The Use of English Collocations in Written Translation – A Case of University English-Majored Students

The current study attempted to investigate English collocations used in written translation among fourth-year English majors at a university. The participants included forty-one fourth-year English-majored students and three translation teachers who are teaching English – Vietnamese translation courses in the university. To gather data, the researchers used two instruments: a test of English collocations including five types of collocations: an adjective going with a noun, a verb going with a noun, a noun going with a verb, a noun going with a noun and a verb going with an adverb. After data analysis, it revealed that just slightly over half of the student participants were able to find the correct collocations in written translation. Furthermore, a noun going with a verb and a noun going with a noun are the two main lexical errors made by most of the students. The interview was also used to seek translation teachers’ perspectives on students’ ability to translate texts in general and strategies of translation in particular. They perceived that most of their students were not aware of collocations in written translation and still had difficulties in both grammatical and lexical collocations. Finally, some recommendations on improving English collocations in written translation were also made based on these findings.


Introduction
"Knowledge is power. Information is liberating." (Annan, 1997). In the age of globalization, the borders of nations no longer exist. Also, the thirst for knowledge is increasing. Much knowledge of mankind is available not only in English but also in many different languages. Accordingly, translation plays an essential role in removing the language barrier and conveying information to readers. As stated by Saroukhil, Ghalkhani, & Hashemi (2018, p. 101), translation is "unavoidable in today's modern world" since "human communication equals translation" (Steiner, 1975, p. 49). Besides, translation is the transferring of verbal language messages into non-verbal forms (Jakobson, as cited in Munday, 2009). According to Munday (2001), the nature of translation is multilingual and interdisciplinary, covering languages, linguistics, communication studies, philosophy and a variety of cultural studies. However, translation does not appear to be easy for university students, even translators. Parks (2010) once said that to produce a novel, considerable endeavors of organizing and imagining are always required but to translate demands further. Wilkins (1972, p. 111-112) said that "while without grammar very little can be conveyed, without vocabulary nothing can be conveyed". This emphasizes how significant vocabulary is. In addition, how to use language naturally, it is a matter of collocation. Since words are not used alone but in combination with others. Collocation is not only a more habitual co-occurrence of certain words but the way they are used by native speakers to produce natural speech and writing. When words are heard together often or used in combinations regularly, they become stuck and may not be changed any longer (even their order) (Meidasari, 2007). Thus, any variations in organization of a collocation may sound strange to the native speakers. The phrase Ladies and gentlemen is a good example. It is heard many times and the order of each word becomes stuck and never changes any more. Also, if the order of words was reversed, e.g. Gentlemen and ladies, it would sound unnatural in the English context. From this example, collocation can be seen as crucial for language learners. Therefore, the use of improper collocations may affect the validity of translation (Jabbari & Kavoosi, 2017;Obeidat & Sepora, 2019). Depending on the sense, collocations are applied properly. For instance, translators must decide which of the words hurricane, typhoon or tornado that best describes the nature phenomenon -storm -based on the region it occurs.
Nonetheless, although many students are aware of the importance of collocations in translation, they do not use collocations in their translations frequently and sometimes they also make some errors (Obeidat & Sepora, 2019). Furthermore, collocations appear to be a challenging issue for second-language learners even for those who achieve a higher state of proficiency (Arnaud & Savignon, 1997;Barfield, 2009).
In short, "a careful study of collocational patterns in translated text can shed light on the cultural forces at play in the literary marketplace, and vice versa." (Kenny, 1998, p. 5). Therefore, the important role of collocation in translation is undeniable.

Objectives of the study
This research is an attempt to examine a group of fourth-year English majors' ability in using English collocations in written translation. What is more, to ensure the quality of the results collected from the test of collocations, the current study endeavours to explore the translation teachers' opinions on their students' ability in using English collocations in their written translation. These two aims were given so that the researchers can gain more information about this group of students in using collocations in their written translation in particular. Therefore, we proposed three research questions as follows: 1. Are fourth-year English majors able to use certain types of collocations in written translation? 2. How do teachers respond to the students' use of English collocations in written translation?

Definitions
Two approaches to the study of collocations are phraseological approach and frequency-based approach. The first deals with the restriction on substitutability and differs from free combinations. To illustrate, ill treatment is a collocation, however, it may be odd if ill is replaced by other synonyms such as sick, infirm or ailing. For example, people say ill treatment, not sick treatment. Differently, in the approach of frequency-based, collocations are seen as word combinations that co-occur frequently rather than by chance.
In the phraseological approach, scholars argue that lexical components are semantically related to each other. In particular, one of these components is called a "base" and is considered as a dominant component, the second one is a "collocate" and it depends upon the "base" (Hausmann, as cited in Tutin, 2008;Mel'čuk, 1998;Vasiljevic, 2014;Evert, 2008;Suwitchanphan & Phoocharoensil, 2014). Moreover, according to Nattinger and DeCarrico (1992), after a base and a collocate are established, neglecting all the functional words is necessary to concentrate on only the collocations.
Since the definition of collocations varies among researchers, it is important to clarify how the term is. O 'Dell & McCarthy (2017) stated that collocation is a semantic and syntactic word combinations that sound natural to the natives. Likewise, in A dictionary of Linguistics and Phonetics (Crystal, 2008), collocations are clarified as a term of lexicology which refers to the juxtaposition of lexical items. Substantial increase/decrease, deep dissatisfaction, of great importance are examples for English collocations. Nevertheless, it would sound odd to say "strong increase/decrease" or "strong dissatisfaction" or "of big/high importance". These exemplary phrases above illustrate vividly that collocations often string together. Correspondingly, Halliday and Hasan (1976) used the term collocation to indicate lexical items that are associated in some way in language. Sharing the same point of view, Lewis (1993) put forward the idea that the habitual individual words co-occurrence is collocations. Contrastingly, in frequency-based approach, Lewis (1997) and Bateni (2010) delineated collocations as sets of words that are combined naturally rather than frequently. In addition, Nattinger and DeCarrico (as cited in Suwitchanphan & Phoocharoensil, 2014) defined collocations as a series of specific words and often occurs mutually.

