EFL learners’ interlanguage can be distinguished from the native speakers’ language by the fact that their interlanguage consists of features which show the incomplete mastery of the language. It is characterized by deviations called ‘errors’. As they go through the process of language learning, they make a lot of mistakes in their speech and writings. One of their areas of difficulty relates to ‘collocations’, a concept which was first put forth by J. R. Firth in 1957 who defined them as “actual words in habitual company” (Firth, 1957, p. 183).
EFL learners do not seem to pay attention to collocations; they usually focus on the individual words, disregarding sets of other words with which these individual words co-occur (Mirsalari and Shokouhi, 2010). Bahns (1993) sees the notion of collocation as an ignored element in syllabus design which results in learners’ deficiency of the knowledge of collocations.
1.2. Statement of the problem
L2 learners often have particular problems with word combinations, even at a relatively advanced level (Brown, 1974; Channell, 1981; Cowie, 1978b). Most students would, for example, have no difficulty producing the word diary when the meaning they wish to express requires it. They, however, have considerable difficulty in trying to figure out whether they should say *maintain a diary, *conduct a diary, or keep a diary?
This deficiency seems to be, to some extent, due to the lack of collocational knowledge among EFL learners, and to a large extent, the inadequate emphasis given to collocational patterns in their textbooks, and the type of instruction they receive. Bahns (1993) believes that when learners lack collocational knowledge, they resort to the L1 as the only resource, and thus perform well on those collocations that have L1 equals than on others. Lewis (2000) contends that EFL learners not only need to know what is right to say, but also what is wrong to say. Finding out the learners’ collocation errors and informing them of their errors by their teachers can raise their awareness of collocations and thus improve their knowledge of collocations. Therefore, the present study investigated English majors’ grammatical collocation errors and explored types of grammatical collocation errors which appear to be more persistent in their performance as EFL learners.
1.3. Purpose and scope of the study
The importance of collocations and the difficulty they pose on EFL learners has been the center of attention by several researchers in the field of EFL/ESL. They believe that the knowledge of collocations is essential for EFL learners in order to have a mastery of the language (Brown, 1974; Channell, 1981, Howarth, 1998; Nattinger, 1980). Accordingly, they have argued that teaching collocations has an important role in helping EFL learners improve their level of mastery in English. However, few studies have been conducted in Iran about errors in the performance of Iranian EFL learners on different types of grammatical collocations. Thus, the present study was planned to investigate the distribution of 6 types of grammatical collocation errors in Iranian EFL learners’ performance, to find the most frequent types of grammatical collocation errors, and to see how their knowledge of grammatical collocations develops as they advance to higher levels of proficiency so that practical solutions can be offered to minimize the types of grammatical collocation errors which are persistent in EFL learners’ interlanguage.
1.4. Research questions
The present study was an attempt to find possible answers to the following interrelated questions:
1- What is the distribution of grammatical collocation errors in Iranian EFL learners’ performance by their level of proficiency?
2- Which types of grammatical collocation errors are more persistent in Iranian EFL learners’ performance?
3- How does Iranian EFL learners’ knowledge of grammatical collocations improve with their level of proficiency?
1.5. Significance of the study
Learning the grammar of a language in general, and the grammatical collocations of that language in particular, is one of the most important aspects of learning a foreign language. According to Shokouhi and Mirsalari (2010), if collocational associations are not properly taught and learned, the resulting irregularities may mark the learners’ language problematic and non-native like. Most Iranian EFL learners seem to have good knowledge of English grammar, but they have many problems with production of grammatical collocations. They need to know which words co-occur and how we can use them in our actual communication. Nattinger and DeCarrico (1992) emphasize the importance of collocations for language acquisition and use. They believe that learners’ attention should be drawn to word combinations from the beginning stages of learning.
Those who doubt the value of teaching collocations might argue that collocations need not be explicitly taught as learners will simply acquire them along with relevant vocabulary (Mackin, 1978). However, many studies, such as that of Bahns and Eldaw (1993), have responded to this argument by showing that learners’ knowledge of collocations lags far behind their knowledge of vocabulary in general.
