T.C.
SELÇUK ÜNİVERSİTESİ SOSYAL BİLİMLER ENSTİTÜSÜ
YABANCI DİLLER EĞİTİMİ ANABİLİM DALI İNGİLİZCE ÖĞRETMENLİĞİ BİLİM DALI
AN EFFECTIVE WAY OF TEACHING VOCABULARY:
COLLOCATIONS
YÜKSEK LİSANS TEZİ
DANIŞMAN
YRD. DOÇ. DR. ECE SARIGÜL
HAZIRLAYAN GÖKHAN ŞİMŞEK
TABLE OF CONTENTS
ACKNOWLEDGEMENTS………...i
ABSTRACT………...ii
CHAPTER I – INTRODUCTION……….………...………..1
1.1. Background of the Study ………..1
1.2. Statement of the Problem ……….2
1.3. Goal and Scope of the Study ……….3
1.4. Significance of the Study ………...4
1.5. Research Questions ………...5
CHAPTER II – LITERATURE REVIEW ………...……….6
2.1. Introduction ………...6
2.2. Vocabulary Teaching in ESL / EFL Settings ………..7
2.2.1. Historical Overview ………....………...10
2.3. The Definition and Categorization of Collocation ………..…….16
2.3.1. Definitions of Collocation ………16
2.3.2. Types of Collocations ………...…24
2.3.3. Classification of Collocations ………..…26
2.4. The Importance of Collocations in EFL Context ……….29
2.5. Difficulties in Collocational Use ……….33
2.6. Empirical Studies on Collocations ……….36
CHAPTER III – METHODOLOGY ……….……….….40
3.1. Introduction ……….40
3.2. Subjects ………..………..40
3.3. Materials and Procedure ………....40
4.1. Data Analysis Procedures ………...………...44
CHAPTER V – CONCLUSION ………...………48
5.1. Conclusions and Discussions ………..………48
VI. BIBLIOGRAPHY ………..……...………...…52 APPENDICES ………55 APPENDIX A ………...………..55 APPENDIX B ……….………57 APPENDIX C ……….………60 APPENDIX D ……….………61 APPENDIX E ……….………63 APPENDIX F ………..………64 APPENDIX G ……….………65 APPENDIX H ……….66 APPENDIX I ………..68 APPENDIX J ………..69
ACKNOWLEDGEMENTS
I would like to express my deepest gratitude and special thanks to my advisor Assist. Prof. Dr. Ece SARIGÜL who helped and guided me in completing my thesis.
My thanks goes to Lec. Gülgün SERTKAYA and Dr. Osman DÜLGER for being so kind to let me implement the data analysis procedure in their lessons.
I owe my special thanks to my mother and father for their endless help, patience and encouragement.
Finally, I would like to thank my cousin Hasan BABALIK who has been a great support to me for parts especially related to the statistical analysis.
ABSTRACT
In foreign language teaching vocabulary has been a neglected area for a long time. Since 1970s, the perspective on vocabulary teaching has changed. It must be highlighted that as far as communication is concerned, vocabulary is just one of the components in the whole system. Vocabulary should be recognized as a central element in language instruction from the beginning stages. With the recognition of the importance of vocabulary, many techniques and approaches to teaching and learning vocabulary have emerged. One of these is teaching vocabulary through collocation.
Teaching vocabulary through collocation is a comparatively new technique. The purpose of this study is to investigate whether teaching vocabulary through collocations will result in better vocabulary learning than teaching vocabulary using classical techniques such as using definition, synonym, antonym and mother tongue translation. The study was conducted at Selçuk University, Faculty of Education, ELT Department. The participants were 79 undergraduates of first-graders.
In the first chapter, the background of the study, statement of the problem, goal and scope of the study, siginificance of the study and research questions are presented.
The second chapter includes literature review which centers on the place of vocabulary in foreign language teaching and the definition and types of collocation.
In the third chapter, the method of the study, participants, materials, procedure and information about the statistical analysis are given.
CHAPTER I : INTRODUCTION Introduction
The significance of vocabulary was not noticed by the people who devoted themselves in linguistics such as teachers, theorists, researchers and the others involved in EFL . But fortunately, by the help of scholars who managed to address English as a Foreign Language (EFL) teachers by their experimental studies and pedagogical material provided an arising interest in vocabulary teaching and this has been one of the most valued issues of English Language Teaching since then.
It is widely believed that the amount of words which someone knows directly equals to his / her vocabulary knowledge. However, vocabulary knowledge refers to much more skills than that. It rather means a learner’s knowledge about the possible relationship of the words in question, which one fits more with which. The word “combinality” merits to be given more attention simply because one of the first subjects in which EFL students have difficulty and tend to make mistakes is the collocational properties of words.
1.1. Background of the Study
Word combinations such as cause damage, sharp reaction, scream silently which are called “collocation” were originally introduced by Firth and he directed the attention of theorists and linguists to this phenomenon.
However, it is after the increasing number of contribution that some scholars (e.g. Sinclair, 1991; Nattinger, 1992; Lewis, 1993) make with their significant theoretical and pedagogical works that the importance of collocation has been fully realized. Today, it is almost an undeniable truth by the scholars and foreign language teachers that collocations are integral elements of second and foreign language teaching.
Many researches and studies have been conducted on collocation up to now. Scholars like Nattinger (1988), Sinclair (1991) and Lewis (2000) are among the ones who first categorized collocations and made invaluable contributions with their studies.
Also, some of them (Meara, 1984; Carter, 1987; Nesselhauf, 2003) have focused on the receptive (comprehension of word meaning) and productive (the relationship between words ,i.e. collocation) vocabulary skills of EFL / ESL learners not only revealing the learner errors in production but also offering some solutions to keep the collocational errors at minimum as an end product.
Dictionaries also are the most important sources that learners and instructors can apply to access lexical information. Almost every learner of a language as a foreign language has got at least one dictionary. But these typical dictionaries are not much helpful in the field of collocation and their application. They just help decoding, that is, make the unknown word known by definitions. However, after the recognition of collocations more dictionaries focusing on collocations have appeared and been served to the use of all learners interested in their more complex and advanced phraseological structure.
In short, collocations still need to be examined thoroughly in the aspect of learner needs since most researchers have handled this issue from other perspectives. The focus of this study is to fill just a part of this gap by having students get consciousness toward collocations and helping learners retain these combinations of words.
1.2. Statement of the Problem
The purpose of this study is whether using collocation has any positive effect on the retention of vocabulary by letting students have an interest and understand the importance of collocations itself, that is, a conscious-raising toward collocations. As words are the smallest meaningful part of a language, then collocations are located at the hub and should be handled
with great care and importance. Many linguists (Lewis, 2000; Nation, 2001) are aware of this fact and state that language knowledge is directly linked with collocations.