Classification
Based on syntactic feature, Baker (2011), Benson, Benson & Ilson (1986 and Bahns (as cited in Boonyasaquan, 2009) grouped collocations into two main types. The first group is grammatical collocations. They are phrases consisting of collocational meaning go through the paper (to examine carefully) and the literal meaning go through the paper (to move from one side to another side of the paper). Based on the fixedness of words, collocations can be categorized into two types: strong and weak collocation. Strong linking of phrases that functions as single words is called strong collocations. Linking of two common words, each of which can collocate with the others, is called weak collocation. (Lewis, 1997;O'Dell & McCarthy, 2017). For Benson et al. (1986), this kind of collocation is known as free lexical combination.
Based on another tactic of categorizing collocations, Sinclair (1991) classified collocations into upward collocation and downward collocation. In view of his classification, the term node is introduced to refer to the word being studied and the term collocate is used to indicate the environment of node. When a is node (e.g. walk, come, drive) and b is collocated (e.g. by), this is called downward collocation. For example: walk by, come by and drive by. On the contrary, if b is node (e.g. by) and a is collocated (e.g. now, chance, accident), this is called upward collocation. For example: by now, by chance and by accident. While upward collocation tends to be "the elements of grammatical frames, or superordinates". (p. 116), downward collocation provides a word's sematic analysis.
Notwithstanding that meaning appears to be a "property" of words, a word often contains more than one meaning. Hence, contextualizing it in its most common collocations is a must to grasp its appropriate meanings (Baker, 2011, p. 57;Vasiljevic, 2014;Meidasari, 2007). For example, to explain the word cold, people are likely to think of the state of being low in temperature such as cold weather, cold water or cold wind. Nevertheless, pursuing further investigation into the collocational meaning, cold can combine with person to indicate the unfriendly person, or go with call to describe a type of a phone call to an unacquainted person for the purposes of selling something.
This characteristic of collocation complies with the idiom principle of Sinclair (1991), in which collocations are explained to be ready-made blocks of words and reject any reorganizations. Thus, a word's collocations play a major role in identifying the proper meanings or "the company that a word keeps contributes to its interpretation." (Greenbaum, 1996, p. 427). Therefore, the arbitrariness of collocation constitutes a major impediment for EFL learners, Benson et al (1986, p. xxv) claimed that "non-native speakers cannot cope with them".
In addition to the arbitrariness, collocation's meaning is often unpredictable. It is impossible to predict which word is preferred by the others (Seretan, 2011). Moreover, there is no clear linguistic, semantic and syntactic relationship to state that these participating words are collocating (Cruse, 1986).
In contrast to earlier preceding scholars, Hill (2000, p. 53) found that "to an important extent vocabulary choice is predictable". For this account, he relies on popular or the most widely-known combinations of words. When we talk about the act of drinking, people might think of the most common word have. So, the listeners may be able to guess a great deal of probabilities such as coffee, tea, juices or even cocktails. However, the listeners may also consider other sorts of liquid such as oil, acid or shampoo but he repudiates these expectations and explains that the latter juxtapositions are not linguistically probable.
The unchangability of the components Since collocation meaning is perceived as a whole unit, then its lexical constituents might not often accept any variations in syntax or semantics (James, 2013;Makinina, 2018). For example, a business person may be away on a business trip, but not on a business journey or business expedition although trip, journey and expedition are fundamentally synonymous. Also, native speakers supposed that combinations such as congratulate on, warn against, pride on are naturally combined with each other and any substitutions, such as *congratulate for, warn at, pride about, are not accepted (Benson et al, 1986). Furthermore, the term fixed collocations is treated as idioms. This is certainly true in the case of the phrase "to and fro", of them none can be replaced by other words. (O'Dell & McCarthy, 2017).

Significance of Collocation
When emphasizing the significance of collocation, Lewis (2000) stated that mastery of collocation may bring native speakers of a language and advanced EFL learners achievements in language competence -"having a sufficiently large and sufficient phrasal mental lexicon" (p. 177). Besides, Deuter, Greenan, Noble & Phillips (2002) claimed that no natural English either speech or writing is absolutely collocation-free. Mastering collocational knowledge helps learners communicate or translate naturally and to sound more native-speaker-like. Collocationally rich language is more precise (Mounya, 2010). "Adherence to the collocational conventions of an FL contributes greatly to one's idiomaticity and nativelikeness, and not doing so announces one's foreignness" (James, 2013).
Moreover, one of the most important requirements in language learning is to expand vocabulary and collocation is considered to be one of the crucial factors in language competence (Vasiljevic, 2014). Furthermore, Hill (2000, p. 54) recognized that ELF learners face problems of listening and speaking because of "density of unrecognized collocations", which separates them from the native speakers who have met regularly these prefabricated chunks. Therefore, memorizing the formulaic phrases will enable the native speakers to make use of idiomatic language in language comprehension and to produce language fluently and naturally and more quickly (Pawley & Syder, 1983;Wray, 2002;Lewis, 2000;Hill, 2000;Rao, 2018).
Collocation is a vitally important aspect of language; in translation, therefore, "If grammar is the bones of a text, collocations are the nerves, more subtle and multiple and specific in denoting meaning" (Newmark, 1988, p. 213). Similarly, Dai (2009) concluded that translators might need to be based on collocations so as to make a right decision on word's style and field.
Through these prime statements, we can come to conclude that there are ways to learn how to better use collocations. Learners can choose to learn chunks or idiomatic expressions of the language to make sure that words that are collocated often appear in their mind. Then, when teaching, the teacher needs to have his or her students listen to chunks of collocations to imprint these in their mind. If learners are frequently exposed to such language input, it is sure that they can master collocations. Like what Deuter, Greenan, Noble & Phillips (2002) said, the teacher needs to raise his or her students' awareness that when a language is presented in written or in speech, it is usually collocations that make it sounds natural.

