The Use of Grammatical Collocations by Advanced Saudi EFL Learners in the UK and KSA

This study attempts to investigate the production of English grammatical collocations amongst Saudi students majoring in English in the KSA and those in the UK. It also shows the most frequent types of errors that may occur as well as some possible reasons for their occurrence. For this purpose, the researcher analysed essays written by the participants. The results reveal that Saudi EFL learners in the UK do grammatical collocation errors less than those who learn English in the KSA. Additionally, the highest number of errors in both groups was recorded on the grammatical collocations patterns, noun + preposition and adjective + preposition. It seems that L1 interference plays a crucial role in students' erroneous responses, especially those which contain a preposition. For instance, the majority of noun + preposition, adjective + preposition and preposition + noun are used incorrectly throughout the essays. Furthermore, the avoidance phenomenon in SLA may be used by Saudi students. They tend to avoid using some grammatical collocation categories such as adjective + thatclause and noun + that-clause since they are beyond their English proficiency level. Finally, the lack of knowledge of grammatical collocations is another possible reason behind such errors. Educational leaders, curriculum designers and teachers need to shed light on these types, especially as the English language curricula used in the KSA do not pay a great deal of attention to grammatical collocations.


Introduction
An important aspect of language learning is learning the vocabulary of that language and its appropriate use.Within the area of vocabulary, researchers have stressed the importance of word combination, especially collocations.Benson, Benson and Ilson's (1986) classification has drawn researchers' attention to collocations.For instance, McCarthy (1990, p. 12) pointed out that "in vocabulary teaching there is a high importance of collocation …… the relationship of collocation is fundamental in the study of vocabulary, and collocation is an important organising principle in the vocabulary of any language".Therefore, this study is a serious attempt to examine the problems that Saudi EFL learners encounter in using grammatical collocations in the UK and KSA.The ultimate goals are to find out which grammatical collocation patterns are the most problematic as well as identify the reasons beyond the errors.

What is Collocation?
Many definitions of the term collocation have been proposed in the literature.Firth (1957, p. 183), who first introduced the term in his work, defined collocation as "the company that words keep".Additionally, McCarthy (1990, p. 12) defined collocation as "a marriage contract between words, and some words are more firmly married to each other than others".Also, Woolard (2000, p. 29) defined collocations as "the co-occurrence of words which are statistically much more likely to appear together than random chance suggests".Moreover, Nation (2001, p. 317) explained that the term collocation is used to refer to a group of words that come together, either because they commonly occur together like take a chance, or because the meaning of the group is not obvious from the meaning of the parts, as with by the way or to take someone in.Additionally, according to Oxford Collocation Dictionary (2002), collocation is a means of combining words in a language to produce natural-sounding speech and writing.Incorrect combinations such as heavy wind or strong rain do not sound natural in English.

Features and Classification of Collocations
In terms of the features of collocations, Boonyasaquan (2005, pp. 11-13) summarised the main features of collocations as follows: 1-Collocations are frequent co-occurrences of items between which no word can be added.For example, in knife and fork, it is not usual to add a word to this collocation like, knife, spoon and fork.
2-Collocations consist of components that cannot be replaced by a synonym or word of similar meaning.For example, John makes a cake; but not John makes a pancake.
3-Collocations are binomials that cannot be reversed.The order of a collocation is more or less fixed, for example, bread and butter not *butter and bread.
4-Some collocations can be predicted, for example, if a person hears a collocation apply and shrug, s/he automatically expects that for and shoulder will follow respectively.
It is worth noting that collocations have specific features; their components cannot be replaced or changed.
In terms of their types, Hill (2000) classified collocations based on their co-occurrence as follows: 1-Unique collocations: these collocations are fixed and cannot be replaced by any other words, such as to foot the bill but to foot the invoice is obviously wrong.
2-Strong collocations: these collocations are very strong but not unique.They usually have few other collocates, for example, moved to tears or reduced to tears.
3-Weak collocations: they refer to a number of word co-occurrences that can be easily guessed, such as a white shirt, a blue shirt, a small shirt, a long shirt, etc. 4-Medium-strength collocations: these collocations can sometimes be weak collocations such as to hold a conversation and to make a mistake.Learners already know each individual word like to hold and a conversation, but they are unable to use the words together as a collocation.
Additionally, collocations fall into two major groups: grammatical collocations and lexical collocations (Benson et al., 1997, p. 20).With regard to grammatical collocations, they consist of a noun and adjective or verb plus a preposition or a grammatical structure, such as an infinitive or a clause.Benson et al. (1997) distinguished between several structural types of grammatical collocations as follows: 1-Noun + preposition e.g.blockade against 2-Noun + to-infinitive e.g.He was a fool to do it.
3-Noun + that-clause e.g.He took an oath that he would do his duty.4-Preposition + noun e.g. by accident, in advance.
6-Adjective + to -infinitive e.g. it was important to work.
7-Adjective + that-clause e.g.she was afraid that she would fail.
In terms of lexical collocations, the main combinations of them are nouns, adjectives, verbs and adverbs.The obvious difference between lexical collocations and grammatical collocations is that the former do not contain prepositions, infinitives or clauses.Benson et al. (1997) made a distinction between several structural types of lexical collocations as follows: 1-Verb (which means action) + noun /pronoun/ prepositional phrase e.g.inflict a wound, come to an agreement.

