Examining ESL Learners’ Knowledge of Collocations

It is generally agreed that collocational knowledge is an important language form for language learners in order for them to be proficient and fluent in the target language. However, previous studies have reported that second language (L2) learners lack collocational competence and they encounter difficulties in learning and using collocations. The present study not only investigates the overall collocational knowledge of Malaysian ESL learners, but more specifically, their productive and receptive knowledge of lexical and grammatical, which so far have not been further explored. Additionally, the ESL learners’ performance on three different types of collocations: verb-noun, adjective-noun, and verb-preposition is also investigated. Results of the study reveals a few interesting findings with respect to the Malaysian ESL learners’ overall knowledge of collocations, in particular their productive and receptive knowledge of collocations in relation to the three different types of collocations (verb-noun, adjective-noun, and verb-preposition). Pedagogical implications with regard to collocations and recommendations for future research are also put forward. investigated both the receptive and productive aspects of collocational.


INTRODUCTION
Having an extensive choice of vocabulary items is seen as a prominent element in the use of a language. Nevertheless, it is not sufficient for learners just to know the words without knowing how to use them (Phythian-Sence and Wagner, 2007). Effective use of a large repertoire of vocabulary would undeniably help learners to improve all the language skills, namely, speaking, listening, reading and writing. According to Lewis (1997), although words have always been the pillars of a language, it is the use of collocations which will assist learners to achieve communicative competence.
A 'collocation' is commonly described as a string of two or more words that have the tendency to co-occur (Cruse, 1986). According to Gledhill (2000), collocation has been defined in various depending on the specific aim of the observer. McCarthy (1990,) for instance, simply sees collocation as "a marriage contract between words," (p. 12), while Lewis (2000) defines it as "the way in which words co-occur in natural text in statistically significant ways" (p. 132). Despite various definitions of collocations by different scholars, collocations, generally, can either be lexical or grammatical (Shammas, 2013;Ebrahimi-Bazzazz, Arshad, Ismi, and Nooreen, 2014).
The use of a large quantity of word combinations like collocations is also believed to speed up language processing and hence create native-like fluency (Aston, 1995). As a matter of fact, having language knowledge is perceived to include having collocational knowledge as well (Nation, 2001). Altenberg (1998) emphasizes on the need for learners to have collocational competence or collocational knowledge as he asserts that a natural language could consist of as high as 80% of word combinations.
Nevertheless, Bahns and Eldaw (1993) claimed that learners' knowledge of collocational competence is far below their general vocabulary knowledge. In fact, numerous studies in L2 acquisition research have shown that lack of knowledge and use of collocations at different levels of proficiency could affect learners' communicative competence and language performance in a negative way (Bahns and Eldaw 1993;Stubbs 2004;Wray 2002;Nasselhauf 2005;Ozaki 2011). Therefore, the aim of this paper is to further investigate the collocational knowledge of Malaysian ESL learners, by focusing on their receptive and productive knowledge of both lexical and grammatical collocations, which to date have not been explored.
To be more specific, this paper intends to answer the following questions: 1) Is there a significant difference between the participants' productive and receptive knowledge of collocations? 2) Is there a significant difference between the participants' performance on the verb-noun, adjective-noun, and verb-preposition collocation tests?
IJALEL 9(1):1-6 3) What type of collocations in the receptive and productive test that learners found difficult to understand and produce?