Equivalence in Translation
In order to support the acquisition of collocations in translation, the paper also aimed to briefly go through a matter of the use of equivalence. It is obvious that foreign language learners may not have any problems in translation if all the words or sentence structures in the source language can be found in the target language with the same meanings and vice versa. However, in reality, this does not happen. That is why translation is a very interesting topic to explore. Therefore, the following ideas will focus on equivalence in translation.
According to Pym (2014, p. 8) equivalent means "equal value" that can be gained at different levels. In addition, Newmark (1991) who traced specifically the advances in levels of equal value, showed that depending on the complexity of the source language text, translation equivalence may require the replacement of not only words, collocations, clauses or sentences but also paragraphs and texts. Therefore, when translating English sentences into Vietnamese ones, translators may produce either grammatical or lexical substitutes. Despite that, substitutes of grapheme are not by no means considered to be translation.
While the aforementioned scholars shift a strong emphasis on lexical units of the term equivalence, two other categories were recommended by Nida and Taber (1982), namely, dynamic equivalence and formal equivalence. Dynamic equivalence, on the one hand, presents that "the receptors of the message in the receptor language respond to it in substantially the same manner as the receptors in the source language" (p. 24). In other words, the reader in the target language can also understand the meaning of the translated texts as the reader in the source language. Although, it might not be possible to fully understand the meaning due to historical or cultural differences, reaching "a high degree of equivalence of response" is essential (p. 24).
Nonetheless, Newmark (1981) replaced the Nida's terms of equivalence by his new terms, communicative and sematic equivalence, respectively. The main difference in these two categories is that while communicative equivalence focuses on effects, semantic equivalence concentrates on contextual meaning. Putting the matter another way, the communicative equivalence addresses target language readers' needs and satisfies them by transferring foreign elements into their culture. In contrast, semantic equivalence attempts to retain as many properties of the source language as possible.
1. Denotative equivalence deals with the extra-linguistic content 2. Connotative equivalence concerns the lexical choices 3. Text-normative equivalence focuses on distinct types of texts 4. Pragmatic equivalence concentrates on the message's receivers 5. Formal equivalence relates to the form and the aesthetic of the text In short, according to Kashgary (2011), if translation is believed to be a process, then equivalence is considered to be the heart of the whole process. In line with Kashgary, many researchers such as Catford, Pym, Newmark and Nida also agree that equivalence is the center of translation issues. In addition, although equivalence still remains its defects, it is important to emphasize that equivalence is still one of the climactic "definitory axes" of translation since it plays a pivotal role in orienting problems that translators may encounter (Panou, 2013, p. 5). Nonetheless, Pym (2007) argued that only if the two languages appear to some degree of linguistic similarities, then there is an equivalence.

Translation Process
According to Nord (2005), translation happens to meet the needs of customers who want to understand an original text, written in a source language by a source language writer or transferred by a source language sender (who transmits a text in order to carry a particular message) in the condition of source culture. Therefore, a text is created not only to be translated but also to perform a specific communicative purpose of customers.
In terms of semantic aspect, Newmark (1991) proposed another way of processing translation, in which the more important words, phrases, collocations are, the more closely they must be translated. In contrast, the less important words are, the less closely they will be translated.

Purposes of Translation
The formation and development of translation aim at enabling communication of different languages (Massey, 2009).
In About translation by Peter Newmark (1991), the purposes of translation can be treated under five sections. Firstly, translation provides important insights and harmonies among countries, associations and people. Secondly, translation contributes to the successful, accurate transmission of knowledge. Thirdly, translation plays a role in mediating in different cultures based on "a common humanity, respecting their strengths, implicitly exposing their weaknesses." (p. 44). Fourthly, the core purpose of translation is to translate the great and universal works of the world in diverse fields "in which the human spirit is enshrined and lives" (p. 44). Finally, translation is treated as a significant skill in language acquisition.
In addition to the skill in language acquisition, Duff (1989, pp. 6-7) identified five reasons that translation should be taught in school. Firstly, the first language may have a strong effect on the insights of the syntactic features of both first and second language. Secondly, translation is a natural, ordinary and essential activity and is present in all fields and should of course be included in the school curriculum. Thirdly, knowledge and skills of second language may be increasingly acquired through learning translation since "language competence is a two-way system". Fourthly, standardised translations provide learners with Saccurate knowledge about second language. Finally, translation is a worthy skill to learn owing to four main reasons: the encouragement in discussing and criticizing, the enhancement of essential qualities (flexibility, accuracy and clarity), the relevant material for helping overcome the language structure problems and the need of translators.