Iranian EFL learners commit errors when producing English collocations, especially grammatical ones. Their errors show that they tend to depend on their personal language strategies which facilitate their learning. Their problems seem to be in part due to their lack of knowledge of grammatical collocations and, to a greater extent, to insufficient emphasis given to collocations in their textbooks as well as the kind of instruction they receive. Therefore, grammatical collocations must be given sufficient pedagogical significance in Iranian EFL/ESL programs (Shokouhi & Mirsalari, 2010)
1.6. Definition of the Key Terms
Collocation: the habitual occurrence of a word with another word or words with a frequency greater than chance or a word or group of words that habitually occur together (e.g. heavy drinker) ( Oxford English Dictionary).
Lexical Collocations: Combinations of nouns, verbs, adjectives and adverbs (Benson, Benson and Ilson, 1997).
Grammatical Collocation: A noun, an adjective or a verb, plus a preposition or a grammatical structure such as an infinitive or a clause (Benson, Benson and Ilson, 1997).
1.7. Outline of the thesis
The present study is presented in five chapters. Chapter I provides an overview of the study and states the purpose of the study along with the research questions. Chapter II reviews different definitions of collocation in the literature and presents the kinds of collocations categorized from different perspectives. It also mentions the properties of collocations and the importance of learning English collocations. It also provides an overview of studies on English collocations. Chapter III describes the participants of the study, the instruments used to collect the required data, and the data collection procedures. Chapter IV presents the results of the study. Chapter V summarizes the findings, discusses the results, mentions the limitations and the implications of the study, and also provides suggestions for further research.
The present chapter reviews the studies on collocations and presents their findings. In section 2.2, different definitions of collocation in the literature are mentioned. Section 2.3 deals with kinds of collocations and puts forward the classifications of collocations by researchers in the literature. Section 2.4 presents the properties of collocations, and the importance of learning English collocations is discussed in section 2.5. Section 2.6 provides an overview of collocation research in terms of three approaches to the studies on collocations; i.e. lexical, semantic, and structural. The empirical studies on English collocations are categorized and summarized in section 2.7. Finally, section 2.8 concludes the chapter with what is done and what remains to be done on English collocations, part of which will be the focus of the present researcher.
2.2. What is a collocation?
The term collocation has its origin in the Latin verb collocare (to arrange; to set in order). It was first introduced by the British linguist J.R. Firth in 1957. Firth, who is known as the father of collocations, defined collocation as “the company that words keep” (Firth, 1957, p.183)
There have been diverse definitions of collocation in the literature. Halliday (1966, cited in Li, 2005) argued that when two words co-occur “independent of grammatical types and likely to take place over sentence boundaries”, this co-occurrence is called collocation (p. 7). Later, adopting the notion of Halliday’s lexical set, Halliday and Hasan (1976) defined collocations from the aspect of discourse. They proposed another definition for collocation and defined it as “cohesive effect of pairs of words, such as flame . . . candle, king . . . crown, and hair . . . comb” (p.7).
Yorio (1980) referred to collocations as ‘conventionalised language forms’ and Alexander (1984) called them ‘fixed expressions’. A further definition of collocations was provided by native speakers on the basis of their views towards English. For example, Nattinger and DeCarrico (1992) define collocations as habitual associations of words that co-occur with mutual expectancy. Lewis (1997) claims that collocations are word combinations that occur naturally with greater than random frequency. He believes that collocations co-occur, but not all words which co-occur are collocations.
However, Gitsaki (1999) believes that the literature on collocation lacks a unique definition of the term. In general, the term collocation refers to a sequence of words or terms that occur together more often than you would expect would happen through mere chance, like ‘take a shower’. Collocation imposes restrictions on how words can be used together; for example, ‘strong tea’ is considered to be acceptable, but ‘powerful tea’ is not (Li, 2005).
2.3. Types of collocations
Collocations have been categorized from different perspectives, since the term ‘collocation’ has been the subject of research in many areas, such as semantics, corpus linguistics and systematic linguistics. In one classification, many researchers agree that different collocation types should be placed on a continuum (Fontenella, 1994; Herbst, 1996; Howarth, 1998a; Nattinger & DeCarrico, 1992; Palmer, 1981). There are three criteria for categorizing collocations: semantic transparency, degree of substitutability, and degree of productivity (Carter, 1987; Howarth, 1998a; Nattinger & DeCarrico, 1992; Nesselhauf, 2004). Based on these criteria, if we consider collocations on a continuum, at one end we have free combinations with the greatest semantic transparency, the highest degree of productivity and substitutability. At the other end are idioms which cannot be interpreted from their literal meaning and are not easily productive or substitutable. Different types of restricted collocations fall between these two ends.