7-Verb + adverb e.g. appreciate sincerely, apologize humbly
Here, it is worth noting that Benson et al.'s (1997) classification is universally accepted and it is also widely used in many studies on collocations such as (Shamsudina, 2013).

Major causes of Collocational Errors
Several researchers (Tarone, 1981;Farghal & Obiedat, 1995;Liu, 1999) conducted several studies on the use of collocations by EFL learners to identify areas of difficulty and possible causes of errors.They discovered that the main reasons for errors are related to paraphrasing, overgeneralization, L1 interference, interlingual transfer, intralingual transfer, and lack of collocational knowledge in English.For instance, Howarth (1998), who built a corpus that contains essays written by EFL learners and native speakers of English, stated that both groups have insufficient knowledge of collocations in general.When compared with native speakers of English, EFL learners produced a lower number of conventional collocations but a higher one of incorrect combinations such as pay effort.Additionally, Gitsaki (1997) used essay writing, a translation test, and a blank-filling test to measure collocational knowledge of 275 Greek EFL learners in junior high school at three proficiency levels of post-beginning, intermediate, and post intermediate.The results showed that L2 learners had deficiency in producing acceptable collocations and that the development of collocational knowledge was found to be affected by L1-L2 differences.Grammatical and lexical collocations that were simple and frequent in everyday use were acquired early and more complex grammatical collocations were acquired later.Hill (1999, p. 4) stated that lack of collocational competence can be a cause of EFL learners' problems in learning English collocations.One reason for the learners' problems in learning English prepositions is that they try to learn the meaning and use of prepositions individually without paying attention to their collocational properties (Flowerdew, 1999).Zarei (2002) pointed out that Iranian EFL learners have problems with English collocation.He categorised English collocations into ten types, namely, noun + verb, adjective + noun, verb + noun, noun + noun, combination of all patterns, idioms, fixed expressions, adjective + noun, verb + noun.He found that most problematic categories for Iranian EFL learners are 'adjective + noun', 'noun + noun', 'noun + verb', 'verb + noun' combinations, and idioms and fixed expressions.
Finally, Koya's (2005) study showed that Japanese learners had difficulty with some collocation categories, such as adjective-noun collocations, compared to others.Koya (ibid) concluded that there are several factors which influence the development of collocational knowledge.These factors include the differences between L1 and L2, arbitrariness, maturation, complexity, language proficiency, teaching, and saliency.

