which is the odd one?
1. BRIGHT idea green smell child day room
2. CLEAR attitude need instructions alternative day conscience road
3. LIGHT traffic work day entertainment suitcase rain green lunch
4. NEW experience job food potatoes baby situation year
5. HIGH season price opinion spirits house time priority
6. MAIN point reason effect entrance speed road meal course
7. STRONG possibility doubt smell influence views coffee language
8. SERIOUS advantage situation relationship illness crime matter

134
APPENDIX 6
SAMPLES OF COLLOCATIONAL ERRORS EXTRACTED FROM
STUDENTS EXAM PAPERS IN BASIC TRANSLATION

Source text Students translations Suggested corrections


Approximation
Vic ph rng, ph hy Deforestation, ozone damaging or ozone (layer)
tng ozone hay thi cc radioactive release result in global destruction
cht phng x ra bin environmental pollution.
u gy ra nhim mi Global pollution is caused by forest
trng ton cu. destruction, ozone layer ruin and radiation
rejection to ocean.
B tc ng mnh bi Be strongly influenced by pollutants and fishing methods
nn nhim v cc destructive fish-catching methods
phng thc nh bt Being strong impacted by pollution and
c mang tnh hy dit destructed means of catching fishes
Being strongly affected by the pollution
and some fish-hunting methods...
Nhn ti l nguyn kh Talent people are life-sustaining element - the (source of) power
quc gia of a country - the essence
Talented man is a precious core of a
country
rt nhiu bi many stretch of aqueous rock and underwater coral reefs
ngm san h b ph coral have been destroyed. They are not
hy, khng cn l mi the habitat for fish anymore.
trng sng cho c na.
Thomas Edison tin rng Thomas Edison believed that his deaf concentration
bnh ic ca ng disease had strengthen his focus ability.
lm tng thm sc tp Thomas Edison believes that his deaf
trung ca ng y. increased his focusing ability.
Thi Lan l nc duy Thailand is the only country that has never - powerful European
nht ng Nam been governed by any super western countries
cha bao gi b cai tr countries in Eastern Asia. - European superpowers
bi mt cng quc u
Chu.
Nh cc dc phm Thanks to the modern pharmacy, modern medicine
hin i, bnh lao v tuberculosis and malaria are not as
bnh st rt khng cn dangerous as before.
nguy him nh trc Thanks to modern chemical, tuberculosis
y na. and malaria haven't been dangerous like
before.
...chuyn trang du lch the diligent travelling web of The the Travel section
ca bo The Guardian Guardian (U.K) selected banh mi Sai
(Anh Quc) bnh Gon
chn bnh m Si Gn the tourist column of The Guardian
newspaper (UK) chose Sai Gon bread

135
Use of synonyms
Tim mch l mt trong Heart disease is one of the most popular common disease
nhng bnh ph bin disease of human beings.
ca nhn loi Cardiovascular is one of the popular
diseases of human being...
Cardiovascular trouble is one of the cardiovascular problems
human common diseases.
B tc ng mnh bi Being strongly affected by pollution and destructive fishing
nn nhim v cc devastating fishing method... methods
phng thc nh bt
c mang tnh hy
dit
rt nhiu bi many submarine coral reefs have been underwater coral reefs
ngm san h b ph destroyed and no longer fish's habitat.
hy, khng cn l mi
trng sng cho c na.
bng ch dinh healthy, suitable diet can help appropriate diet/regimen
dng hp l cng c controlling the disease.
th khng ch cn bnh
ny.
Nhng ngi chin s The unnamed soldiers died for freedom of unknown soldiers
v danh kia cht their countrymen.
ng bo h c sng Those unname soldiers had sacrificed
trong t do. themselves for their countrymen to live in
freedom.
L1 transfer
c mt ngun nhn so that there could heave a talented and a rich resource of able
lc di do c nng lc potential human resource to help other and talented people
v c ti phng s people and develop the country
cho t nc v nhn In order to have a rich talented, potential,
dn full of ability human resource to serve
country and citizens...
rt nhiu bi ...a lot of underground coral reef has underwater coral reef
ngm san h b ph been destroyed, which limits living
hy, khng cn l mi environment for fishes.
trng sng cho c na. many underground rock of coral are
damaged and are not environment of fish.
Many rocky field underwater and coral
reef are destroyed
Tuy nhin, bng ch However, suitable nutrient system can diet/regimen
dinh dng hp l control this disease.
cng c th khng ch However, with a proper nutrition
cn bnh ny. system can constrain this disease.
nc ta, t ngy m In our country, since economy-door open Since the government
ca kinh t day opened the economy /
In our country, from the beginning of the economy was
opened-economic day opened