However, there are different classifications of collocations proposed by researchers which will appear in the following sections.
2.3.1. Collocational restrictions
Palmer (1981) proposes three kinds of collocational restrictions and classifies collocations based on the restrictions on words. These restrictions are as follows:
(1) Restrictions based fully on the meaning of the item, such as green cow.
(2) Restrictions based on range; i.e., a word may be used with a set of other words having some semantic features in common. This accounts for the unlikeliness and unacceptability of ‘the pretty boy’ (since we use pretty with words that denote females).
(3) Restrictions that are collocational in the strictest sense. Such restrictions involve neither meaning nor range, as addled with eggs and brains (p. 79).
2.3.2. Cowie and Mackin’s classification
In their Dictionary of Current Idiomatic English, Cowie and Mackin (1975) classify collocations into four categories: (a) pure idioms; (b) figurative idioms; (c) restricted collocations; and (d) open collocations. This classification is based on idiomaticity; pure idioms being the most and open collocations being the least fixed.
2.3.3. Wood’s classification
Wood (1981, cited in Li, 2005) classifies collocations into four categories, on the basis of a semantic criterion and a syntactic criterion: (a) idioms; (b) collocations; (c) colligations; and (d) free combinations. He believes that a collocation is less frozen than an idiom and a colligation such as ‘off with his head’ is more restricted than free combinations. On the other hand, a free combination is fully compositional and productive.
2.3.4. Lewis’s classification
Lewis (1997) categorizes collocations into four groups of strong, weak, frequent, and infrequent collocations. He believes that strong collocations, such as drink coffee, are more fixed and restricted than weak collocations such as nice weather. On the other hand, the distinction between frequent and infrequent collocations is based on their frequency of occurrence in a corpus.
2.3.5. Howarth’s categorization of lexical collocations
Howarth (1998b) distinguishes between the following types of lexical collocations:
(1) Free combinations, which derive their meaning from composing the literal meaning of individual elements and their constituents are freely substitutable, such as blow a candle.
(2) Restricted collocations, which are more limited in the choice of compositional elements, such as blow a fuse.
(3) Figurative idioms, which have a metaphorical meaning as a whole, which can be derived from the literal interpretation of that idiom, such as blow your own trumpet.
(4) Pure idioms, which have a whole meaning that cannot be derived from the meaning of its components, like blow the gaff.
2.3.6. Benson, Benson and Ilson’s classification
Researchers in the field of collocations widely adopt the classification of collocations proposed by Benson, Benson and Ilson (1997). They classify English collocations into two major groups of lexical and grammatical collocations. Lexical collocations consist of combinations of nouns, verbs, adjectives and adverbs. Grammatical collocations, on the other hand, consist of a noun, an adjective or a verb, plus a preposition or a grammatical structure such as an infinitive or a clause. In their classification, Benson, Benson and Ilson (1997) distinguish seven types of lexical collocations (Table 2.1); while grammatical collocations consist of eight groups (Table 2.2), the last group of which consists of 19 verb patterns (Table 2.3).
Table 2.1. Lexical collocations categorized by Benson, Benson, and Ilson (adapted from Chen, 2008, p. 18)
TypePatternExampleL1V (creation/activation) + N/ Pron./ Prep. Phrase1. Verb denoting creation:
come to an agreement, compose music
2. Verb denoting activation:
fly a kite, launch a missileL2V (eradication/cancellation) + Nreject an appeal, break a codeL3Adj. + N
N + Nstrong tea
aptitude testL4N + Vbombs explode, bees stingL5N (unit) + Na bouquet of flowersL6Adv. + Adj.hopelessly addictedL7V + Adv.apologize humbly
Table 2.2. Grammatical collocations categorized by Benson, Benson, and Ilson (adapted from Chen, 2008, pp. 18, 19)
TypePatternExampleG1N + Prep. (exclude N + of combinations)blockade against, apathy towardsG2N + to infinitive
(five most frequently encountered syntactic patterns)It was a pleasure to do it.
They had the foresight to do it.
They felt a compulsion to do it.
They made an attempt to do it.
He was a fool to do it.G3N + that-clauseHe took an oath that he would do his duty.G4Prep. + Nby accident, in advanceG5Adj. + Prep. (exclude past participle + by)They were fond of children.