The Importance of Learning Collocations
The comprehension and production of collocations are considered the most important areas in L2 learning process.Fillmore (1979) regarded fluency as a genetic term, which contained all properties of speakers' competence and performance in language.Furthermore, Kane (1983) pointed out that the significance of collocations is due to making the spoken and written language more stimulating and alive.In addition, Aghbar (1990) emphasized the importance of collocations and stated that knowledge of formulaic language (proverbs, idioms, and collocations) is necessary in the overall fluency of the language of non-native speakers.In other words, the mastery of collocations helps EFL learners become native-like (Bahns & Eldaw, 1993;Howarth, 1998).
However, the acquisition of collocations is not as simple for EFL/ESL learners as it is for native speakers of English.Palmer (1979) stressed that collocations are problematic for both native and non-native speakers of English.In fact, McCarthy (1990, p. 13) pointed out that "even very advanced learners often make inappropriate or unacceptable collocations".Nattinger (1988) concluded that knowledge of collocations can aid learners in memorising and defining the semantic area of a word.It also helps learners predict what kinds of words would be found together.He pointed out that teaching collocations will lead to fluency in speaking and writing because it shifts learners' concentration from individual words to larger structures and to the social aspects of the interaction.Moreover, Crystal (1992, p. 105) stated that; Collocations provide a major difficulty in mastering foreign languages.Nevertheless, the more fixed a collocation is, the more we think of it as an 'idiom', a pattern to be learned as a whole, and not as the 'sum of its parts.Bonk (2000) reported that some collocations may be difficult to be produced correctly even by some of the best language learners.The best language learners may make mistakes in producing L2 collocations or sometimes are not sure whether a certain word combination is possible or not.
In sum, based on the findings of the previous studies, collocations play a crucial role in learners' L2 fluency and accuracy, language appropriateness and word knowledge.

Statement of the Problem and Purpose of the Study
Much research has been conducted on the acquisition of English collocations by EFL learners.Based on the researcher's experience, being a teacher in the KSA, the methods of teaching English to Saudi EFL learners in general are mainly based on syntactic principles.The grammatical aspects of English are overemphasized.Therefore, learners can produce sentences that are grammatically correct, but contain mistakes because of incorrect use of word combination.Also, the focus on drills or repetition of individual words rather than chunks and collocations makes it difficult for Saudi EFL learners to express their ideas perfectly and communicate effectively with native speakers of English.As a result, grammatical and lexical collocations are considered to be the most frequent aspects of language learning and language teaching.
Therefore, this study attempts to explore the familiarity of Saudi EFL learners in the UK and the KSA with grammatical collocation.Their knowledge was measured through analysing their errors when using grammatical collocation patterns in written essays.As a result, suggestions and solutions were provided to account for these errors.This study focused on grammatical collocations rather than lexical ones as Saudi EFL learners encounter more problems with the first type.This may be due to many reasons such as the unclear use of these sorts of collocations in schools.After having a wide revision of the materials which have been taught to Saudi students, it could be noticed that few activities and tasks tackle this issue.It is always believed that students pick up these patterns through exposure to or use of English language.Hence, the results of this study may help language teachers and curriculum designers identify areas of difficulty and account for errors and problems in learning collocations.

Research Questions and Hypotheses
The initial hypothesis is that performance of Saudi EFL learners who study in the UK is better than those who study in the KSA in terms of the amount of errors and grammatical collocation patterns they produce.The two groups have been chosen to be in the KSA and the UK because in the former, students are not exposed to English in the same degree compared to the latter.The researchers were keen on seeing the effect of the exposure on the learning of collocations.
The present study addressed the following questions: A) Do Saudi EFL learners in the KSA encounter problems in the use of grammatical collocations more than those learning in the UK? B) Which type of grammatical collocation among the ones selected (Benson et al., 1997) is the most problematic for these groups of language learners?C) What are the main reasons behind the grammatical collocations errors which Saudi EFL learners do?

Participants
Two groups, consisting of a total of 10 Saudi undergraduate students, at stage two, participated in the study.Five of them study at the school of English at Manchester Metropolitan University in the UK, whilst the other five study at Najran University in the KSA.There was no much difference in their proficiency level as all of them scored either 5.5 or 6 in IELTS.Their essays, which were gathered to collect the data, show a good level of proficiency in English, so it can be argued that all participants have a minimum level of proficiency to be undergraduate students majoring in English language.
In terms of ethical issues, the researcher reassured the participants that their data will be handled with care and consideration.In particular, the researcher informed them that no one will have access to the participants' assignments and their work will never be used for any other purposes apart from this research.A coding system was put in place to refer to each student, which would hopefully protect identities.The essays were kept in password-protected documents and will be discarded when no longer required.