It has been accepted that vocabulary learning is an important skill in language learning. However, despite its importance vocabulary instruction has not received the attention it deserves in EFL / ESL instructional contexts (Zimmerman, 1997). In many institutions, for example, many teachers do not recognize the variety of vocabulary teaching techniques available. One of the teachers’ responsibilities is to provide learners with effective opportunities that will enable them to learn more vocabulary items and retain them for a longer time. Traditionally, vocabulary instruction has been mostly incorporated into reading lessons and has been mostly taught through dictionary definitions, synonyms and antonyms. Guessing meaning from the context is something which has been frequently used. However, there are many other vocabulary teaching techniques that teachers can use, such as word families, collocations and formal groupings. The efficiency of these techniques needs to be empirically investigated (Nattinger, 1988). This experimental study was set up to investigate whether teaching vocabulary in collocations will result in better vocabulary learning than teaching vocabulary using mere definitions.
However, students tend to make mistakes while using collocations which mostly stem from their L1. Almost all students adapt their vocabulary knowledge of L1 to L2 which a linguistic term called “mother tongue interference” and as a result some odd sentences may appear. Therefore, students should be given the chance to develop collocational competence.
1.3. Goal and Scope of the Study
The goal of this study is to find out whether teaching vocabulary through collocations will result in better vocabulary learning than teaching vocabulary using classical techniques such as definition, synonym, antonym and mother tongue translation. Our purpose is to show
the contribution of collocations to vocabulary learning of students in English as a Foreign Language (EFL) classes. Our suggested hypothesis is that learning vocabulary through collocations is an effective strategy that positively contributes to the development of vocabulary learning.
If it could be shown that teaching vocabulary through collocations improves the vocabulary learning more than classical techniques, teachers of English could be encouraged to spare some more classroom time for this type of training in their classes and to assign more importance to the application of certain learning strategies in vocabulary development.
The scope of this study is to discuss the vocabulary teaching techniques that the teachers use and to discuss a relatively new technique in order to make his process more effective and more meaningful.
1.4. Significance of the Study
Language learners are supposed to know thousands of words to be dominant over the language, that is, grasp it in all its aspects, to be able to use and understand it efficiently. That is why, seeking for plausible solutions for students’ problems about vocabulary building by the help of appropriate techniques is vital on the part of the teacher. That being said, most of the teachers who know this fact fail to let their students restore the amount of vocabulary they desire in spite of a great deal of time spent on this issue. The critical point and –almost always the omitted one- is that to know the definition of a word does not necessarily mean knowing this word. A much more successful result can be acquired by the words’ various use in context, their connotational aspects and so forth. Thus teachers should use more productive and effective ways to enable students to retain more words in less time at hand. Teaching vocabulary through collocations may be of great help to teachers to realize this.
This study aims to create and add new perspectives about collocations for EFL teachers and encourage them to develop or/and use vocabulary teaching techniques in order to build an awareness towards collocations in their students and to enhance retention of vocabulary.
1.5. Research Questions
This study intends to find answers to the following questions:
1) Does presenting new words through collocations result in a better learning of the words than presenting them without collocations?
2) Does presenting new words through collocations enhance retention of new vocabulary items?
CHAPTER II : LITERATURE REVIEW 2.1. Introduction
Vocabulary learning is central to language acquisition, whether the language is first, second or foreign. Although vocabulary has not always been recognized as a priority in language teaching, interest in its role in second language (L2) learning has grown rapidly in recent years and specialists now emphasize the need for a systematic and principled approach to vocabulary both by the teacher and the learner. The increased interest in this topic is evidenced by a rapidly expanding body of experimental studies and pedagogical material, most of which addresses several key questions of particular interest for language teachers.
As stated before, many researchers have conducted a great deal of study to know about the complex procedure of vocabulary acquisition and develop some techniques to enhance this process by speeding it up. One of these techniques is the presentation of vocabulary along with collocations and see the influence of it over the retention of vocabulary in compliance with the very aim of this study.
The first part consists of a brief retrospective of vocabulary and collocation as an integral element of it and their increasing importance over the years in ESL / EFL settings. The second part is based simply on the definition and categorization of collocation. The third section handles the importance of collocations in vocabulary teaching in EFL context. The last section is about the problems of collocation that both learners and teachers can come across frequently as a part of applied linguistics and also factors affecting EFL learners’ performance while dealing with collocations.
2.2. Vocabulary Teaching in ESL / EFL Settings
Vocabulary teaching is concerned with the selection and presentation of words for learners. Neglected for much of the twentieth century in favour of pronunciation and grammar. It has re-emerged since the 1980s as a central factor in language teaching.
Vocabulary played a central role in Grammar-Translation and early Direct Method approaches but Audiolingulism and the Structural Syllabus subordinated vocabulary to the requirements of pattern practice. Notional / Functional Syllabuses and Communicative Methodology gave no special emphasis to vocabulary. In the 1980s, research into lexis and Discourse Analysis combined with arguments from Psycho-Linguistics and L1 literacy research to reassert the importance of vocabulary in language learning. Computerized databases (such as COBUILD – the Birmingham University research project in lexical development with a corpus of over 20 million words of spoken and written English) gave researchers and materials writers access to powerful tools for vocabulary analysis. Proposals were made for a lexical syllabus (Willis, 1990) and approach (Lewis, 1993), assuming that “Language consists of grammaticalised lexis, not lexicalised grammar”. In practice, a lexical syllabus is hard to develop: grammatical structures are easier to select and sequence for teaching vocabulary. Furthermore, while it is obvious that vocabulary is of more use than grammar at the early stages of second language learning, supporters of grammar point out that successful processing and production of language will always rely heavily on grammatical knowledge.
It is estimated that an educated native English speaker has a vocabulary of about 50,000 words. Attempts have been made to determine a common core vocabulary for non-native learners. The 1930s Vocabulary Control Movement was concerned with delineating a minimum adequate vocabulary, primarily based on frequency counts. Ogden and Richard’s Basic English Project (1930) listed 850 basic words which would allow learners to express
complex ideas. West’s more influential General Service List (1953) consisted of the 2000 words that comprised 80% of the words in any written text.