They were hungry for news.G6Adj. + to infinitiveIt was necessary to work.G7Adj. + that-clauseIt was nice that he was able to come.G819 verb patterns(See Table 2.3)
Table 2.3. 19 patterns of the Group 8 Collocation Type (adapted from Chen, 2008, pp. 19-21)
G8PatternExampleAV + D.O. + to + I.O. = V + I.O. + D.O.
(Verbs allow the dative movement transformation)He sent the book to his brother: = He sent his brother the book.BVt. + D.O. + to + I.O.
(transitive verbs which do not allow the dative movement transformation)They returned the book to her: *They returned her the book.CVt. + D.I. + for + I.O. = V + I.O. + D.O.
(Transitive verbs with for allow the dative movement transformation)She bought a shirt for her husband. = She bought her husband a shirt. DV + Prep. + O
Compound V + prep.
*Excluded: free combinations
*Excluded: verb + by/with + ‘means’ or ‘instrument’to act as, to serve as
catch up to, break in on
*to walk in the park
*They came by train. We cut bread with a knife.EV + to infinitive
*Excluded: verbs are used in phrases of purposeThey begin to speak.
*He was running to catch a train.FModals + V
*Excluded: dare, help, needWe must work. He had better go.GV + V-ingThey enjoy watching television.HVt. + O + to infinitiveShe asked me to come.IVt. + O + VShe heard them leave.JV + O + V-ingHe kept me waiting two hours.KV + sb’s + V-ingPlease excuse my waking you so early.LV + that-clauseThey admitted that they were wrong.MVt. + D.O. + to be + Adj./pp/NWe consider her to be very capable.
We consider her to be well trained.
We consider her to be a competent engineer.NVt. + D.O. + Adj./pp/NShe dyed her hair red.
The soldiers found the village destroyed.
Her friends call her Becky.OV + I.O. + D.O.The teacher asked the pupil a question.PV + Adv./Prep. Phrase/N phrase/clauseHe carried himself well.
She put pressure on them.QV + wh- words + to infinitive/clauseHe asked how to do it. = He asked how he should do it.RV + NShe became an engineer.SV + AdjectiveThe soup tastes good.
The present study adopts Benson, Benson and Ilson’s (1997) classification because they gave careful considerations to lexical and grammatical collocations. In addition, the researcher attempts to investigate the performance of Iranian EFL learners on six types of grammatical collocations proposed by Benson, Benson and Ilson (1997).
2.4. Properties of collocations
Collocations are typically characterized as having the properties of being arbitrary, language specific, recurrent in context, and common in technical languages (Smadja, 1993).
2.4.1. Collocations are arbitrary
One of the properties of collocations is that they are arbitrary. It means if we replace one of the words in a collocational word pair with a synonym, it may result in an infelicitous combination. Thus, for example, a phrase such as warm greetings is acceptable, but hot greetings in which hot is substituted for its synonym warm is an unacceptable combination (Benson, 1989).
2.4.2. Collocations are language specific
Another property of collocations is that they are language specific, which means the arbitrary nature of collocations persists across languages. For example, the phrase régler la circulation in French refers to a policeman who directs traffic, the English collocation (McKeown & Radev, 1998).
2.4.3. Collocations are recurrent in context
In addition to being arbitrary and language specific, collocations occur frequently in similar contexts. Thus, while defining collocations is not an easy trait, it is possible to observe them in samples of the language. Generally, collocations are those combinations of words which co-occur frequently in the same environments, but do not include lexical items which have a high overall frequency in language (Haliday & Hasan, Cited in McKeown & Radev, 1998).
2.4.4. Collocations are common in technical language
The last property of collocations is that they are common in technical languages; i.e. words which do not take part in a collocation in everyday language often form part of collocation in technical language. Thus, when discussing computers, file collocates with verbs like save, delete, and create; but these collocations of the word ‘file’ do not exist in other sublanguages (McKeown & Radev, 1998).
2.5. Importance of learning English collocations
The field of EFL has seen a lot of studies on collocations. Many researchers have stressed the importance of collocations in language learning (Brown, 1974; Channel, 1981; Howarth, 1998b; Koosha & Jafarpour, 2006; Lewis, 2000; Nation, 2001; Pawley & Syder, 1983). Some believe that knowledge of collocations is one major component of fluency and such knowledge is essential for the mastery of communication and helps ESL/EFL learners become more native-like (see, for example, Bahns & Eldaw, 1993; Howarth, 1998b).