Related Studies
Phoocharoensil (2014), in his report "Exploring learners" developing L2 collocational competence", studied the collocational competence of Thai EFL learners. The data was selected from essays of 90 first-year participants who study different majors. Overall, these results indicate that collocational problems truly exist due to the wrong use of learning strategies which are native language transfer and synonym. Furthermore, the data also reveal that students with both low and high levels are impacted by their mother tongue. Additionally, using synonyms to replace any unknown items is believed to cause collocational errors. Hence, Thai students need to be introduced to the linguistic differences between the two languages and should have a list of L2 collocations which cannot be translated into L1.

Lan (2015) conducted a study of "Insight into students' use of lexical collocation in Vietnamese-English translation".
Her purpose is to analyze the advantages gained from the correct application of collocations in the process of Vietnamese-English translation among third-year majors. 34 students of Thang Long University were asked to participate in a quantitative survey by completing monolingual questionnaires. The results of the study revealed that despite the full awareness of the significance of collocations in successful translation, students are likely to make mistakes in combining words. This is due to word-for-word translation and the wrong technique of amassing words since vocabulary is memorized in isolation. Therefore, the researcher advocates that changing the study habits would be the best method for all problems.
The thesis "Collocations: a neglected variable in EFL" of Farghal and Obediat (1995) aimed at addressing the collocational problems of English-majored juniors and seniors of Yarmouk University. 34 participants were selected to join the research by taking two translation tests. The results discovered that the participants lacked awareness of collocations as they treated collocations as single words. Hence, the researchers suggest students should consider collocations as multi-words units such as idioms or fixed expressions.
To investigate the frequency of unacceptable collocations of Persian students, Saliminejad & Karimkhanlooei (2018), in his study "A study on the type and frequency of unacceptable collocations in the English-Persian translations of Hemingway's Masterpiece: For Whom the Bell Tolls", listed the unacceptable collocations gathered from the translations of For Whom the Bell Tolls (Hemingway). The findings unveiled that the most common errors in collocation use are Adjective + Noun, Subject + Verb, Verb + Object patterns.

Participants
The study was conducted with the participation of 41 fourth-year students coming from three classes of English course of 2016. They are final year students who have completed three courses of translation subjects and are studying Specialized translation course. Moreover, three teachers who are teaching translation subjects were invited to participate in the research. However, the teachers did not join the test of collocations. They were invited to take part in the interview in which they gave their opinions about their students' knowledge of collocations in written translation classes.

Research Instruments and Materials
The study employed two instruments: a test of collocations for the student participants and an interview with the teacher participants. First, the test consisted of 15 questions embracing only one correct collocation and the remaining options were pseudo-collocates. Furthermore, all the items serving for the test were adopted from the book English Collocations in Use Advanced (O'Dell and McCarthy 2017). To comply with the first research question, the researchers adopted five types of collocations: an adjective going with a noun, a verb going with a noun, a noun going with a verb, a noun going with a noun and a verb going with an adverb. The second instrument used to collect data for the study was conducting an interview with three translation teachers who gave their opinions on their students' use of collocations in written translation classes. The questions were designed based on common opinions about use of collocations in written translation classes. Then, they were appoved by the same four translation teachers at the School.

Data Gathering Procedure
After that, a test for students was designed by gathering collocations from the book English collocations in use advanced (O'Dell & McCarthy, 2017). Also, the Vietnamese meanings of each collocation were translated by the researchers. After being scrutinized, the test was delivered to the test-takers via Google Form and the student participants had one week to complete this test. The reason we administered the test via Google Form because of the Coronavirus. Next, the researchers started to make an appointment with the translation teachers at the School of Foreign Languages to conduct an interview with them. Then, they gave us proper time to do this. Each teacher was interviewed for about 15 minutes. Once all the data from the student participants and teacher participants were collected, the researchers started to make an appointment for putting the data into analysis. We came to their department and sat with each of them for the interview and we were recording their answers for analysis afterward. We also took some notes during this interview.

Data Processing Method
In order to analyze the data, the researchers used the Microsoft Excel to process the fifteen -multiple choice question test. Each question contained four choices. The correct answers of all these fifteen questions had a choice B of the correct answer. So the researchers calculated the numbers of the correct response of each type of collocations and other unaccaptable choices of each questions of each type of collocations in the test and the counts were transformed into percentages. For the interview, the approved questions used to interview the three tranlsation teachers were written in a notebook which was used to conduct an interview with each of the teacher and finally we copied down the teachers' opinions on the students' use of collocations in translation for analysis and then compared these to what the students could really did on their test of picking the right collocations.