Such word lists and frequency counts avoid the issue of multiple meanings: it has been calculated that the 850 words of Basic English have 12,425 meanings and that each of West’s 2000 words has, on average, 21 meanings. Which meaning(s) should be taught and in what order? High frequency of us may be less important than coverage (the contexts in which the word is used). Learnability is also a consideration: factors like spelling, syntactic or phonological difficulties can make a word difficult to learn. “Familiarity” is another important issue, bringing together the concepts of frequency, concreteness and meaningfulness. In addition, low-frequency words are precisely those which demarcate topic and therefore carry essential meaning. It is clear that trying to identify a common core vocabulary for all learners is almost impossible; while students of general English may benefit from learning such a core, students with specific needs will have different vocabulary requirements. Decisions will be affected by whether students need access to spoken and / or written language and by whether lexical items need to be in the active / productive vocabulary, which is always smaller than the receptive / passive one.
Learners cannot be taught all the vocabulary they will need and therefore must develop inferential strategies for dealing with unfamiliar vocabulary: e.g. by means of cloze and words-in-context exercises. Research reported by Carter (1987) suggests more proficient learners benefit most from these techniques. For all learners, the issue of how much unknown vocabulary impedes comprehension is an important one; Nation and Coady recommend West’s guideline of a maximum of %2 unknown words in a written text.
Exploitation of what linguists identify as the crucial area of lexical relations in teaching vocabulary is another issue, including consideration of cohesion and coherence. Meaningful and appropriate context is vital, helping learners to develop an awareness of
lexical patterns such as collocation and sense relations (e.g. synonymy). Componential analysis can help by grouping vocabulary into lexical fields according to common features or attributes. Word sets and grids have been developed for advanced learners, based on research that recall of words is often according to conceptual mapping of categories or semantic fields. Lower-level students may find lexical relations confusing; research, for example, argues against teaching pairs of opposites together, as only one item tends to be retained. However, creating associations within the language. (e.g. by organizing vocabulary according to topic and studying word formation) can promote learnability. Interlanguage associations are also useful, particularly in the early stages. Traditionally, students learned paired L1-L2 word lists; recall of these can be improved by association of target words with native words plus graphic or mnemonic representations. The more words are analyzed or enriched by association, the greater the possibility they will be remembered. Recycling vocabulary taught in similar and different contexts is, of course, also crucial to learning.
Analysis of words can be enhanced by efficient dictionary use. Developments in lexicography mean there is a range of dictionaries for non-native learners of English. Bilingual dictionaries are useful in the beginning stages and should than be used to check insufficiently understood explanations from monolingual dictionaries. The latter give considerably more information about entries and learners must be taught how to exploit these features.
Concordancing texts on computers also develops analytical skills. A concordance shows all the occurrences in context of a given word-form in a particular corpus and therefore allows learners to discover the range and frequency of uses of that word in the corpus.
2.2.1. Historical Overview
There is now general agreement among vocabulary specialists that lexical competence is at the very heart of communicative competence, the ability to communicate successfully and appropriately (Coady and Huckin, 1997). Given the current focus on vocabulary study, many non-specialists might be surprised to learn that, in past years, this area of teaching was often neglected because it was thought that vocabulary could simply be left to take care of itself. Although by the late 1970s and early 1980s more and more voices began to challenge this view (Judd 1978; Meara 1981; McCarty 1984; Laufer 1986), in 1988, Carter and McCarty were still taking note of the relative neglect of vocabulary in previous years. By then its reputation as the poor relation in language teaching was rapidly coming to an end.
The low status of vocabulary study and vocabulary teaching was in large part due to language teaching approaches based on American linguistic theories that had been dominant throughout the 1940s, 1950s and 1960s. Most influential in the early years was Charles Fries’s Teaching and Learning English as a Foreign Language (1945), based on American structural linguistics, which emphasized grammatical and phonological structure. Fries believed that grammar should be the starting point of language learning, and he also adopted the view that learning was a matter of habit formation. His audio-lingual method incorporated these ideas by paying systematic attention to intensive drills of basic sentence patterns and their pronunciation. Because the emphasis was on teaching grammatical and phonological structures, the vocabulary needed to be relatively simple, with new words introduced only as they were needed to make the drills possible (Larsen Freeman 2000; Zimmerman 1997). The assumption was that once students learned the structural frames, lexical items to fill the grammatical slots in the frames could be learned later, as needed. Although the shift to generative (transformational) linguistics in the 1960s brought about revolutionary changes in linguistic theory, triggered by Chomsky (1957), it did little to challenge the idea that the role of lexis was secondary to that of grammar. Chomsky rejected the behaviorist notion of habit
formation and supplanted it with a rationalist framework, the central assumption being that language is represented as a speaker’s mental grammar, a set of abstract rules for generating grammatical sentences. The rules generate the syntactic structure and lexical items from appropriate grammatical categories (noun, verb, adjective etc.) are selected to fill in the corresponding slots in the syntactic frame. The interests of generative linguists centered mainly on rule-governed behaviour and on the grammatical structure of sentences and did not include concerns for the appropriate use of language. Language learning approaches based on this theory viewed learning as rule acquisition, not habit formation, and emphasized grammatical rules. Vocabulary was afforded somewhat more importance, but the focus on rules of grammar still served to reinforce the idea that lexis was somewhat secondary (Carter and McCarty 1988).
Hymes (1972), while not rejecting Chomsky’s model, extended it and gave greater emphasis to the sociolinguistic and pragmatic factors governing effective use of language. Hymes was especially concerned with the concept of communicative competence, which emphasized using language for meaningful communication, including the appropriate use of language in particular social contexts (for example, informal conversation at the diner table versus formal conversation at the bank etc.).The teaching approach that evolved from these notions, referred to as communicative language teaching, promoted fluency over accuracy and consequently shifted the focus from sentence-level forms to discourse level functions. (e.g requests, greetings, apologies and so on). Once again, though, vocabulary was given secondary status, taught mainly as support for functional language use. As in previous approaches, it was generally assumed that vocabulary would take care of itself (Schmitt 2000).
This picture has changed dramatically within the last two decades. The challenge to the status quo began in the late 1970s and early 1980s and by the late 1980s and early 1990s,
vocabulary studies were developing exponentially and vocabulary teaching was coming into its own. One reason for the resurgence of interest on the part of researchers was that computer-aided research was providing vast amounts of information that had not previously been available for analysis, such as information about how words behave in actual language use, larger units that function in discourse as single lexical items, and differences between written and spoken communication. Further, psycholinguistic studies were providing insights concerning mental processes involved in vocabulary learning, such as memory, storage and retrieval. Interest in these issues led in turn to related studies concerned with developing more effective vocabulary teaching and learning strategies.
Today, it is accepted that learning word meanings can not be achieved only through the use of a dictionary, and that vocabulary acquisition is a complex process. This understanding has led to a considerable emphasis on vocabulary. The principal reasons for the present focus on vocabulary, according to Allen (1983:5) are these:
First, many ESL and EFL classes have revealed disappointing results although a great deal of time has been devoted to vocabulary teaching by teachers; second, recent research into word meanings which has dealt with lexical problems, indicates that these lexical problems frequently interfere with communication and that not using the right words results in a communication breakdown.