LITERATURE REVIEW
What is Collocation?
The term collocation was first introduced by Firth (1957) in his theory of meaning in which he mentions "you know a word by the company it keeps". Cruse (1986) distinguishes collocation as 'syntagmatically simple' and 'semantically simple'; the former refers to an expression composed of one word in its normal sense to another restricted word (e.g. table a bill, save time), while the latter refers to "a single choice of meaning with an unpredictable or non-compositional sequence of words (e.g. let the cat out of a bag, spill the beans)" (in Gledhill, 2000, p.9). Halliday (1961) provides what is considered the most all comprehensive definition of collocation, framing the term as below: 'Collocation is the syntagmatic association of lexical items, quantifiable, textually, as the probability that there will occur at n removes (a distance of n lexical items) from an item x, the items a, b, c …Any given item thus enters into a range of collocation, the items with which it is collocated being ranged from more to less probable' (p.276) In simpler terms, collocation refers to a language-specific phenomenon whereby two or more words co-occur. It is composed of a node (i.e. head word) and a collocate or the word that occur with the node (Shin & Nation, 2008), for example, 'table a bill' in which 'table' is the node and 'bill' is the collocate. Collocations can comprise both lexical and grammatical words: verb + noun (break a leg), adjective + noun (dark horse), preposition + noun (after life) and adjective + preposition (popular with). Benson, Benson and Ilson (1986) further categorize collocation under lexical and grammatical collocations. Both the node and collocate of lexical collocations are content words (i.e. noun, verb, adjective and adverb), while grammatical collocations are made of a content word and a function word (e.g. infinitive to, preposition as shown below: Lexical collocations Verb + Noun take action; make appointment Adjective + Noun heavy meal; white lie Noun + Noun a round of applause; head judge Grammatical collocations Preposition + Noun on point, at anchor Adjective + Preposition angry at; happy for Verb +Preposition work for, graduated from

Collocational Competence Among ESL learners in Malaysia
In 1992, Nattinger and DeCarrico introduced the term 'collocational competence' to refer to the ability native speakers have to process language fluently and idiomatically (Pawley & Syder, 1983 in Henriksen, 2013) and fulfil basic communicative needs, thus, enabling them to communicate effectively in the social setting (Wray, 2002). The same competency is argued to be equally important for L2 learners. Nevertheless, according to Henriksen (2013), collocational competence is "acquired late and often not mastered very well by even fairly competent L2 language learners" (p. 29). He also states that collocational competence has received increasing attention in the last decade, and more focus has been invested on the area in SLA research and in second and foreign language teaching publications (e.g. Granger & Meunier, 2008;Lewis, 2000;Laufer & Waldan, 2011;Nesselhauf, 2003;Schmitt, Dörnyei, Adolphs & Durow, 2004;Siyanova & Schmitt, 2007).
In the Malaysian context, collocational competence has also gradually gained more attention. Earlier studies on collocational competence focused mainly on learners' competency of specific types of collocations. Among those studies were by Ang, Hajar, Tan and Khazriyati (2011), Abdullah and Noor (2013), Joharry (2013), Kamarudin (2013;2018), Zarifi and Mukundan (2014); and Yunus and Awab (2011). Ang et al. (2011) and Abdullah and Noor (2013) investigated the use of verb-noun (VN) collocations among the Malay ESL learners in Malaysia. Both studies reported problematic use of VN collocations among the learners. Abdullah and Noor (2013) concluded that even though more advanced ESL learners tended to use almost equal percentage of collocation of several lexical verbs as the native speakers, the usage was observed to deviate from the norms of the native speakers.
Another grammatical aspect researched was phrasal verbs (Kamarudin, 2013;Zarifi & Mukundan, 2014). Kamarudin (2013) in her investigation on the use of six phrasal verbs with particle UP in the EMAS corpus compared the Malaysian learners' use of phrasal verbs to that of the native speakers from Bank of English (BoE) corpus. The findings revealed that wrong usage of common phrasal verbs (e.g. pick up, wake up, get up) has strong association with the learners' lexical knowledge, their awareness of common collocates, familiarity with the context of use and most important their mother tongue. The appropriateness in the use of phrasal verbs was also found to improve over time, suggesting that learners had benefited from longer exposure to the target language. Zarifi and Mukundan (2014) conducted a corpus-based analysis of the creativity and unnaturalness in the use of phrasal verbs among Malaysian ESL learners. Learners were found to use phrasal verbs quite sparingly, but some of the phrasal verbs created were unnatural. Yunus and Awab (2011) in their investigation on collocational competence among law undergraduates focused specifically on prepositional colligations. The study investigated learners' competency on prepositional colligations that were most frequently used in law documents. The findings provide evidence that colligations of preposition, which is crucially required in the study of Law, still present a major challenge to the law undergraduates. Another collocation study by Joharry (2013) focused specifically on the collocation and semantic prosody of the lemma CAUSE. The study aimed to shed lights on Malaysian learners' awareness of the negative prosodic feature of CAUSE. The research concluded that the collocational behaviour of the lemma CAUSE by Malaysian ESL learners was more inclined to negative evaluation.
More recently collocational research in Malaysia has expanded to include investigation on the influence of individual differences on the acquisition of formulaic language. Halim and Kuiper (2018) adopted the semantic plausibility metric to measure Malaysian ESL's acquisition of restricted Verb-Noun collocations in written English. The majority of the learners managed to only score average points in the cloze tests conducted, confirming that ESL learners in general experience difficulties in acquiring formulaic language.
Empirical studies conducted thus far on the knowledge of collocations among different groups of ESL learners in this country have revealed producing and using appropriate multi-word units is challenging to the learners regardless of their level of English. The challenge is attributed to mainly to learners' lack of collocational knowledge (Howard, 1998). According to Ang et al. (2017), collocation is difficult to ESL learners as it contains "combination restrictions that ensure the collocability of the component words" (p. 116); this feature contributes to the difficulty of producing appropriate word combinations among these ESL learners. There also exist differences in the use of collocation between ESL learners and native speakers (Abdullah & Noor, 2013;Halim & Kuiper, 2018;Joharry, 2013;Kamarudin, 2013). Even though more advanced learners tended to produce collocation in the aggregate similar to that of the native speakers, their use was considered unnatural (Abdullah & Noor, 2013;Halim & Kuiper, 2018;Zarifi & Mukundan, 2014) and often times erroneous (Ang et al. 2011;Yunus & Awab, 2011;Kamarudin, 2013).
The studies reviewed nonetheless have shed some lights in the current state of collocational competence among ESL learners in this country; however, none have investigated both the receptive and productive aspects of collocational competence. The present study aims to fill this gap by examining these aspects further and provide the much-needed empirical evidence with respect to collocational knowledge among the Malaysian ESL learners. To be more specific, the present study mainly focusses on the learners' receptive and productive knowledge in relation to lexical and grammatical collocations.