136
...mi quan h tng ...the contrast relation between paradoxical relationship
phn gia mi trng environment and development always be
v pht trin lun l bi considered as a mathematic problems for
ton t ra cho cc nh economists...
kinh t
Thomas Edison tin rng Thomas Edison believed that his deaf - deafness
bnh ic ca ng disease had strengthen his focus ability. - hearing loss
lm tng thm sc tp Thomas Edison believed that is his deaf
trung ca ng y. disease increasing more his concentration.
Bn khng nn lm You should not break the friendship small amount of money
hng tnh bn hu ch v because of a small money.
mt mn tin nh. You should not make a damage for your
friend ship because of only small money.
...Mo con i hc chng ...Petty cat don't bring anything to school. small/tiny piece of bread
mang th g. Ch mang Only bringing one pencil. And bringing
mt chic bt ch. V with a small bread.
mang mt mu bnh Baby cat goes to school with nothing. A
m con con. pencil is in his hand only. And just with a
small bread.
Bring nothing Kitty attends school. Except
one little pencil. And one little bread.
...chuyn trang du lch ...the major tourist page of the Guardian The Travel section
ca bo The Guardian (UK) voted for Sai Gon bread
(Anh Quc) bnh the tourism page of the Guardian
chn bnh m Si Gn (England) voted Sai Gon bread
...bnh m Si Gn thuc Sai Gon bread was voted to be one among (top 10) most delicious
10 mn n ng ph ten most attractive street food in the and tasty street foods
ngon v hp dn du world... that attract tourists
khch nht th gii. Banh Mi Sai Gon (Saigon Bread) 10 most around the world
delicious and attractive street food in the
world.
...bi Mo con i hc the little small cat go to school of Phan
ca Phan Th Vng Anh Thi Vang Anh is the only work bread food
l tc phm duy nht is appeared.
mn bnh m c xut
hin
Word coinage
Vic ph rng, ph hy The destruction of forest and ozone layer the emission of
tng ozone hay thi cc or radio-active emition to oceans can lead radioactive substances
cht phng x ra bin to the global pollution
u gy ra nhim mi Deforestation, ozone destroyment or destruction of ozone
trng ton cu. radioactive emition are all cause global layers or emission of
pollution of environment. radioactive waste
bng ch dinh it is possible to prevent this disease by nutritious diet
dng hp l cng c sensible nutritional diets.
th khng ch cn bnh
ny.

137
Nhng ngi chin s Those noname soldiers dead in order that unknown soldiers
v danh kia cht their countrymen can live in freedom.
ng bo h c sng
trong t do.
V thuc khng hiu Because of uneffected medicine the the medicine was not
qu nn cng ty dc pharmacy company... effective
phm ng li xin li Because of uneffective medicines, so
v rt li vc-xin. pharmacist company gives respectful
apology and returns their vaccine.
Formal confusion
Nhn ti l nguyn kh Talent people are life-sustaining element talented people
quc gia of a country
Vic ph rng, ph hy Cutting down forest, destroying ozone environmental pollution
tng ozone hay thi cc layer or releasing radioactivity into seas all
cht phng x ra bin cause global environment pollution
u gy ra nhim mi Forest devastating, ozone layer destroying radioactive substances
trng ton cu or radiation substance emitting causes the
global pollution.
Thi Lan l nc duy Thailand is the only country in South East a powerful European
nht ng Nam Asia which has never been dominated by a country
cha bao gi b cai tr Europe developed country.
bi mt cng quc u Thailand is the only country in South East
Chu. Asia which has never been dominated by a
developed Europe country.
V thuc khng hiu qu Since the medicine doesn't work the pharmaceutical
nn cng ty dc phm effectively, the pharmacy company company
ng li xin li v rt apologized and took back the vaccination.
li vc-xin. Since the medicine did not work
effectively, the pharmacy company
apologized and took it back.
Rt kh chp hnh sao It is difficult to take a photo of Venus dense clouds
Kim v n thng xuyn because it is usually veiled by densely fog.
b che bi nhng m It is hard to take picture of Venus because
my dy c. it is usually veiled by densely clouds.
th mn n ph bin Banh mi Sai Gon which is the most unfortunate fate
nht Si Gn c s phn popular food in Sai Gon has a unfortune
kh "hm hiu" fate.
...mn n ng ph ... 10 most delicious and tasteful street tasty street food
ngon v hp dn du food which attracts tourists most in the
khch nht th gii. world.
...Mo con i hc chng Little cat went to school, bringing nothing a small loaf of bread
mang th g. Ch mang except for: A small pencil. And a small
mt chic bt ch. V load of bread.
mang mt mu bnh
m con con.
d l tui i hc no - ...Althoug ages of school - for instance in cities/towns in the South
nht l Si Gn v cc Saigon and the south city also at least once of Vietnam
th min Nam - cng time eat bread.

138
t nht mt ln gm bnh No matter what age you are, especially
m. when you live in Sai Gon and south town
you certain ate bread at least one time.
th mn n ph bin the most popular food in Saigon - literary works
nht Si Gn c s phn Saigon bread - is partially pitiful because it
kh "hm hiu" khi t is not brought into literature work much.
c vn nhn thi s a
vo tc phm
Others
rt nhiu bi many coral oust ranges were destroyed, Underwater coral reefs
ngm san h b ph unabling fish from living.
hy, khng cn l mi
trng sng cho c na.
B tc ng mnh bi (Many coral supplants are destroyed) by destructive fishing
nn nhim v cc the heavily impact of pollution and fishing methods
phng thc nh bt determinant devastation...
c mang tnh hy dit
Mi quan h tng phn the contrary relationship between a problem / dilemma
gia mi trng v pht environment and development is always
trin lun l bi ton t the asked problem for economists...
ra cho cc nh kinh t
...chuyn trang du lch the diligent travelling web of The the Travel section
ca bo The Guardian Guardian (U.K) selected banh mi Sai Gon
(Anh Quc) bnh as one of the ten most delicious and
chn bnh m Si Gn attractive street foods to tourists in the
thuc 10 mn n ng world.
ph ngon v hp dn du the diligent travelling web of The
khc nht th gii. Guardian (UK) selected it as one of the ten
most delicious street foods attracting
visitors best in the world.
th mn n ph bin the dish is popular in Saigon to have an unfortunate fate
nht Si Gn c s phn dim name as poet let into poem.
kh "hm hiu" khi t the most popular food in Sai Gon is
c vn nhn thi s a bread which has a bad view when it
vo tc phm. appears a little in interature term.

139

You might also like