Nation (1990:1-2) supports the idea that vocabulary should be taught in a systematic and principled approach due to the following reasons:
1. Because of the considerable research on vocabulary we have good information about what to do about vocabulary and about what vocabulary to focus on.
2. There is a wide variety of ways for dealing with vocabulary in foreign or second language learning.
3. Both learners and researchers see vocabulary as a very important if not the most important element in language learning. Learners feel that many of their difficulties in both receptive and productive language use, result from an inadequate vocabulary.
He also argues that the language tasks in which students with inadequate vocabulary will be involved will cause them to suffer from frustration and concludes that vocabulary has vital importance in reading and therefore giving attention to vocabulary is unavoidable.
Nattinger (1988) states that comprehension requires understanding the words and storing them and also committing them to memory whereas production requires retrieving them from memory and using them in appropriate situations. Hence, our aim in teaching vocabulary should be to strengthen this memory storage.
Learning vocabulary is something more than memorizing lists of words. To know a word in a target language as well as the native speaker knows it may mean the ability to:
a) recognize it in its spoken or written form; b) recall it at will;
c) relate it to an appropriate object or concept; d) use it in the appropriate grammatical form; e) pronounce it in a recognizable way in speech; f) spell it correctly in writing;
g) use it with the words it correctly goes with, in correct collocations; h) use it at the appropriate level of formality;
i) be aware of its connotations and associations
(Wallace, 1982:27)
Wilkins lamented the neglect of vocabulary in the audio-lingual years. While it is true that to learn nothing but words and little or no structure would be useless to the learner, it would also be useless to learn all the structure and no vocabulary: “Without grammar very
little can be conveyed, without vocabulary nothing can be conveyed” (Carter and McCarty, 1988:42)
According to Carter and McCarty, Wilkins’s work is significant for his desire to bring to vocabulary teaching which have become a major feeding ground for vocabulary practitioners in the 1970s and 1980s. Also, they state that Twaddell’s arguments around the same time that it is impossible to teach learners all the words they need to know, and so it is important to teach them guessing strategies that will enable them to tackle unknown words and lose their reliance on dictionaries is the beginning of viewing vocabulary learning as a language skill, of shifting the responsibility to the learner. By the mid 70s, we have a picture of a growing concern with vocabulary teaching and learning. What is more, we have the beginnings of the view of the vocabulary as a skill in which the learner is actively involved and a concern with what and how the learner might learn.
The move away from seeing vocabulary as lists of items to be learnt raises the question of precisely what it means to learn vocabulary. Richards tries to tackle this issue, considering some of the knowledge that is assumed by lexical competence. (Carter & McCarty, 1988)
Carter and McCarty (1988) state that by the end of the 70s, vocabulary teaching gained importance. Its place within language teaching had been reasserted, insights from lexical semantics had been brought to bear in the incorporation of notions such as sense-relations and collocation into teaching materials, the learner had been brought to the center stage and the lexicon was beginning to be seen as a resource for the needs of the learner and for strategic use in the gaining of communicative objectives.
Nation (2005) points out the main problem with vocabulary teaching is that only a few words and a small part of what is required to know of a word can be dealt with at any one time. He also adds as follows:
The first decision to make when teaching a word is to decide whether the word is worth spending time on or not. When deciding how to spend time on a word, it is useful to consider the learning burden of the word. Part of effective vocabulary teaching involves working out what needs to be taught about a word. This is called the learning burden of a word.
(Nation, 2005:2-3)
McCarty (1984) points out that attention has recently been turned to the problems of vocabulary in foreign-language teaching, and a steadily growing amount of work is beginning to challenge assumptions that have relegated vocabulary teaching to a secondary position in the priorities of language teaching. At the beginner or lower intermediate level, the treatment of vocabulary as a teaching area in itself is sadly lacking. Beginner courses do present many new orthographic words in a carefully controlled way, but do nor generally deal with words lexically and they view vocabulary acquisition as a cumulative by-product of the teaching of structures or the communicative functions of sentences.
Deveci (2004) states that the importance of vocabulary acquisition has always been recognized, although at times, vocabulary was treated as separate from grammar and other skills. However, the communicative and natural approach emphasized the importance of vocabulary development, which resulted in more interest in vocabulary teaching. He also points out that we can not use structures correctly if we do not have enough vocabulary knowledge.
Zughoul and Abdul-Fattah (2003) state that development of word lists for English based on frequency distribution directed the attention of teachers, ELT practitioners and curriculum specialists. Mastering the upper ends would be adequate for the development of a good measure of proficiency in English. Besides, the upper ends would be efficiently exploited for learning English phonology and syntax. This restricted view of vocabulary,
beside its notional falseness from a statistical point of view, has caused learner lexical deficiency and incapability to function adequately in real life situations. Work in the area of corpus linguistics has convincingly triggered the urge for a reconsideration of the pedagogues have been on the wrong track in their assumptions about the role of vocabulary frequency distribution. The counter argument has led to the recognition of a fundamental role for lexical learning. The area of collocation within the realm of lexis is of prime importance and forms a serious problem for language learners.
2.3. The Definition and Categorization of Collocation 2.3.1. Definitions of Collocation
The term ‘collocation’ is used in widely different and often rather vague senses in linguistics and language teaching. Among the many diverse uses of the term, two main views can be identified (Klotz 2000: 63.; Nesselhauf 2004). In one of these two views, a collocation is considered the co-occurrence of words at a certain distance, and a distinction is usually made between co-occurrences that are frequent (or more precisely, more frequent than could be expected if words combined randomly in a language) and those that are not. This view has therefore been called the ‘statistically oriented approach’ (Herbst 1996: 380) or the ‘frequency-based approach’ (Nesselhauf 2004a). In the other view, collocations are seen as a type of word combination, most commonly as one that is fixed to some degree but not completely. This view has been referred to as the ‘significance oriented approach’ (Herbst 1996: 380) or the ‘phraseological approach’ (Nesselhauf 2004a). The frequency-based approach goes back to J. R. Firth and has been developed further in particular by Halliday and J. Sinclair.