Data Collection
In order to investigate the participants' grammatical collocation knowledge, ten assignments were gathered from the participants to be analysed based on an Error Analysis approach (henceforth EA).The participants were asked to provide me with previous essays that ranged between (1800-2000 words) on slightly different topics in the same genre so they generally have to produce the same amount of words.The slight differences in the topics may affect the quantity of the grammatical collocations produced in each essay.For example, some areas may require the use of more grammatical collocations in comparison with others.Yet, the essays were still written in the same genre.Both groups had the same choices of topics under the same genre.Therefore, it can be claimed that there was no big distinction between the essays produced by both groups.In addition, the researcher consulted native speakers of English to check and confirm any of the observations included in the data.
The data was collected from the participants' assignments as essays are considered a rich source of data for research (Punch, 2005, p. 184).Also, it is clear that the data is authentic since the participants handed in their previous assignments without previous knowledge of them being used in this study.The assignments were read several times to detect any grammatical collocation error.These errors were highlighted and then counted in each paragraph.Then, the errors were classified according the seven patterns of collocation on the basis of Benson et al.'s (1997) classification of grammatical collocations as mentioned in section 2.2.Meanwhile, the correct ones were written in two brackets.Next, the total number of errors in each assignment was counted separately.Before counting the errors detected in the essays, a native speaker of English was asked to check the categorisation used to divide them into groups.These errors were linked to the initial letter of the participants' names, so it was easy to distinguish between the participants' errors to observe, analyse, classify and compare and contrast them later.

Quantitative Approach
The researcher adopted a quantitative method to test the hypothesis and calculate the numerical differences between the two groups.Statistical analysis of grammatical collocation errors, which were dealt with from a quantitative perspective, measures the frequency of the highest level of grammatical collocation patterns.
The quantitative method in this study allowed the researcher to collect the data from the respondents in a numerical format, to exercise objective judgments and to achieve a high level of reliability and accuracy (Seliger & Shohamy, 1989).Employing the quantitative method also allowed the researcher to arrive at a more objective conclusion by minimising the subjectivity of judgment as can be observed, especially when we look at the most frequent problematic pattern Saudi EFL learners encounter as follows.

Error Analysis (EA)
EA has been chosen to detect and then analyse the grammatical collocation errors as it was recommended by Mitchell and Myles (2004, p. 29-30).This approach, in Second Language Acquisition (SLA), studies the types and causes of language errors which are based on the level of proficiency in language skills, linguistic levels such as vocabulary and grammatical forms i.e. insertion and omission, type such as errors in performance and competence, and cause such as interference (Mitchell & Myles, 2004, p. 28-31).According to Ellis (2003, p. 47), EA studies the types and causes of language errors (perhaps they can be defined as deviation from the norms of L2 or they may result from lack of knowledge of L2).As Ellis (2003, p. 133) described, EA explains learners' language in relation to its difference from L2, therefore, it merely shows any external variation.According to Saville-Troike (2006, p. 38-39), in EA, L2 learning was viewed independently of L1.However, EA does not provide a complete picture of how learners acquire language, because it describes learners' language purely as a collection of errors.The most likely causes of errors in SLA are inter-lingual (between languages) factors; resulting from negative transfer or interference from L1 to L2, and intra-lingual (within language) factors.Saville-Troike (2006, p. 40) added that the absence of errors may result from learners' avoidance of difficult structures.EA assumes errors are contributing to language learning, whereas, in fact, they can have a negative influence on learners' social communication/ role.The EA approach will help analyse the obtained data.It may derive accurate results and lead to a reliable and logical conclusion.

Data Analysis and Statistics
The data was calculated and analysed in terms of percentage based on the types of grammatical errors.Also, the data was compared and contrasted in order to see the extent to which some grammatical collocation errors were more common for each group.Thus, each essay has been read several times in order to count the grammatical collocations used by the participants.After that, all grammatical collocations used have been written on a separate sheet to be checked in the Oxford Collocation Dictionary (2002).The reason for this process is to ensure that they were used properly by the participants.Additionally, to confirm the results obtained by the researcher based on the essays, three native speakers of English were consulted about any possible grammatical collocations that may have been used throughout the essays and also whether they were used correctly or not.
The EA approach applied to the students' essays is designed to examine the above points and to provide data showing which type of errors the participants tend to make.The observation of some errors pertaining to Arabic is conducted to see the extent to which L1 interference plays a role in producing some grammatical collocation errors.It could be expected that some patterns of these errors tend to occur more than others.Also, it could be suggested that there might be a difference between errors made by Saudi EFL learners in the UK and those in the KSA.These points are explored in the following section.