Test
In the test delivered to 41 fourth-year English majors, there were 15 items. Five out of 15 were adjective + noun combinations. Four questions dealed with verb + noun collocations. The numbers of questions about noun + verb collocations were two. Furthermore, two questions for each of the rest questions were about noun + of + noun and verb + adverb combinations. After being received, the data was analyzed as follows. Then, the results were displayed and explained in the charts below. As shown in the bar chart, there were two adjective + noun collocation questions that fewer of the participants found it hard to select the correct answers in this catergory. In addition, among the three remaining questions, the combination "giao tiếp phi ngôn ngữ" experienced a large number of the students opting the right collocation. Furthermore, the word groups "một quan chức cao cấp" and "kẻ thù nguy hiểm" were also selected by the majority of the participants, which accounted for 29 and 27 respectively. Overall, most of the proper collocations were the students' selection.
Here are the English versions of these collocations: Sử tăng nhanh (substantial increase), giao tiếp phi ngôn ngữ (nonverbal communication), một quan chức cấp cao (a senior official), nguồn góc không rõ ràng (unknown origin), and kẻ thù nguy hiểm (dangerous enemy). The pie chart illustrates the proportions of each answer that the test-takers chose. The question asked students to choose the correct answer for the meaning "sự tăng nhanh", and only 10 students chose the correct answer (accounting for about 24.4%). Moreover, there was a larger number of the participants deciding to choose "strengthened increase" and "strong increase" as the right answers, which comprised 34.1% and 31.7%, respectively.
Students tended to choose the answers depending upon the meaning of every element in the phrases. For example, in this case, the words "strengthened" and "strong" both refer to physical strength, however, the word "substantial" indicates an increase in value. Moreover, the word "strengthened" is not usually associated with the noun "increase".
Therefore, "substantial increase" is the right collocation (  From the graph, it is clear that there were 16 participants (accounting for about 39%) who chose the right collocation.
Although this was the most highly selected combination, the number of the participants opting for the second option which was "unclear origin," still accounted for a considerable number (approximately 31.7%). In contrast, the other two options made up quite low percentages ("uncertain origin", 17.1%; "unidentified origin" 12.2%).
explained why there are quite many students who preferred this combination. On the other hand, if we try to explain the word "unknown" in isolation, we can see that it means less or even not known. Therefore, it is unlikely that the majority of the students know the word "unknown", in this case, it must referr to "unidentified" and "uncertain".
Taking all into consideration, the results recorded only two out of five questions which accounted for the largest number of the students who selected the improper collocations. These figures can still prove that the type of adjective + noun collocations was still the complexity of word combinations.
4.1.1.2 A verb going with a noun (A totality of four collocations to be tested)

Figure 4. The results of verb + noun collocation questions in the test
The bar chart illustrates the results of the test of verb + noun collocations. In general, as can be seen from the graph that half of the questions for this type of collocation shared nearly similar number of the participants who opted for the proper collocations. Specifically, the number of the participants choosing the right word combination for the meaning "Thả diều" and those picking the right word grouping for the meaning "Tuyên bố độc lập" constituted 35 and 34 respectively. Moreover, for the meaning "Tiến hành điều tra", there was almost half of the participants (20 students over 41) choosing the correct answer. However, the collocation "Chịu được áp lực" was the question chosen by just a few students.

Make an investigation Implement an investigation
The chart describes the percentage of the participants dealing with the collocations of the Vietnamese phrase "Tiến hành điều tra". From the chart, the proportion of the students who chose the right collocation ("conduct an investigation") experienced the significant choiceapproximately 48.8%. Followed by this great percentage, the option "implement an investigation" also reached the second rank in the choice preferred by the students and it obtained around 34.1%. Additionally, "make an investigation" was the preferable selection of only seven students, which made up about 17.1%. Although the number of the participants who selected the right collocations was less than half, the proportion of the correct collocation chosen by the participants was still considerable.
The Vietnamese phrase "chịu đựng được áp lực", experienced a great number of choices of its English version "suffer pressure" . However, withstand and suffer are not the synonym ones. Oxford (2020) defines withstand as the state of having enough strength so that it cannot be harmed and its synonym is resist. On the other hand, suffer basically means being affected. Therefore, the selection of which word, withstand or suffer, to collocate with pressure will possibly lead to different meanings. Notwithstanding, for the given Vietnamese verb "chịu đựng được", the English verb withstand collocates with pressure will best describe the original meaning. From that, we can say that students may be familiar with the phrase suffer pressure and they might neglect or misunderstand the given meaning. Hence, that causes their wrong choice of collocations.
In general, from the analysis, verb + noun collocations witnessed few errors made by the participants. Take all into consideration, students might face less trouble with verb + noun collocations. The researchers found that verb + noun collocation poses a challenging problem on EFL learners due to low frequency of intensifiers.

Figure 6. The results of noun + verb collocation questions in the test
The chart describes the number of the participants choosing the right collocation. Overall, it is clear that most students failed to pick the correct collocations for the two questions.
Here are the English versions: sóng vỗ (waves lap) and muỗi chích (a mosquito bites). The chart depicts the percentage of each answer that the participants selected. Interestingly, only seven participants picked the correct answer (waves lap), which accounted for 17.1%. In addition, the most selected answers were waves clap and waves flap, which comprised 36.6% and 34.1% respectively. Furthermore, 12.2% wass the proportion of the least opted option.

Waves lap
Waves flap

Waves wrap
The proportion of the responses to the collocation "muỗi chích" was illustrated in the pie chart. Overall, a large number of participants chose the correct collocation. Specifically, the collocation "a mosquito bites" made up a considerable percentage, 43.9%. However, there was also a substantial proportion of students whose answer was a "mosquito stings", and it comprised26.8%. The percentages of the rest options, namely, "a mosquito sticks" and "a mosquito injects" were similar at 14.6%. For noun + of + noun collocation questions, the data continued to illustrate the low number of the participants choosing the right collocations.
Here are the English versions: hành động bạo lực (acts of violence) and một đàn chó (a pack of dogs).