METHODOLOGY Participants
A total of 21 students from a selected public university in Malaysia participated in the study. They consist of 10 male and 11 female students undertaking a Diploma in Food Technology program, and their English language proficiency is generally at the intermediate level.

Instruments
There are 2 instruments used in the present study -the Receptive Collocation Test and the Productive Collocation Test. The same 48 target collocations are used in both tests. In order to assess the participants' receptive knowledge of collocation, a receptive test of English lexical collocations is employed. The receptive test is adopted from Alsakran (2011), and the reliability index of the test is very high r =.92. The test is comprised of 75 items (48 target collocations, 27 items of mismatched collocations that act as distractors). The test items consist of different types of lexical collocations such as noun+ noun and verb+ noun. The participants have to evaluate whether the underlined part of each sentence is acceptable or not. Then they have to circle the number corresponding to the unacceptable sentence (see Appendix I).
The second instrument is a productive English collocation test (Haqiqi, 2007) which is to measure the participants' productive collocational proficiency. The productive test is highly reliable with a reliability index of r=.89. The test consists of gap-filling items, which include the 48 target collocations that examine three different types of collocations: verb-noun collocations (16 items), adjective-noun collocations (16 items), and verb preposition collocations (16 items). The initial letters of collocations are provided as clues to the right answer. Each item allows only one correct answer. In the verb-noun and adjective-noun collocations items, the initial letters of the target collocations are given as a clue, and in the verb-preposition items, the meanings of the phrasal verbs are provided. This is to ensure that the participants choose only the target word and to prevent guessing (see Appendix II).

Data Collection Procedures
The data collection starts by administering the Productive Test, which takes about 30 minutes for the participants to complete. Following this, the Receptive Test is given to the participants, and they are given another 25 minutes to do the test. In order to ensure all participants take both tests, thus, each of them is given a number and both instruments (Productive Test and Receptive Test) are also numbered. During the tests, participants are not allowed to use dictionaries and to leave blank any item that they are not sure of. This is to prevent them from guessing.

Data Analysis
All items in both tests (Receptive Test and Productive Test) are manually checked and marked as correct or incorrect since all items only allow one possible answer. The total score for each instrument was 48 for the productive test and 48 for the receptive test (the 27 items that act as distractors are not counted). Items that are not answered are considered as incorrect. Apart from that, errors such as incorrect use of verb tenses (e.g. It usually take) and spelling errors (e.g. narow escape), were not considered. To analyze the collected data, a statistical tool (SPSS) is used for statistical analysis including descriptive statistics.