Saudi Students Learning English in the KSA
The data has been coded and meta-coded so as to be sorted into categories for discussion.The categories below are arranged with a view to answering the research questions.The data is analysed based on the following classification of seven patterns of grammatical collocations by Benson et al. (1997).
Through identifying the errors done by the participants, the following findings can be seen.Table 1 shows the percentage of grammatical errors by Saudi students studying in the KSA.Firstly, the number of grammatical collocation errors which occurred throughout the five essays was of forty-eight different errors in all types out of the total number of grammatical collocations used which was one hundred-sixteen.The word count of the essays was between ten to eleven thousand words.The findings showed that the participants tend to do grammatical collocation errors mostly on the noun + preposition pattern collocation as seen in table 2. Errors on noun + preposition pattern were the highest (45.8%) out of the total percentage of errors, followed by the adjective + preposition (18.7%), the preposition + noun (14.5%), the noun + to-infinitive (10.4%), the adjective + to-infinitive (8.3%), the adjective + that-clause (2.1%), and the noun + that-clause (0%) respectively.Based on table 2, most of the participants' errors were on the noun + preposition pattern, whereas the lowest percentage of errors was on noun + that-clause pattern.
Based on EA approach, some errors tend to occur because of L1 interference.This can be seen in the following quotation taken from errors found in the participants' essays: a) "The reaction of the teachers should be *suitable to the students who ………..……" First of all, the error above is an example of adjective + preposition pattern.The use of the grammatical collocation *suitable to instead of suitable for in the above context is due to L1 interference.Suitable for + noun can be literally translated into Arabic (See Appendix A for full description of Arabic sounds) as seen in the following: 1. haðaa laisa munaasib la-k this not suitable to-you 'this is not suitable for you' When translated into English, the preposition ʔila in Arabic means to.The participants may have chosen to use to as it is a possible translation to this pattern in Arabic.Also, the same problem appears in the following example: b) "Many students show negative *attitude from a given subject due to being demotivated …………….." This error found in example (b) is an example of noun + preposition pattern.The use of the grammatical collocation * attitude from instead attitude to may be due to L1 interference; the participants may have translated the Arabic equivalent literally into English.The equivalent of attitude to in Arabic is as mawqif min.
2. ʃ-ʃaʕb ʔtaxaða mawqif min l-hukuumah the-people took attitude from the-government 'the people's attitude to the government was negative' The preposition min in Arabic can be translated into English as from.The participant may have chosen to use from as it is the only translation to this pattern in Arabic.It can also be observed that sometimes there is an absence of some prepositions in Arabic with regard to some patterns.In other words, there is no one-to-one equivalence in Arabic and English with respect to some prepositions.This can be observed in the following example: c) "Students should be informed *on advance that ………………." This error could also have happened due to L1 interference.The use of *on advance instead of in advance can be caused by L1 interference but in an indirect way.For instance, there is no equivalent to in in the Arabic translation of the collocation in advance as follows: 3. ana dafaʕt l-mablaɣ muqaddaman

I paid the-amount advance 'I paid the amount in advance'
Since the Arabic equivalent is not existent, it can be claimed that the preposition found in the collocation in advance has no equivalent in Arabic, so the participant may have chosen on by chance.
Finally, Saudi EFL learners seem to avoid using Noun + that clause.In this study, the participants seem to avoid two types of grammatical collocations, namely, adjective + that-clause (2.1%) and noun + that-clause (0%).For instance, one of the participants could have used 'took an oath that he ……' instead of 'swore that he'.Also, they could have used 'it was imperative that these issues were dealt with quickly………' instead of 'these issues were dealt with urgently and quickly as they are very important'.This avoidance can be because of two reasons.The first one is that they know that if they attempt to use it, they may get it wrong, so they choose the easy, or shall we say the safe alternative.The second one is that they do not know such grammatical collocations at all as some of them are not used in Arabic such as noun + that-clause and adjective + that-clause.The second claim can be suggested due to the fact that there are few statements generated by the participants in their essays.