Figure 10: The percentage of answers chosen by students
As can be seen from the graph, while the correct answer accounted for 26.8%, the incorrect one made up 61.0%. Additionally, also a few students chose activities of violence and acting of violence, which comprised 4.9% and 7.3% respectively. In general, the students encountered a problem with the noun + of + noun collocation.  The chart describes the percentage of each answer chosen by students. Overall, the percentage of the most selected option exceeded the sum of the percentages of the rest of the answers. In detail, "a herd of dogs" accounted for the most remarkable proportion, which constituted about 53.7%. Furthermore, 29.3% was the ratio of the correct collocation "a pack of dogs". Moreover, "a team of dogs" and "a group of dogs", respectively, accounted for approximately 4.9% and 12.2%.
4.1.1.5 A verb going with an adverb (A totality of two collocations to be tested) Figure 12. The results of verb + adverb collocation questions in the test The chart below highlighted the different numbers of the participants' choices in the two verb + adverb questions. It can be seen that, the question "Cứ xử bốc đồng" obtained more than half of the participants who selected the right collocations. In contrast, there were only 16 students opting the right collocation for the remaining question.
Here are the English versions: cư xử bốc đồng (behave impulsively) and bắt đầu lại từ đầu (start afresh). Figure 13. The proportion of each option for the question "Bắt đầu lại từ đầu" As can be seen from the chart, there was only 39% of the students who opted the correct collocation. In addition, there was 29.3% of the participants choosing the "start initially". Furthermore, while the number of the students selecting "start originally" accounted for 22%, the figure for this group of participants whose option was s "start newly" at 9.8%. This can suggest that these English -majored students have a habit of employing word-for-word translation. This is

A herd of dogs A pack of dogs A team of dogs A group of dogs
even true that the meaning of a cluster of Vietnamese words and its order give a fixed meaning and then these participants follow this rule and apply this to English texts. For example, bắt đầu lại từ đầu (bắt đầu is a verb(start), lại is an adverb (again), and từ đầu is a prepositional phrase (from the beginning. This can look like this: start again from the beginning). Although this sound acceptable, it is not an English collocation.

Interview
To answer the second research question, the researchers used the interview to examine the teachers' remarks on using English collocations in students' translation texts. The findings from the interview revealed that some of the comments on students' use of collocations in written translation are somewhat in contrast to students' abilities of applying collocations obtained from the data in the students' questionnaire.

Situation
All the teachers reported that the students use English collocations in a low frequency. Furthermore, they also made some errors in their translation. According to these teachers, students often encounter difficulties with both grammatical and lexical collocations. In Vietnamese-English translation, students are unlikely to find proper collocations and they tend to translate literally or sometimes coin totally new word combinations. Furthermore, the teachers agreed that students' mistakes also stem from the unwarranted interference of their mother tongue. Additionally, this corresponds to the findings of Jabbari & Kavoosi (2017) and Do (2011) which both lexical and grammatical collocations are tough for students. Moreover, students are said to be unaware of the significance of collocations in written translation. Additionally, some of them are misinterpreting collocations in English-Vietnamese translation. However, the majority of the teachers reported that after being taught and trained by teachers, the number of collocations used in translation texts by the students increased. This means their translation ability is being improved.

Reasons
According to these three teachers, lack of vocabulary range, specifically collocations, is the ground for students' errors in using collocations in tranlsation. In other words, the limitation of words prevents students from explaining their thoughts clearly and accurately. This might be due to students' learning habits, which means the students learn new words in isolcation not in chunks or how a word collocates with another word . As a result, they generate words based on their available vocabulary in their memory and this results in the strange combinations of words.

Teachers' constructive advice
The first teacher Students had better increase the quantity of English collocations in translation. Furthermore, they also need to learn words in groups rather than in isolation. Additionally, due to the time restriction for class meetings, students should plan their own self-study to gain more knowledge and skills of translation. Moreover, students should pay greater attention to the differences between Vietnamese and English versions.

The second teacher
Students need to read more, particularly academic texts because those will help widen the vocabulary and collocation range. Furthermore, learning topic-specific vocabulary is also important. In addition, translation theories are essential for translation subject. Therefore, equipping oneself with background translation theories will help students study more effectively. Moreover, a native speaker-published dictionary may contribute to the improvement of the collocation use. Additionally, owing to the involvement of inter-culture, translation requires a translator to have insights of not only target languages but also source languages.

The third teacher
Since collocations are significant in written translation, increasing the quantity of collocations in written translation is crucial. Moreover, students should accumulate collocations by themselves, because gaining collocations from teachers and course books is not enough.
All in all, these remarks suggested the following concerns. Firstly, students should raise greater awareness of the role of collocations not only in translation, but also in language production in general. Moreover, learning autonomy plays a crucial role in acquiring knowledge. In other words, due to time restriction for class meetings, teachers are not able to share all knowledge for them. Therefore, self-study contributes enormously to great gain of collocations. Secondly, students should treat words in phrases or chunks and learn them in their surrounding combinations. Furthermore, widening their topic-specific vocabulary is also advocated by most teachers as it helps their translated texts sound more natural. Last but not least, in order to achieve success in translation, it requires the improvement of both target language and the source language. This is because knowing the first and second language well helps learners to pick up the right use of words or idiomatic meanings of phrases of words in either language. For example, in Vietnamese, people say 'drink medicine', but in English, English people say 'take medicine'. This analysis is considered to back up the findings of the data collected from the test of collocations that the student participants did and helps us know more about these teachers' opinions on their students' current ability to use collocations in written translation.