RESULTS AND DISCUSSION
This section will present and discuss results of the data analysis in relation to the three research objectives stated earlier.

Is there is any Significant Difference between the Participants' Productive and Receptive Knowledge of collocations?
Table 1 above compares mean scores of the receptive and productive collocation tests. Results indicate that the mean score for the receptive test is very much higher (M=46.14) than that of the productive test (M=28.90). In order to further confirm that the difference in the mean score is significant, a paired sample t-test was carried out and results are presented in Table 2.
Results of the paired sample t-test above confirm that there is a significant difference between the respondents' receptive and productive knowledge of collocations (p<.05). This indicates that even though the respondents show a much higher level of understanding with respect to collocations, the production of this language form is still a great concern.

Is there any Difference in the Learners' Performance in the 3 different types of Collocations?
Results presented in Table 3 shows that the mean score for the 3 types of collocations: Verb+Noun (V+N), Adjective+ Noun (Adj+N) and Verb+Preposition (V+Prp) collocations. Findings indicate that V+Prp collocations has the lowest mean score (M=8.48) while the highest mean score is for the V+N collocations (M=10.9). This indicates that there is a difference in the learners' performance in relation to the 3 different types of collocations tested. In other words, of the 3 types of collocations investigated, the respondents have more difficulties in producing the V+Prp collocations in comparison to the other two types of collocations.

What type of Collocations in the Receptive and Productive Test that Learners Found Difficult to Understand and Produce?
As for the receptive test, further analysis was conducted to find out if there are any specific collocational items in the test that show low mean score. Results in Table 4 below present items with mean score less than M=0.90.
Results indicate all the 6/16 items with the mean score <.90 are V+Prp collocational items, which include held up, set off, came to, pick on, takes after and talked out. Linguistically, this type of collocations is called idiomatic phrasal verbs in which meanings are not transparent and cannot be understood simply by combining the meaning of each element in the combination. Hence, it is not surprising that the learners scored lower in this type of collocations indicating that V+Prp collocations are difficult for the learners to understand.
Finally, an analysis was conducted to find out whether learners also have difficulties in producing the V+Prp collocations. As expected, results indicate that there is a higher number of V+Prp collocations (12/16 items) in the productive test with the mean score <.90 as shown in Table 5. This finding clearly explains that the V+Prp collocations are not only difficult for L2 learners to understand, but they are even   more difficult to be produced by the learners. This finding is consistent with previous studies which reported that V+Prp collocation is a challenging language form to ESL learners (Kamarudin 2013, Zarifi & Mukundan, 2014.

CONCLUSION AND RECOMMENDATION
As mentioned in the Literature Review section, previous studies on the use of collocations have often revealed a lack of collocational competence among the ESL learners. The present study has further investigated and confirmed the problems with respect to Malaysian ESL learner' collocational knowledge, in particular their receptive and productive knowledge of collocations. Although the results show that generally they are able to understand collocations (receptive knowledge), they still face difficulties in producing the correct forms of collocations (productive knowledge). The verb-preposition collocations which is a grammatical collocation, is found to be the most difficult one for the learners to understand and produce, as compared to the other two lexical collocations. The considerable difference between learners' receptive knowledge and productive knowledge, and between grammatical collocations and lexical collocations, suggests a need in tackling this issue by giving more emphasis in the teaching of grammatical collocations such as phrasal verbs in language classrooms. The above findings also indicate a need for language teachers to think of ways to enable learners to learn, understand and produce appropriate collocations in suitable contexts. One of the ways to do this is by encouraging the use of dictionaries on collocations that provide examples of lexical items with different collocates and context of use. This can foster the learners' development of collocational knowledge and competence.
Language teachers may also use concordances that are used in corpus-based studies, or which can be extracted from concordance tools, such as WordSmith Tools and Antconc to design concordance-based exercises that would induce learners to notice the use of both lexical and grammatical collocations in contexts. By observing the language patterns in concordances, learners would be able to detect their own collocational errors. This awareness is indeed necessary in increasing learners' receptive as well as productive knowledge of collocations and would further help them in enhancing their communicative competence as a whole.