Brown (1974) believes that collocations are so important that learning them improves EFL/ESL learners’ language proficiency and oral fluency, as well as their reading speed and listening comprehension. She also emphasizes the role of learning collocations in understanding language chunks used by native speakers and so enabling learners to use those chunks in their performance. Like Brown (1974), Pawley and Syder (1983) also believe that the more collocations language learners use in useful chunks, the more they produce native like structures.
Nattinger (1988) argues that when learners learn collocations, it helps them in committing word combinations to memory and defining the semantic area of a word. This phenomenon enables learners to know and predict what kind of words would be found and used together. He also points out that learning collocations with pragmatic functions facilitates speaking and writing, since they shift learners’ attention from individual words to larger structures of the discourse.
Lewis (2000) argues that learning words in combination better helps language learners develop communicative competence than learning words in isolation. Nation (2001, cited in Li, 2005) considers collocation as an important factor in developing learners’ fluency and stresses that “all fluent and appropriate language requires collocational knowledge” (p. 15). He proposes that learning a word involves learning the set of other words that co-occur with it, i.e. its collocations. He believes that collocation knowledge is part of vocabulary knowledge; therefore, language learners should learn collocations in order to be able to produce the target language appropriately.
What can be concluded from the findings of previous studies is that collocation has an essential role in language learning; it not only helps learners achieve native-like fluency, but also enhances their language competence.
2.6. Overview of collocation research
Studies investigating collocations can be classified in terms of three approaches to collocation research: (1) lexical approach; (2) semantic approach; and (3) structural approach. The following sections explain these three approaches in detail.
2.6.1. The lexical approach
The lexical approach focuses on the idea that words receive their meaning from the words they co-occur with. This approach sees lexis as separate from grammar. The pioneer of this approach was Firth who developed this idea in 1957. After Firth, Halliday (1966) and Sinclair (1966) developed his theory and emphasized the importance of lexical collocations.
Halliday (1966) stresses that collocation is the co-occurrence of two words, independent of grammatical types and likely to take place over sentence boundaries. He added the term ‘set’ to the notion of collocations; a term which is defined as a group of words that collocate with one word. For example, hot, cold, rainy, and nice belong to the same lexical set, because they all collocate with the word ‘weather’.
Sinclair (1966) defines collocation as “the occurrence of two or more words within a short space of each other in a text” (p. 170, cited in Chen, 2008). He added three terms to the notion of collocation: (1) node, which refers to the lexical item under investigation; (2) span, which refers to the items on both sides of the node; and (3) collocates, which refer to those items found within the span.
Nesselhauf (2003) believes that researchers like Firth, Halliday, and Sinclair who adopted this approach defined collocation as ‘co-occurrence of all frequencies’ and ‘ a relationship between lexemes.’
In general, the lexical approach emphasizes lexical analysis as the best way to describe collocations.
2.6.2. Semantic approach
This approach focuses on the investigation of collocations on the basis of semantic framework without considering grammar. Lyons (1966) is one of the proponents of this approach who believes that the lexical approach was inadequate, because it failed to explain the reason why some lexical items collocate only with certain items. For example, why is it that black coffee is correct, but black tea is not.
The semantic approach claims that it is the semantic properties of the lexical items that are responsible for determining the words it collocates with. There were some criticisms toward this approach, because there are some collocations that are restricted. For example, there is nothing in the meaning of the word smoker that makes it collocate with heavy, rather than with strong or powerful.
2.6.3. Structural approach
It is the structural approach which emphasizes the significance of grammar in the study of collocations. It takes collocation to be determined by its structural patterns. According to this approach, lexis and grammar are inseparable and they complete each other (Bahns, 1993). Unlike the lexical and semantic approaches whose proponents only studied a small number of collocations (mainly verb + noun and adjective + noun collocations), the studies in the structural approach considered more patterns of collocations, like adverb + noun, adverb + adjective, and adjective + preposition.
With respect to grammar and lexis, Kjellmer (1992) established the extent to which an individual word class is collocational or non-collocational and showed that prepositions, articles, singular and mass nouns, and the base form of the verbs are collocational in nature; while adjectives, adverbs, and singular proper nouns are not. He concluded that English words are scattered across a continuum; those items whose contextual company is entirely predictable stand at one end, and those whose contextual company is completely unpredictable stand at the other end.
Regardless of any approach adopted, a number of studies were carried out about collocations. The following sections mention these studies in brief.