Discussion
With the aim of determining the status of using English collocations in written translation, some findings revealed that students have basic understandings of collocations and they know a number of collocations.
By conducting a collocation test, the researchers wanted to discover whether forty-one fourth-year majors chose the right collocations that best fited the provided Vietnamese meanings. The findings revealed that selecting the proper collocations seems not to be the most aching problem. To illustrate, twenty-one out of forty-one students (nearly 51.22%) chose more than half of the correct collocations. Furthermore, the remaining students (twenty test-takers) (about 48.78%) chose more than half of the wrong collocations. Through the data, we can see that students know many patterns of lexical collocations, as evidenced by the average of over 51% of correct answers. This supports the statement of González Fernández & Schmitt (2015) as they said that EFL learners may know more collocations than they thought. In addition, these findings are opposite to the assertion of other authors (Farghal & Obeidat, 1995;Bahns & Eldaw, 1993;Laufer & Waldman, 2011;Nguyen & Webb, 2016) when they claim that collocational knowledge of non-native English speakers is low.
However, from the findings, it is noteworthy that nearly half of the participants were unable to choose correct English collocations for many common Vietnamese words or phrases. The results suggested that their translation was still greatly influenced by their mother tongue and they employed therefore literal meaning of each isolated word to attach to each other orderly to make a cluster of words as a whole meaning. This issue can be brought into consideration when teaching translation in class so that many more students work hard on collocations and form a habit of checking appropriate collocations which bring natural ways of translated texts. For nouns + verbs and nouns of + nouns, these two categories experienced the largest number of students who chose incorrect collocations. For the question "hành động bạo lực", according to Oxford (2020), act and action bear the similarity in meaning, which describes something that someone does. However, their major difference is the pattern. While action cannot collocate with the preposition of or adjectives but can go with the verb take, act is in reverse. Moreover, activity refers to what somebody does to gain something, and it cannot collocate with the preposition of. Therefore, the most appropriate collocation is acts of violence.
In addition, with the phrase "một đàn chó", many students preferred a herd of dogs. In Vietnamese, "đàn" can go with various types of animals such as pets -dogs or cats, cattle -cows or buffalos, carnivores -lions or tigers, herbivores -rabbits and so forth. However, in English, there are also various words to mean "đàn" such as a pack, group, herd, hound, swarm, flock, pride, school, pod, or troop (O'Dell & McCarthy, 2017). Furthermore, herd is used for large herbivorous animals only. Therefore, maybe due to lack of register, the major number of the students misused this type of collocation.
On the other hand, for noun + verb collocation, the students also encountered this difficulty. For the question "sóng vỗ", the students preferred the two verbs clap and flap to collocate with the noun waves. This explicitly presents the influence of first language ontheir translation texts. In Vietnamese, clap, flap and lap all means "vỗ", but this is in marked contrast to English where their meanings are entirely different. Clap demonstrates an action of hitting hands together in several times to make sound. Meanwhile, flap is a movement of wings to fly, lap, in case that collocating with waves, means moving or hitting the previous waves of the latter waves. In this question, it is evident that the mother tongue has a strong impact on students' translatability, especially in word choice. This bears little resemblance to the results found by Ünver (2018), Lewis (1997), where the intervention of the mother tongue causes the inaccurate translation of English phrases.
For the question "muỗi chích", although the majority of the students picked the correct answer, the number of students who picked the option a mosquito stings was still high. According to the explanation of Oxford (2020), both the verbs sting and bite have the same definition which is the act of creating a hole in the skin. Despite the similar meaning, each one can collocate with a restricted kind of animal. For instance, sting can go with nettles, wasps or bees. On the contrary, bite only collocate with some of the insects such as mosquitos or spiders. These are examples of what fixed collocation is (Lewis, 1997).
From the analysis, the students have an understanding of collocations and they know a number of collocations. However, according to most teachers, the reasons students used few and wrong collocations in translation are due to the subjectivity of learners in learning vocabulary and using inappropriate translation techniques. The teachers shared that most students translated their texts using word-for-word mainly. Although word-for-word translation is still considered to be a translation technique and is still applied (Schwarz, as cited in Nida, 1964;Bateni, 2010), overuse of word-for-word translation will make translation less effective to convey the main massages. Another reason is the influence of the mother tongue on the translation process, especially from Vietnamese to English. All in all, the teachers confirmed that the students are not truly aware of the significance of collocations in written translation (Abdelhay & Dafallah, 2017). The analyzed data support the reports of the earlier researchers that although they are university students, they also encounter difficulties in collocations (Altenberg & Granger, 2001).
This result is opposite to the one obtained from the view of students. This can be explained that although students know a quite great number of the collocations in this study, they might not remember to bring them out of the mind when translating their Vietnamese texts. On the contrary to the test, all test-takers were provided a collocation stem for each question but in translation they had to retrieve the entire phrases by themselves without any vocabulary hints.
Collocations pose a threat on sophisticated production skills of language and also translation texts. Therefore, increasing the load of vocabulary is a must to improve the quality of translation. Firstly, students should learn words in chunks or phrases. Since many Vietnamese students'style of learning English is focused on individual words. Therefore, students may not understand the meaning of words in a fixed context (Nguyen & Webb, 2016;Tran, 2012). Moreover, students also have to know that, in some cases, the amount of vocabulary may contribute generously to the well-qualified translation. Secondly, if we considered translation as a school subject, it would require higher fundamental knowledge. Hence, when studying translation, students need to master many aspects of a language. For example, skills like listening and speaking can help enhance oral translation while reading and writing can help ameliorate written translation. But translators and interpreters need to widen their knowledge of collocations to smoothen their translated works.