Saudi Students Learning English in the UK
This section shows the number of grammatical errors by Saudi students studying in the UK as seen in the following table.
Firstly, the number of grammatical collocation errors which occurred throughout the five essays is twenty-nine different errors.The percentages of these patterns were calculated out of the total grammatical collocation errors made by the participants as mentioned in table 4. The findings showed that the participants tend to make grammatical collocation errors on the noun + preposition pattern.This type has the highest percentage of errors (58.6%), followed by the preposition + noun (17.2%), the adjective + preposition (10.3%), the noun + to-infinitive (6.8%), the adjective + to-infinitive (6.8%), the adjective + that-clause (0%), and the noun + that-clause (0%) respectively.Based on table 2, most of the participants' errors were on the noun + preposition pattern, whereas the lowest percentage of errors was on the noun + that-clause pattern.
On the basis of EA approach, some errors tend to occur because of L1 interference.The following example, taken from the participants' essays, illustrates this problematic issue: d) "If students are *good in English, they feel happy ………….." Firstly, the error above is an example of adjective + preposition pattern.The use of the grammatical collocation *good in instead good at is due to L1 interference.The idea is that the grammatical collocation good at can be literally translated into Arabic as jayyid fii as seen in example 4: e) "*By reference to the language competence and performance, students vary ………….." On the other hand, some errors do not occur due to L1 interference.Seven errors out of twenty-nine done by Saudi students studying in the UK were due to other reasons apart from L1 interference.The percentage of these errors is (24.13%).Some examples of these errors are as follows: Again, Saudi EFL learners in the UK seem to avoid using two types of grammatical collocations which are adjective + that-clause (0%) and noun + that-clause (0%).As mentioned above, this can be due to the fact that they are beyond their level.In other words, they know they have difficulties with it.An example which shows this avoidance is: 'many researchers were astonished that this idea………' rather than 'this idea is always astonished by many researchers ……..'.Also, it can be claimed that this kind of avoidance may not deliberately occur only because students know they have difficulties with it.Rather, this may happen because they do not know such grammatical collocations at all as there were few statements found in their essays.

Comparing the Performance of Saudi EFL Learners in the KSA and UK
Based on the analysis of the data, it is clear that the noun + preposition pattern has the highest percentage of errors out of the total errors in both groups.With regard to Saudi EFL learners who study in the KSA, the percentage was (45.8%) in comparison with those who learn English in the UK (58.6%).L1 interference plays a central role in causing many errors, especially those which have a preposition as a part of their patterns.Many examples are analysed and tested based on EA approach.Overall, the majority of noun + preposition (attitude to), adjective + preposition (good in) and preposition + noun (in advance) were used incorrectly throughout the essays.
Also, Saudi EFL learners in the UK made less grammatical collocation errors than those who learn English in the KSA.The first group made twenty-nine errors in comparison with the other one which made forty-eight.This indicates that exposure to English can help students minimise their grammatical collocation errors as there was no difference between these two groups except that one group studies in the UK, while the other in the UK.Additionally, the researcher tried to account for the participants' errors.It is clear that L1 interference plays a central role in causing many errors, especially those which have a preposition as a part of their pattern.For instance, the majority of noun + preposition, adjective + preposition and preposition + noun were used wrongly throughout the essays.The variety and difference of meanings of each preposition in Arabic may have contributed to producing plenty of errors.The most prominent prepositions in Arabic which cause this problem are fii 'in' and ʔila 'to', as explained in the previous two sections.Also, some errors are not connected to L1 interference.The percentages of these errors were (25%) for those who study in the KSA and (24.13%) for those who study in the UK.Finally, the avoidance phenomenon in SLA can be another cause for not using some grammatical collocation patterns such as adjective + that-clause and noun + that-clause.Saudi EFL learners do not use some types of grammatical collocations since they know they have difficulties with them or because they are beyond their level.The percentages were almost (0%).
Furthermore, a lack of knowledge in grammatical collocations is also a main reason behind such errors.Educational leaders, curriculum designers and teachers need to take into account that some grammatical collections are more problematic for Saudi EFL learners.Thus, they have to shed light on these types, especially as English language curricula used in the KSA do not pay a lot of attention to grammatical collocations.
Finally, in terms of the number of correct uses of grammatical collocations, Saudi EFL learners in the KSA did sixty-eight out of one hundred-sixteen.In other words, the percentage of the correct uses on all types is (58.6%) in comparison with the percentage of the incorrect uses which is (41.4%).Those who study in the UK, Saudi students did one hundred and two out of one hundred thirty-one.In other words, the percentage of the correct uses is (77.9%) in comparison with the percentage of the incorrect uses which is (22.1%).Clearly, it can be suggested that those who study in the UK tend to use grammatical collocations more than those in the KSA.As a result, Saudi EFL learners in the UK have a higher percentage of correct uses of grammatical collections these types of collocations, than those in the KSA.Here, it can be argued that exposure can play a crucial role in the acquisition of grammatical collocations in English.Saudi EFL learners in the UK are exposed to English language more than those in the KSA.