2.7. Empirical studies on English collocations
A number of experimental studies have been conducted on collocations in EFL (Aghbar, 1990; Bahns & Eldaw, 1993; Biscup, 1992; Chen, 2002 & 2008; Gitsaki, 1996; Huang, 2001; Koosha and Jafarpour, 2006; Kuo, 2009; Li, 2005; Mahmoud, 2005; Mirsalari & Shokouhi, 2010; Morshali, 1995; Yuan & Lin, 2001). Among these studies, the focuses were on measuring EFL learners’ knowledge of collocations, investigating the relationship between EFL learners’ knowledge of collocations and their language proficiency, finding the most difficult types of lexical and grammatical collocations, and finding the main sources of errors in learning collocations.
2.7.1. Measuring knowledge of collocations
Most of the studies mentioned above (Aghbar, 1990; Bahns & Eldaw, 1993; Channel, 1981; Chen, 2008; Gitsaki, 1996; Huang, 2001; Mirsalari & Shokouhi, 2010; Morshali, 1995; Yuan & Lin, 2001) aimed at investigating EFL learners’ knowledge of collocations. The findings of these studies show that EFL learners have insufficient knowledge of English collocations. Channel (1981) conducted a study on eight advanced EFL learners to investigate their knowledge of collocations, using a blank-filling test of collocations. She found that the learners’ productive knowledge of collocations was far below their receptive knowledge. She therefore argued that it is necessary to encourage learners to pay more attention to collocations. Aghbar (1990) used a cloze test to examine knowledge of verb-noun collocations by faculty members, native undergraduates, and advanced ESL students at Indiana University of Pennsylvania. He found that among the participants, ESL learners provided the least number of appropriate collocations.
In a similar study, Bahns and Eldaw (1993) investigated German learners’ knowledge of fifteen English verb-noun collocations, using a translation and a cloze task. The results of their study revealed that all the participants had insufficient knowledge of lexical collocations because of their poor performance on the tests. Farghal and Obiedat (1995) investigated Jordanian EFL learners’ knowledge of lexical collocations. They administered a blank-filling test of collocations to one group and an Arabic translation task to another group; the first group consisted of 34 seniors and juniors majoring in English, whereas the second group consisted of 23 seniors at a teacher training college. The findings showed that both EFL learners and English teachers had insufficient collocational knowledge.
Morshali (1995) conducted a study on the learning of English lexical collocations by Iranian EFL learners to find out the effect of proficiency level on collocation use. She found that Iranian EFL learners’ knowledge of collocations lagged behind their vocabulary knowledge. The analysis of the data also revealed that there was no significant relationship between language proficiency level and the knowledge of English collocations.
Gitsaki (1996) used three tests to measure collocational knowledge of 275 Greek junior high school students at three proficiency levels of post-beginning, intermediate, and post-intermediate. He used an essay writing task to measure learners’ productive knowledge of collocations, a translation task to measure cued production of collocations and a blank-filling task to measure accurate use of collocations. The analysis of the data revealed that L2 learners had deficiency in producing acceptable collocations.
In a more recent study, Yuan and Lin (2001) investigated collocational knowledge of 32 EFL learners majoring in English and 56 non-English majors at Ming Chuan University in Taiwan. The participants were given a translation test consisting of 15 verbal collocations. The results of the study showed that English majors only produced 34% of the target collocations, while non-English majors produced 30% of the collocations. They concluded that both groups lacked collocational competence.
The findings of Chen’s (2008) study were similar to the findings of Yuan and Lin’s (2001) study. He investigated English collocation competence of 440 non-English major freshmen in Taiwan and found that they did not show sufficient knowledge of collocations.
From the studies mentioned above, we can see that findings were quite consistent. It has been revealed that EFL learners, both English majors and non-English majors, have insufficient knowledge of collocations, mainly because of the fact that collocation has been neglected in EFL classrooms, and as a result, learners fail to learn collocations. This highlights the need for emphasizing collocation as the most useful type of prefabricated speech in EFL classrooms.
2.7.2. Types of collocation errors
Several studies show that certain types of collocations are more difficult for non-native learners to produce in their writings or in tests of collocations and point out which types of collocation errors may occur more frequently in EFL learners’ performance (Bahns, 1993; Biskup, 1992; Chen, 2002; Hassan Abadi, N. D.; Huang, 2001; Li, 2005; Mirsalari & Shokouhi, 2010; Nesselhauf, 2004; Zarei, 2002).