The term ‘‘collocation’’, as is well known, was first coined in its modern linguistic sense by the British linguist J. R. Firth, along with the famous explanatory slogan: ‘‘you shall judge a word by the company it keeps’’. In an article entitled “Modes of Meaning” (Firth
1957), he outlines how the study of ‘‘meaning by collocation’’ can contribute to a formal and contextual approach to word meaning: Meaning by collocation is an abstraction at the syntagmatic level and is not directly concerned with the conceptual approach to the meaning of words. One of the meanings of night is its collocability with dark, and, of dark, of course, collocation with night. Later writers on collocation have picked up different aspects of Firth’s ideas. A number of different but related definitions of ‘‘collocation’’ have been forthcoming. Sinclair, who was a student of Firth’s at London University, sees it as follows: Collocation is the occurrence of two or more words within a short space of each other in a text ( Sinclair 1991). We might call this the ‘‘textual’’ definition. One item collocates with another if it appears somewhere near it in a given text. Another definition is given by Leech in his discussion of ‘‘Seven Types of Meaning’’, one of which is ‘‘collocative meaning’’: Collocative meaning consists of the associations a word acquires on account of the meanings of words which tend to occur in its environment ( Leech 1974). This could be referred to as the ‘‘psychological’’ or ‘‘associative’’ definition. It is part of a native speaker’s communicative competence to know what are normal and what are unusual collocations in given circumstances. Through lifelong exposure to a language, native speakers acquire what Firth calls ‘‘expectancies’’ (1957) of which items commonly co-occur with which others in texts. The contribution of collocation, in psychological terms, to meaning is also emphasized by Aitchison, who says that ‘‘humans learn word-meaning from what occurs alongside’’ (1994: 21). The learner, child or adult, faced with an unknown word looks at the co-text to gain clues as to what the unfamiliar item might mean. Meaning is function in context, as Firth used to say. Finally, Hoey highlights another aspect of the concept: collocation has long been the name given to the relationship a lexical item has with items that appear with greater than random probability in its (textual) context. (Hoey 1991) We might call this the ‘‘statistical’’ definition of collocation. It is a good working Definition of the concept for those studying
corpus linguistics, where large quantities of text can be made available for computer analysis. The co-occurrence of two items becomes interesting if it seems to happen for a purpose, and especially if it is repeated, if there are ‘‘patterns of collocation’’. The habitual associations of a word with other items can thus be studied both by calling up concordances of that word and by obtaining lists of its most frequent collocates. Firth himself gives the example of the word “time” whose common collocates include “saved”, “spent”, “wasted” and also “presses” and “flies” and even the word “no”. By these means, it is possible to build up a picture of the way in which each and every lexical item in a language behaves. The majority of modern learners’ dictionaries make use of computer facilities for lexical analysis and include information on collocation. ‘‘Collocation’’ is usually used to refer to the co-occurrence of two single words. However, Firth himself originally used the term to refer to the co-occurrence of items at all grammatical levels, not just the word level. If we concentrate on higher levels of language, there is theoretically no qualitative difference between word with word, word with phrase, phrase with phrase, even phrase with clause and clause with clause collocation.
The phraseological approach has been strongly influenced by Russian phraseology. Typically, researchers adopting this approach work in the fields of lexicography and/ or pedagogy; among the main representatives are A. P. Cowie and F. J. Hausmann. For the frequency-based approach, Sinclair’s view of collocations will be discussed, for the phraseological approach, that of Cowie. Sinclair defines collocations as “the occurrence of two or more words within a short space of each other in a text” (1991: 170). A short space, or ‘span’, is usually defined as a distance of around four words to the right and left of the word under investigation, which is called the ‘node’ (e. g. 1991: 170; Jones & Sinclair 1974: 21f.). If, for example, in a given amount of text, the word house is analysed, and the word occurs in an environment such as He went back to the house. When he opened the door, the dog barked, the words went, back, to, the, when, he, opened, they are all considered to form collocations
with the node house; these words are then called ‘collocates’. Sinclair distinguishes two types of collocations, namely ‘significant’ and ‘casual’ collocations, and sometimes reserves the term ‘collocation’ for the former type (e. g. 1991: 115). Significant collocations are co-occurrences of words “such that they co-occur more often than their respective frequencies and the length of text in which they appear would predict” (1974: 21). In the example above, the and house would probably not be significant collocations, as, although these two words can be assumed to co-occur frequently, “the” is itself a frequent word in virtually every kind of text. The words “dog” and “barked” would, however, very likely constitute a significant collocation, as “barked” is not usually very frequent and, if it occurs, is likely to be found near the word “dog”. Exact formulae of how to determine exactly whether co-occurring words constitute a significant collocation have also been developed by Sinclair and others. Given that even Sinclair sometimes varies in how he defines collocations, it is not surprising that some researchers adopting a frequency-based approach to collocations consider co-occurrences of all frequencies to be collocations (e. g. Halliday 1966; Moon 1998), while others reserve the term for frequent co-occurrences (e. g. Stubbs 1995). Some use recurrence, i. e. co-occurrence more than once in a given corpus, as the defining criterion. Other points of variation in the definition of collocations in the frequency-based approach are also mirrored by variation in Sinclair’s writings. Whereas he uses ‘word’ in the sense of ‘lexeme’ in the above definition, and thus sees collocation as a relationship between lexemes, he previously regarded it as a relationship between ‘lexical items’. This latter view is also shared by Halliday, who exemplifies ‘lexical item’ with the group of derivationally related lexemes STRENGTH, STRENGTHEN, STRONG (1966: 156). According to this view, a strong argument, he argued strongly, the strength of the argument, his argument was strengthened would all be considered instances of the same collocation. A third view on this question is that collocation is a relationship between word forms, i.e. that combinations such as hold tight and
holds tight are two different collocations. A more fundamental aspect in which definitions vary is the question of the nature of the collocation as such.
Unlike the frequency-based approach, the phraseological approach consistently requires that the elements of collocations should be syntactically related. Hausmann even goes so far as to call only those combinations collocations that appear in a pre-defined set of syntactic relations: adjective + noun, (subject-) noun + verb, noun + noun, adverb + adjective, verb + adverb, verb + (object-)noun. He thus also only allows the combination of two lexical elements in the category of collocations, while Benson et al., for example, also permit a lexical word plus a preposition. As in the frequency-based approach, the number of participating items varies between two and two or more, but often this question is not addressed at all. Shrug one’s shoulders, for example, can therefore be viewed as a collocation consisting of either three elements (shrug + ones + shoulders) or of two (shrug + shoulders). As to the question of whether lexemes, word forms, or lexical items/ roots are the elements of collocations, most authors assume that the participating elements are lexemes (with the exception of Cowie). A final important point on which representatives of the phraseological approach differ is on how they view the relationship between the elements of a collocation. Often, the assumption seems to be that there is no difference in the nature of the elements, as for example in Cowie’s definition, which merely requires that one of the elements is restricted but does not specify which one (e. g. Cowie 1992: 51.). What can also be found is a double use of the term ‘collocation’, i. e. its use in both the sense of the frequency-based approach and the phraseological approach in one and the same piece of work. F. R. Palmer, for example, on the one hand reserves the term for free and restricted combinations as opposed to idioms (1981: 77.), and on the other refers to “the collocation of kick and the bucket” (1981: 79), where ‘collocation’ apparently means co-occurrence.