Conclusion
It is clear that Saudi students learning English in the UK perform better than those who learn English in the KSA in relation to the use grammatical collocation errors less.The noun + preposition and adjective + preposition patterns have the highest level out of the total errors for both groups in the UK and KSA.L1 interference (cf.Gitsaki, 1997), avoidance and lack of grammatical collocations knowledge are the main reasons for these errors.Overall, teaching collocations, namely grammatical collocations, is one of the most important areas that is still open to debate.Students' knowledge of English grammatical collocations differs not only based on their L1, but also the way they have been learning these collocations.Students need to be taught cautiously how to use collocations properly.

Recommendations
It might be suggested that educational leaders, curriculum designers and teachers need to take into account that some grammatical collocations are more problematic for Saudi EFL learners, especially those who study in the KSA.They need to find the appropriate strategies, methods and techniques to minimise these errors.Teachers need to choose carefully the activities that focus on the most problematic types of grammatical collocations.Several types of activities can be used, such as textual analysis activities, preparation activities, speaking activities, and dictionary and matching activities.Regarding textual analysis activities, students can be asked to analyse texts in order to draw their attention to grammatical collocations.The students can be asked to find, for example, five useful collocations that occur around a specific topic.It is also possible to go beyond the text and ask them to find further possible collocations with certain items in the text using a collocation dictionary.Additionally, it can be suggested to design gap-fills based on authentic texts.Another suggestion is prediction exercises.This can be done through using a kind of word association technique.A text could be revealed gradually to the students using an overhead projector.The students' mission would be to predict the next word or phrase.Also, some teachers are not qualified enough to use technology to help students enhance their knowledge in collocations.Ministries of Education in the Arab world have recently started providing new teachers with sessions, workshops and seminars that can help them enhance their teaching methods as well as knowledge in technology.Also, teachers need to encourage students to communicate with native speakers and be exposed to English language.As argued above, exposure helps to minimise grammatical collocation errors.For instance, students can be encouraged to visit the UK, the USA, Canada or Australia for a few months to practice their English.
With respect to further research, it could be vital to conduct another study about the acquisition of collocations by students at other levels such as the secondary school or graduate levels.A comparative study of undergraduate and graduate students' knowledge of collocations, either lexical, grammatical or both can be an excellent area of research.Such a comparison may shed light on the causes of grammatical and lexical collocational errors.They can also show the extent to which these types are problematic for Saudi EFL learners.Finally, the difference between the Saudi learners' productive and receptive knowledge of English grammatical collocations is still open to debate since no study has been conducted in this area so far.
good at writing' The preposition fii in Arabic can have two meanings in English which are at or in.The participant may have chosen to use in as it is a possible translation to this pattern in Arabic.Moreover, the following error shown in example (e) is an instance of preposition+ noun pattern.The use of the grammatical collocation * by reference to instead of with reference to is due to L1 interference.The equivalent of with reference to in Arabic is bi lʕawdah ʔila.The preposition bi in Arabic can be translated with two different meanings in English, namely, by or with.The participant may have chosen to use by as it is a possible translation to this pattern in Arabic.

Table 1 .
The number of grammatical errors by Saudi students studying in the KSA

Table 2 .
The percentage of errors on each type by Saudi students studying in the KSA

Table 3 .
The number of grammatical errors by Saudi students studying in the UK

Table 4 .
The percentage of errors on each type by Saudi students studying in the UK