Conclusion
Translation plays an important role in today's integration process, being from a bridge for multi-cultures to a compulsory subject in university-level training programs. Therefore, the products of translation in general and written translation in particular need to be paid more attention. Unlike oral translation, which is limited by time, written translation has a great deal of time to complete and have many factors that contribute to the quality of translation, especially collocations.
Collocation has no concept of true or false based on linguistics, but they are a concept of naturalness. Therefore, collocations must be learned by heart and the more students know about them, the better their translation ability becomes.
This paper aims to investigate the use of collocations in written translation of fourth-year English majors at a university. However, the participants did not need to actually provide their translated sample. The researchers had difficulties in asking the participants' time for it. Thus, the current study with miltiple choice questions and they needed to use their knowledge of collocations to find the correct answers. The researchers believed that this was the first stage of the study and will conduct further investigation of students' ability to use collocations in their translation. The results showed that overall, although students realized the importance of collocations in translation, their knowledge of English collocations is limited. Moreover, collocations are something they still need to better improve in the future. With the suggestions made by the participant teachers, we hope students in the future can find them helpful to improve their use of collocations in their translation in particular and in producing language in general.

Recommendations
Nesselhauf (2005, p. 252) stated "It is essential that learners recognize that there are combinations that are neither freely combinable nor largely opaque and fixed (such as idioms) but that are nevertheless arbitrary to some degree and therefore have to be learnt." Additionally, they should encourage students to own specialized dictionaries like native speaker-published ones such as Oxford, Cambridge, Longman, Collins or Merriam-Webster and a collocation dictionary. A collocation dictionary will help students understand clearly the meaning and the usage of words. Therefore, their use of word combinations in translation may be more precise and accurate.
In addition, a translation curriculum or training must be fully designed. For Saroukhil, Ghalkhani, & Hashemi (2018), the ideal learning plan should include these criteria: 1. Linguistic, extra-linguistic and discourse knowledge of the language 2. Cultural knowledge of the L2 3. Abundant use of reading comprehension techniques 4. Appropriate use of translation techniques 5. Knowledge of translation theories (p. 108) Moreover, the interest and motivation to learn translation are proved to be absolutely necessary. Accordingly, teachers could organize some interesting activities such as voice-dubbing which may arouse the interest of students, instead of doing boring translation exercises (Pham, 2016).
Additionally, enriching background knowledge may contribute greatly to the success of translation. Since "Our world continuously changes and evolves" (Nguyen & Trieu, 2015, p. 31) , it is essential to keep up to date with the new information. Furthermore, students may need to acquire prior knowledge to texts they are going to translate. This can be successfully achieved by means of reading. One of the teacher interviewees reported that the students had better enhance their ability of reading comprehension and particularly, learn the topic-specific words. Again, equipping oneself with knowledge of translation such as types of translation, the purposes of translation and translation techniques might be of help in sharpening well-qualified translation texts.

To teachers
Problems of collocational usage mostly come from the overuse of literal translation. Hence, teachers should raise students' awareness of using collocations in the early stage of any academic courses. If teachers intend to incorporate their translation classes with collocations, they had better attract students' attention to collocations and those teachers whose collocation range is limited should be trained more (Koc, 2006).
On the contrary, students would be asked to create stories based on given collocations, after that, they were told to retell their stories without looking at papers (Lindstromberg & Boers, 2008). Providing collocations in a particular topic might help attract the students' attention, especially romantic topics (Vasiljevic, 2014).
In addition, learning chunks that contain sound repetition may help students remember effectively. For instance, from hero to zero and boy bands are good examples since these kinds of collocation sound more natural (Lindstromberg & Boers, 2008).
Finally, Sethi (2013), in her analysis, indicates that retrieval is paramount of importance. The recovering of stored phrasal units may leave traces on memory. Sharing the same viewpoint, Lindstromberg & Boers's (2008) and Hill (2000) research complements that of Sethi. Lindstromberg & Boers mentioned review is a significant process because it helps create traces on memory. Therefore, teachers should create favorable conditions for students to learn certain target collocations and to repeat those word combinations but in changing circumstances. To illustrate, teachers could let students fill in a text where the collocations are left vacant. Students, then, will recall the collocations and fill in the text. On the contrary, the teacher may present a list of learned collocations and whole new pieces of text and collocations are also left unoccupied and ask students to fill in the blanks.

To students
The findings revealed that the most aching problems are the literal translation and the influence of mother tongue and misuse of vocabulary. Therefore, when translating, students had better treat words in their collocates rather than translate each word individually.
Moreover, students should seriously contemplate to socio-cultural dissimilarities as they might impact the students' way of using words. In addition, it is vitally important to maintain the collocational meanings of the texts rather than the word count (Nida & Taber, 1982).
Furthermore, all the teachers interviewed agreed that students must extend their vocabulary range and learn words in groups, not in isolation, and pay attention to the word register. What is more, learning autonomy is crucial for learning. Owing to the limitation of time, teachers may not share all their knowledge with students. Preparing prerequisite knowledge before studying may encourage students to discuss, express or even share their knowing confidently (Nguyen & Trieu, 2015).
Finally, training yourself to pay attention to all the collocations appearing in texts, conversations, films, songs, lectures and so forth and "record" all would help widen the knowledge of collocations (O'Dell & McCarthy, 2017, p. 12).

Limitations
Despite a great deal of effort to finalize this paper, some of the possible limitations are inevitable. Firstly, the test was delivered to the participants via Google Form. Therefore, the researchers could not exercise greater control over the participants' use of dictionaries and this test was undertaken mainly based on the honesty of the implementers. However, from the recorded data, it hinted no interference of the dictionary sources. Secondly, although more than half of all fourth-year students joined this study, the sample size of the study was quite small, so the data collected were not highly generalized. Thirdly, there were only three teacher interviewees, so the information related to the contemporary knowledge of these students in using English collocations was not much. There should have been a translation test for the student participants to do, so better results of their using collocations would have been produced. Finally, the test only arrived at the phrase level, so the contexts, the objects, and the translation purposes for selecting collocations have not been considered.