Sometimes ‘collocation’ seems to be used purely to describe a phenomenon in a given amount of text; more commonly, it also seems to be considered a more abstract tendency in a language. Further points that are viewed differently by authors adopting a frequency based approach are the number of words involved in a collocation and whether or not these have to be consecutive. A final aspect in which definitions vary is the syntactic relationship of the elements involved in a collocation. In the frequency-based approach, the syntactic relationship between the elements does not usually play a role in deciding whether they form a collocation or not.. A. P. Cowie is a typical representative of the phraseological approach: he considers collocations a type of word combination, i. e. an abstract combination with instantiations in actual texts, and defines them by delimiting them from other types of word combinations, most importantly from idioms on the one side and from what he sometimes calls ‘free combinations’ on the other (e. g. Cowie 1981, 1994; Cowie et al. 1993). At the same time he is one of the most important representatives of the phraseological approach, as his attempts to define collocations and to delimit different kinds of word combinations are among the most precise. Cowie divides word combinations into two main types, ‘composites’ and ‘formulae’. Formulae are combinations with a primarily pragmatic function such as “How are you?” or “Good morning”. Collocations belong to the group of ‘composites’, which are described as having a primarily syntactic function. The distinctions in the group of composites are made on the basis of two criteria, which Cowie assumes to interact closely: the criterion of transparency and the criterion of commutability (or substitutability). Transparency refers to whether the elements of the combination and the combination itself have a literal or a non-literal meaning, and commutability refers to whether and to what degree the substitution of the elements of the combination is restricted. On this basis, he distinguishes the following four types of combinations, stressing, however, that these types are not clearly delimitable, but should rather be seen as forming a continuum:
Free combinations (e. g. drink tea):
– the restriction on substitution can be specified on semantic grounds – all elements of the word combination are used in a literal sense Restricted collocations (e. g. perform a task):
– some substitution is possible, but there are arbitrary limitations on substitution – at least one element has a non-literal meaning, and at least one element
is used in its literal sense; the whole combination is transparent Figurative idioms (e. g. do a U-turn, in the sense of ‘completely change one’s policy or behaviour’):
– substitution of the elements is seldom possible
– the combination has a figurative meaning, but preserves a current literal interpretation Pure idioms (e. g. blow the gaff ):
– substitution of the elements is impossible
– the combination has a figurative meaning and does not preserve a current literal interpretation
The most important variation in Cowie’s use of the term ‘collocation’ is that while he sometimes applies it only to combinations with an arbitrarily limited substitutability in which one element is used in a non-literal sense, he sometimes applies it to free combinations as well. In this case, however, he makes a distinction between ‘open collocations’ (i. e. free combinations) and ‘restricted collocations’. As in the case of Sinclair, Cowie’s variation in the use of the term reflects some of its different uses by different authors adopting a phraseological approach. A number of researchers apply the term ‘collocations’ to both free combinations and restricted collocations. Some of these do not differentiate further, while others, like Cowie, distinguish between ‘open collocations’ (or ‘free collocations’) and ‘restricted collocations’. The number of categories towards the more restricted and opaque end of the scale also varies between authors. Cowie’s distinction between two types of idioms
(figurative idioms and pure idioms) is often not made, and Benson et al., for example, consistently postulate an additional category between collocations and idioms, which they call ‘transitional combinations’ or ‘transitional collocations’ (Benson et al. 1986: 254) An important aspect of variation which is not present in Cowie’s own definition(s) concerns the criteria that are used for the delimitation of collocations from other types of word combinations. The distinction between combinations with a pragmatic function and those without such a function is sometimes explicitly made, but often the question is not addressed. For the distinction of different types of non-pragmatic word combinations, many authors apply the same two criteria as Cowie, i. e. the criteria of opacity and commutability. Some of them, however, use the criteria differently for the distinction between (restricted) collocations and free combinations than for the distinction between (restricted) collocations and idioms. Hausmann, for example, considers arbitrarily restricted combinability the main factor for distinguishing collocations from free combinations, and transparency the main factor for distinguishing collocations from idioms (1989: 1010). Even when the same criteria are used by different authors, however, the delimitations between different types of word combinations are not necessarily identical. Among those authors basing the distinction between free combinations and collocations on commutability, for example, (restricted) collocations are sometimes a broader category than in Cowie’s classification, and sometimes a narrower one. Interestingly, a number of authors use the criterion of frequency of co-occurrence, i. e. the main criterion of the frequency-based approach, in addition to phraseological criteria such as commutability and transparency (e. g. Nation 2001: 317; Herbst 1996: 389; Benson et al. 1986: 253). For these authors, for a combination to be considered a collocation, it has to be restricted, transparent and frequent; Benson et al. even seem to assume that the criteria of restriction and frequency coincide.