To investigate the difficulty EFL learners have with lexical collocations, Biskup (1992) compared collocational performance of 28 German and 34 Polish students of English and found that learners faced difficulties on the type of verbs. Moreover, German learners turned out to be more creative in deducing meaning of collocations and they did not make much transfer errors; whereas Polish learners were not risk-takers and showed more reliance on their L1 and thus made more transfer errors. He concluded that if we create native-like collocations in the learners’ mental lexicon, they will show better performance on collocations and their difficulty in producing collocations will be reduced. Bahns’ (1993) findings supported Biskup’s viewpoints, indicating that learners tend to use their L1 lexical knowledge in transferring to L2.
In a study conducted to analyze Taiwanese EFL students’ knowledge of English collocations and collocation errors, Huang (2001) designed a blank-filling test to measure 60 college students’ knowledge of four types of lexical collocations. The analysis of the data showed that while free combinations were the easiest types of collocations, pure idioms were the most challenging, and students performed equally well on the other two types, i.e. restricted collocations and figurative idioms.
Chen (2002) targeted high school students as his subjects to investigate their collocation errors in writing. The analysis of students’ writing samples showed that their writings included 272 collocation errors, out of which 147 errors were grammatical and 125 errors were lexical, revealing that grammatical collocations were more difficult than lexical ones. It was also found that in lexical collocations, errors of adjective + noun and verb + noun were the most frequent ones. On the other hand, preposition + noun and verb collocations were the most frequent types of grammatical collocation errors.
In a more recent study, Nesselhauf (2004) investigated advanced German-speaking learners’ performance of verb-noun collocations in free written productions and found that wrong choice of verbs and nouns, making wrong combinations, prepositional mistakes, and determiner mistakes were the most frequent kinds of collocation errors. Tang’s (2004) findings are similar to what Chen (2002) found in his study. He conducted a study which focused on the intermediate Chinese EFL learners’ collocation errors and found that adjective + noun and verb + noun errors were the most difficult types of collocations for the Chinese EFL learners.
Li (2005) conducted a study on 61 sophomores studying English at Ming Chuan University in Taiwan. He analyzed 76 copies of the participants’ writing samples to examine their collocation errors and found that errors of verb + noun and verb + preposition + object were the most frequent collocation errors in their writings, while adjective + infinitive errors were the least frequent ones. So, unlike previous studies (Chen, 2002; Tang, 2004) which had found errors of verb + noun and adjective + noun collocations to be the most difficult ones, Li (2005) found that in addition to verb + noun collocation errors, collocations of verb + preposition + object also pose the most difficulty to EFL learners.
In another recent study, Mahmoud (2005) found that 64% of the collocations written by Arab-speaking university students majoring in English were incorrect; 80% of which were lexical collocation errors as opposed to grammatical errors.
Kuo (2009) analyzed typical errors in collocation usage in writing productions by intermediate level EFL students in Taiwan and found three major types of errors, among which approximation was found to be more frequent (49.18%).
In recent years, some research has also been done to analyze collocation errors committed by Iranian EFL learners (Faghih & Sharafi, 2006; Hassan Abadi, N. D.; Mirsalari & Shokouhi, 2010; Zarei, 2002). Hassan Abadi (N. D.) examined the relationship between the performance of learners on lexical and grammatical collocations to find out which subcategories in different types of collocations are more problematic for Iranian EFL learners. A multiple-choice test of collocations consisting of 40 items was adopted and given to 80 Iranian freshmen studying English at Shiraz University. The analysis of the data showed that the learners’ knowledge of grammatical collocations lagged behind their knowledge of lexical collocations. He found that preposition + noun subcategory was the most difficult, while participle adjective + noun subcategory was the least difficult one for the learners.
Zarei (2002) studied over 5000 pages of 78 MA and Ph.D. students’ free productions to discover the patterns of Iranian advanced EFL learners’ problems with English collocations. He found many problems with English collocations among Iranian EFL learners. The findings of his study showed that while the collocations of prepositions were among the most problematic ones, adjective + adverbs and fixed expressions ranked among the least problematic collocation patterns.