2.3.2. Types of Collocations
We have already noted that some collocational phrases are fixed and some allow degrees of variation. A fully fixed phrase is one which allows no syntactic transformation and no internal lexical variation. Sinclair (1991: 110— 11) cites the example of the phrase of course whose constituent words cannot be shifted around, or added to, or altered in any way. It effectively functions as a single unit and it is only the vagaries of English orthography which prevent it from being spelt as one word along the lines of insofar as, an invariable phrase in which three out of four of its constituents are in fact written as one. Compare also in spite of which, with three orthographic words, does the same work as the single word despite. Nattinger & DeCarrico call these items ‘‘polywords’’ and note how some of them maintain a regular grammatical form (are ‘‘canonical’’, in their terminology. Other types of fully fixed phrases include proverbs and sayings, quotations and those idioms which have a frozen form (e. g. on the wagon). They also include the category of fully restricted collocation, that is, phrases in which the appearance of one item implies the co-occurrence of another. For example, if we meet the word “stinking” before an adjective, this can normally only be “rich”, or if we meet the word “blithering” it is almost certain that “idiot” will follow. Variations in phraseology come about in a number of ways. As Sinclair notes, ‘‘many phrases have an indeterminate extent’’, in other words, it is not always possible to say precisely what is properly part of the phrase and what is an optional addition: As an example, consider “set eyes on”. This seems to attract a pronoun subject, and either never or a temporal conjunction like the moment, the first time, and the word has as an auxiliary to set. How much of this is integral to the phrase? (Sinclair 1991: 111) Many phrases allow what Sinclair calls ‘‘internal lexical variation’’. There seems to be little to choose, he says, between in some cases and in some instances. Other phrases allow some variation in word order. We might add that some phrases are susceptible to lexical insertion. Even highly restricted collocational phrases such
as vested interest(s) can have words inserted in them to produce, for example, vested financial interest or vested political interests. Carter (1987) proposes a number of clines in the criteria which are relevant in determining how fixed or how free particular lexical patterns are. One of these is the cline of syntactic structure or word order, similar to that described above. As an example of fixed order he cites phrases such as go it alone and the more the merrier. He points out that the more irregular (or non-canonical) a phrase’s syntax, as in the above examples, the more likely it is to be a fixed phrase. As an example of a phrase with a more flexible pattern he cites break someone’s heart which can take on a number of forms including he broke her heart, they’re heart-broken, she’s a right little heart-breaker. Another of Carter’s clines is that of collocational restriction. The most restricted are phrases such as “stinking rich” and “blithering idiot” where the occurrence of one item is conditioned by the other. Total conditioning however is probably quite rare, except for the special category of ‘‘irreversible (or ‘‘ordered’’) binomials’’, which include cash and carry, bread and butter, ups and downs and so on. At the opposite end of the scale, we have what Carter calls ‘‘unrestricted collocation’’, which ‘‘describes the capacity of particular lexical items to be open to partnership with a wide range of items’’. Most common lexical items (e. g. fat, bright, head) are in this category. In between the two extremes are the items which have a ‘‘semi-restricted collocational range’’, that is, they normally collocate with a limited number of other lexical items. Carter’s examples include the item harbour (as a verb) which collocates with doubts, uncertainty, a grudge, suspicion. There is less difference between the categories of ‘‘unrestricted’’ and ‘‘semirestricted’’ collocational potential than might at first sight appear. Whilst it is true that a word like head— one of Carter’s examples of an item with unrestricted collocationalability— collocates very widely, corpus evidence nevertheless demonstrates its collocational preferences, in the sense that it is found more often with particular items than others. It collocates frequently with bang, scratch, turn, nod, and so on, in non-idiomatic uses,
with “lose one’s”, “come to a”, “go to one’s”, “fill one’s”, “go over one’s” etc in idiomatic expressions. The word head is not unusual in this. In fact, every lexical item enters into particular collocational relations with the rest of the lexis of a language, a behaviour which can be studied and described in terms of frequencies and preferences.
2.3.3 Classifications of Collocations
There have not been many attempts to classify restricted collocations, but the classifications that have been made can be divided into three types. The first type is based on the syntactic characteristics of the collocation, the second on its the semantic characteristics of the collocation and the third on the commutability of its elements. In the first type, restricted collocations are classified according to the word classes in which their elements appear. As already mentioned above, Hausmann (1989), for example, divides collocations into six types: adjective + noun (heavy smoker), (subject-) noun + verb (storm – rage), noun + noun (piece of advice), adverb + adjective (deeply disappointed), verb + adverb (severely criticize), and verb + (object-)noun (stand a chance). Aisenstadt (1981) proposes a similar classification, but divides the verb + noun group further into verb + noun (e. g. make a decision) and verb + preposition + noun (e. g. come to a decision). Benson et al. make the same distinctions as Hausmann, but owing to their broader definition of collocations add the combinations noun + preposition (e. g. interest in), preposition + noun (by accident), and adjective + preposition (angry at). They also make a more fundamental distinction, which is based on the word classes to which the elements of a collocation belong. Collocations in which two lexical elements co-occur are called ‘lexical collocations’, collocations in which a lexical and a more grammatical element (such as a preposition) co-occur, are called ‘grammatical collocations’. The second type of classification is based on the semantic characteristics of the combination. Two different kinds of attempt to classify collocations in this way can be discerned. One is
limited to verb-noun collocations and is based on the nature of the meaning of the verb. Cowie is one of the few researchers who attempt such a classification. He distinguishes between verbs with a ‘figurative’, a ‘delexical’ and a ‘technical’ meaning. Corresponding collocations would deliver a speech, make recommendations, try a case. Cowie’s classification, which was also adopted by Howarth, was probably inspired by Aisenstadt (1979), who makes a similar distinction between verbs with a “secondary, abstract meaning”, verbs with a “grammaticalized, wide and vague meaning”, and verbs with a “very narrow and specific meaning” (1981: 57). From the examples she gives, it may be assumed that the first two categories roughly correspond to Cowie’s figurative and delexical meanings; the last category is exemplified by shrug one’s shoulders. The second type of attempt at classifying collocations on a semantic basis is much more detailed and also more comprehensive in that it applies to all grammatical types of collocations. The classification is based on the notion of what is called ‘lexical functions’. An example of a lexical function is the meaning ‘do’/‘perform’. If this meaning is to be expressed with respect to the noun cry, one of the possible lexemes is let out; with respect to support, lend is possible. Possible collocations thus are let out a cry and lend support (but not *lend a cry or *let out support). In theory, it should be possible to classify most collocations according to lexical functions, i. e. according to the meaning the collocator expresses. The third type of classification that has been established for restricted collocations is based on the commutability of the elements of a collocation. The distinction made by Benson et al. between ‘collocations’ and ‘transitional collocations’ may be considered an example of this type of classification, as it is based on the variability of elements( foot a bill, for example, is assigned to this category, as foot in this meaning can only combine with bill). An attempt to explicitly subclassify collocations on the basis of commutability is made by Aisenstadt (1979, 1981). She divides collocations into two groups depending on whether both or only one of the participating lexical elements are restricted in
their commutability. In the combination “shrug one’s shoulders”, for example, she assumes that both lexical elements are restricted, and illustrates this with the following paradigm: “shrug one’s shoulders”, “shrug sth. ff”, “shrug sth. away”, “shrug one’s shoulders”, “square one’s shoulders”, “hunch one’s shoulder”. In combinations such as to make/ take a decision or “auburn hair”, on the other hand, only one element is considered restricted in its commutability. In the former combinations, the verbs are said to “have a rather wide and vague meaning and collocate with different nouns” (1981), whereas the noun “is restricted in its commutability, though not to one verb only”; in the latter combination, the commutability of “auburn” is considered to be restricted to hair whereas hair commutes freely with a great number of other adjectives” (1981). The most comprehensive classification on the basis of commutability to date has been established by Howarth; it is, however, restricted to verb-noun collocations. Howarth distinguishes five ‘levels of restrictedness’ according to two criteria, namely the number of elements that are restricted in their commutability and the degree of the restriction (1996). These levels are described and exemplified as follows (1996: 102):
1. freedom of substitution in the noun; some restriction on the choice of verb an open set of nouns, a small number of synonymous verbs:
adopt/ accept/ agree to a proposal/ suggestion/ recommendation/ convention/ plan
2. some substitution in both elements a small range of nouns can be used with the verb in that sense there are a small number of synonymous verbs:
introduce/ table/ bring forward a bill/ an amendment
3. some substitution in the verb; complete restriction on the choice of the noun no other noun can be used with the verb in that sense; there are a small number of synonymous verbs:
pay/ take heed
4. complete restriction on the choice of the verb; some substitution of the noun a small range of nouns can be used with the verb in that sense
there are no synonymous verbs: give the appearance/ impression
5. complete restriction on the choice of both elements no other noun can be used with the verb in the given sense there are no synonymous verbs:
curry favour (with)
Howarth also relates these five levels to the semantic categorization of combinations into those with figurative, delexical, and technical verbs, which he adopts from Cowie. He finds that there are no combinations with figurative and technical verbs on level five and no combinations with technical verbs on level one, but that all other combinations of type of verb and level of restrictedness occur (1996: 118), which means that there is no direct correlation between the type of verb in the combination and its level of restrictedness. This attempt to relate aspects of meaning and commutability in some detail in a classification of collocations is probably unique to date.
2.4 The Importance of Collocation in EFL Context
Collocations, i. e. arbitrarily restricted lexeme combinations such as make a decision or fully aware, are one type of a group of expressions whose importance in language has been increasingly recognized in recent years. This group of expressions has been variously called prefabricated units, prefabs, phraseological units, (lexical) chunks, multi-word units, or formulaic sequences. They are made up of more than one word and are lexically and/ or syntactically fixed to a certain degree. Following a period in which, largely due to the wide influence of generative grammar, prefabricated units were considered peripheral in language, it is today widely assumed that their number is vast and that they play a major role in
language processing and use. Bolinger was among the first linguists to point out that a generativist view, which relegates prefabricated units to the periphery of language, fails to account for a considerable part of observable language data. On the basis of numerous examples he claims that our language does not expect us to build everything starting with lumber, nails, and blueprint. Instead it provides us with an incredibly large number of prefabs (1979: 96). He also points out that most of these prefabs are not completely but only partially fixed. Many linguists have since made similar claims, most notably Pawley and Syder. They – also mainly on the basis of a sizeable collection of prefabricated units – come to the conclusion that “by far the largest part of the English speaker’s lexicon consists of complex lexical items” (1983: 215). Further empirical support for this view has come from corpus studies, which have regularly found that most of naturally occurring language, both spoken and written, consists of recurrent patterns, many of which are phraseological (e. g. Altenberg 1998; Sinclair 1991). Corpus studies have also shown that collocations are a frequently occurring type of semi-prefabricated unit. In an analysis of over 5,000 verb-noun combinations in a written 240,000 word corpus, for example, over a third of the combinations were found to be collocations (Howarth 1996: 120). Several important functions have been identified for prefabricated units. First, there is growing evidence that they play an essential role in language learning, as they seem to be the basis for the development of creative language in first language and childhood second language acquisition. Secondly, prefabricated units are essential for fluency in both spoken and written language. Psycholinguistic evidence indicates that the human brain is much better equipped for memorizing than for processing, and that the availability of large numbers of prefabricated units reduces the processing effort and thus makes fluent language possible (Aitchison 1987). Thirdly, the use of prefabricated units supports comprehension, as the recipient can understand the meaning of a passage of text without having to attend to every word (Hunston
& Francis 2000). And fourthly, prefabricated units serve to indicate membership of a certain linguistic group; they fulfill “the desire to sound like others” (Wray 2002). For the adult non-native speaker, the first of these functions probably does not play a major role, as it seems that prefabricated language is not regularly used as a basis for creative language in adult L2 acquisition. However, two of the other functions are at least as essential for non-native speakers as for native speakers. Enhancing fluency through reducing processing effort must be of particular interest for non-native speakers, as they naturally need more processing effort to convey their intended message. Indeed, it has been shown that whether or not L2 production is fluent crucially depends on the learner’s control over a large repertoire of prefabricated units (Towell & Hawkins 1996). The third function, making comprehension easier, is doubtless of importance for every user of a language. While the use of native-like prefabs aids comprehension, non-native-like prefabs can irritate the recipient and draw the attention away from the message. Being perceived as a member of a certain linguistic group that speaks the L2 natively, finally, though clearly not an aim of all non-native speakers, is also important to certain learners of a language. The knowledge of and the ability to use prefabricated units are thus essential for the language learner; unfortunately, however, they also pose considerable difficulties, even for the advanced learner. Language learners often stumble across co-occurrence relations. Any analysis of students’ speech or writing shows a lack of collocational competence. Knowing which subset of grammatically possible utterances is actually commonly used by native speakers is an immense problem for even the most (Wray 1999) proficient of non-natives. There is also wide agreement that prefabs have to be taught (Bahns 1997; Cowie 1988; McCarthy 1990 and many others). In spite of this, many types of prefabs, including collocations, are still not treated adequately in English language teaching today. Although collocations have received increasing attention in language teaching in recent years (Granger 1998; Howarth 1998), we are still far from the development of a
coherent methodology and even further from a wide-spread and systematic treatment of collocations in language teaching materials and syllabi. In recent years, however, a few approaches to language teaching have been developed, which, far from neglecting phraseological units, put them at the centre of teaching: Willis’s lexical syllabus, Lewis’s lexical approach, and the lexical phrases approach by Nattinger and DeCarrico. As with most other suggestions for teaching prefabs, even these approaches are at best based on the analysis of native speaker prefabs; none of them is based on any systematic observation of prefabs in learner language However, if efficient pedagogical measures are to be devised, they need to take into consideration the difficulties learners have with prefabricated units.
More recently, Nattinger & DeCarrico too have noted that: In formulating performance models of language processing, researchers endeavour to offer direct descriptions of psychological categories and processes, attempting to describe languages in terms of how they are perceived, stored, remembered and produced. These researchers feel that the storage capacity of memory is vast, but that the speed for processing those memories is not (Crick 1979: 219), so that we must learn short cuts for making efficient use of this processing time.—( Nattinger & DeCarrico 1992: 31) Principal among these short cuts is the use of prefabs. A still more powerful reason for the employment of semi-preconstructed phrases probably lies in the way it facilitates communication processing on the part of the hearer. Language consisting of a relatively high number of fixed phrases is generally more predictable than that which is not (Oller & Streiff 1975). In real-time language decoding, hearers need all the help they can get. Redundancy in communication is often explained in this way and the collocational principle probably has the same functional origins.
‘‘Every linguist makes room in his scheme of things for lexical units larger than words,’’ Bolinger writes. ‘‘He calls them idioms’’. What is new is that: we are now in a position to recognize that idiomaticity is a vastly more pervasive phenomenon than we ever imagined,