Faghih and Sharafi (2006) selected over one hundred EFL learners from English-major juniors and seniors studying at AlZahra University and the Islamic Azad University of Torbat-Heydarieh. They administered a Michigan test to the entire group of the students to rank them as the high intermediate and the low intermediate learners. The students then took a multiple-choice test on collocations. The analysis of the data indicated that adjective + noun type of collocations pose the most amount of difficulty to Iranian EFL learners.
To find out which types of collocation errors are more frequent in Iranian EFL learners’ performance, Mirsalari and Shokouhi (2010), using a proficiency test, selected 35 students majoring in English translation and literature at Shahid Chamran University of Ahvaz from among 70 junior and senior students. The participants were then given a ninety-item multiple-choice test including four types of lexical collocations (noun + noun, noun + verb, verb + noun, and adjective + noun) and two types of grammatical collocations (noun + preposition, and preposition + noun). The findings of their study showed that the grammatical collocations were more difficult than the lexical ones, and noun + preposition errors were the most frequent and noun + verb errors were the least frequent type of collocation errors for the learners.
In sum, the findings of previous studies show that grammatical collocations have always been more difficult for EFL learners than the lexical collocations and certain types of collocations pose more difficulty to the learners. These findings also point out that in lexical collocations, errors of adjective + noun and verb + noun collocations are the most difficult ones, while in grammatical collocations, preposition + noun and noun + preposition collocation errors are the most frequent ones in EFL learners’ performance.
2.7.3. The relationship between EFL learners’ knowledge of collocations and their language proficiency
The literature on research about collocations lacks empirical studies on the relationship between the knowledge of collocations and general linguistic knowledge. It was not until recent years that such research on collocations in EFL was conducted (Faghih & Sharafi, 2006; Koosha & Jafarpour, 2006; Mirsalari and Shokouhi, 2010).
Faghih and Sharafi (2006) compared the mean and standard deviation of 100 Iranian EFL learners’ scores on the Michigan Proficiency Test with their scores on the collocation test. The results showed a positive correlation of 0.6946 between learners’ overall proficiency and their knowledge of collocations which indicated that there was a high relationship between the learners’ proficiency and their knowledge of English collocations.
Koosha and Jafarpour (2006) investigated Iranian EFL learners’ performance on collocations of prepositions. They selected 200 senior English-majors studying at three Universities in Shahrekord as their participants. A Michigan Test of Proficiency was administered to the participants to determine their level of proficiency. The participants then underwent two types of treatments. To check the effects of those treatments, two completion tests on collocations of prepositions were administered as the pre-test and post-test. The results of the study showed that learners’ performance on collocation of prepositions was positively related to their proficiency level.
In their study mentioned earlier, Mirsalari and Shokouhi (2010) also intended to see whether there was any correlation between Iranian EFL learners’ knowledge of collocations and their general linguistic knowledge. Unlike previous studies (Faghih & Sharafi, 2006; Koosha & Jafarpour, 2006), they found no significant correlation between Iranian EFL learners’ general linguistic knowledge and their knowledge of collocations.
What can be concluded from these findings is that collocations are considered as an important factor in determining EFL learners’ overall proficiency, since the use of collocations is highly correlated with their language proficiency.
2.7.4. Sources of committing collocation errors
Some studies have been conducted in EFL to point out the sources of committing collocation errors (Bahns, 1993; Biscup, 1992; Chen, 2002; Howarth, 1998b; Liu, 1999b). Biscup (1992) conducted a study focusing on finding the main sources of errors in collocations. She observed the performance of learners of English whose L1 was either close (German) or distant (Polish) from English to find out whether that distance would influence performance on L2 collocations. She selected advanced German and Polish students as participants in her study. The participants were asked to render native language collocations into English. The analysis of the data showed that Polish students relied more on their L1, but they produced fewer incorrect collocations compared with their German counterparts who seemed to look for more creative strategies. She came to this conclusion that the cause of collocation errors was related to L1 interference.
Similar conclusion was made by Bahns (1993) and Farghal and Obiedat (1995). They argue that the majority of EFL learners’ collocation errors can be traced back to L1 influence. Later, Howarth (1998b) made new conclusions. He conducted a study comparing the writing samples of native speakers and EFL learners and noticed that the latter made less conventional collocations and more deviant ones. He concluded that the production of overlapping collocations was the main source of EFL learners’ collocation errors.
Liu (1999b, cited in Li, 2005) analyzed Taiwanese EFL learners’ writings to find the causes of their collocation errors. She found seven main factors leading to their collocation errors and divided them into to categories: (1) intralingual