Written by: Rossen Stoitchkov, “St. Kliment Ohridski” University of Sofia
Department for Language Teaching and International Students

The paper deals with collocations of the type verb + noun, which are central to building the language learner’s mental lexicon.  It draws on evidence of language learners’ lack of awareness of the psychological salience and severely limited repertoire of collocations in their output. It further looks at noticing as a central pedagogic activity and as the first step to improving language learners’ knowledge of collocations.

It has been claimed that, in contrast to native speakers, learners hardly make use of prefabricated units in language production:

“In building his utterances, he [the native speaker] makes use of large prefabricated sections. The learner, on the other hand, having automated few collocations, continually has to create structures that he can only hope will be acceptable to native speakers. His building material is individual bricks rather than prefabricated sections.”
(Kjellmer, 1991:124)

One important observation that was made by a number of researchers is that learners use an overall smaller number of restricted collocations than native speakers would have used. Similar findings were reached by Granger (1998) and Kaszubski (2000). Learners vary much more in their use of collocations than native speakers and also frequently appear to create collocations using ‘individual bricks’. This observation indicate that a model of the second language lexicon that simply assumes two types of elements, words and chunks, is inadequate. A model based on the notion of stronger and weaker links between elements appears much more adequate, as it can explain, for example, why learners in many cases combine the appropriate lexical elements, but do not choose the appropriate non-lexical elements or features (e.g. ‘make impression’ [make an impression]; ‘give yourself a try’ [give it a try]; ‘lose one’s tie with sb.’ [lose one’s ties with sb.]; ‘lift one’s spirit [lift one’s spirits]). The conclusion to be drawn is that the links between the lexical elements are fairly strong, but between the lexical and non-lexical elements are fairly weak. What can probably be claimed is that often the links between elements of (semi-) prefabricated units are weaker in the learner’s mental lexicon than in the lexicon of native speakers – both the links between the elements of collocations and those between collocations and larger units of usage. Similar observations have also been made for example by Meara (1984), who concludes from association tests that the learner’s mental lexicon is in general more loosely organized than the native speaker’s lexicon (1984:232). Weaker links would also explain why learners apparently blend collocations more often than native speakers. Although the elements of collocations are linked to a certain extent in the learner’s mental lexicon, the links between these elements can probably be broken up more easily and replaced or supplemented by other existing links.
Another inference about the representation of collocations in the mental lexicon which can be drawn is that the links between collocations and meaning are weaker in the mental lexicon of learners as compared to native speakers.
Before presenting the results of a small-scale experiment I will try to define collocation competence. Collocation competence requires both the development of  awareness and a set of skills.
The awareness which is to be fostered has the following ingredients:

  • Within the mental lexicon, “collocation is the most powerful force in the creation and comprehension of all naturally-occurring text” (Lewis, 2000:45).
  • Collocation is a vital key to language learning. “Languages are learned collocation by collocation rather than word by word” (Palmer,1981:21).
  • Knowing the meaning of a word is useless unless you also know something of how the word is used, which means knowing something about its collocation field.
  • Collocation illustrates the idiom principle, which postulates that a language user has available to him or her a large number of semi-preconstructed phrases that constitute single choices, even though they might appear to be analyzable into segments (Sinclair, 1991:110).
  • Collocation is an essential carrier of meaning.
  • Collocations cannot be divorced from the grammatical context in which they occur.
  • Knowledge of the types of lexical collocations e.g. v + n; adj. + noun; adv. + adj.; v + adv.

The skills that are to be developed are:

  • Thinking bigger than the word – always to look for the two – or three-word expressions
  • Noticing words with the words with which they naturally occur and identifying different types of lexical collocations
  • Expanding knowledge of the collocation field of words already known
  • Recording collocations with their translation equivalents in L1 according to topic
  • Storing collocations as single choices which are easy to retrieve
  • ‘Gapping’ (Hill, 2000:59) verb + noun collocations i.e. identifying verbs and nouns that go together but are discontinuous for syntactic reasons and can be one or more words apart e.g. He had a business which he ran efficiently.; We have a bit of a confession to make.

To test my students’ ability to notice collocations in text, I devised  three tests which were done by a group of 20 Upper-Intermediate students, in which students were asked to:

  1. Identify and underline eight verb-noun collocations in a small-sized text (100 words);
  2. Identify and write the English equivalents of Bulgarian verb-noun collocations to be found in a short text (100 words);
  3. Identify verb-noun collocations which are discontinuous i.e. they are far apart in the sentence and the students had to match the verbs and nouns that go together. (See Appendix)

In the first test students were able to discover up to six collocations. Three students underlined words randomly, which means they were not aware what the task was all about despite the explanations. It seems that the concept of collocation is new to some students and they just can’t grasp what collocation is all about. There are two collocations which no student was able to notice: ‘give sb a ring’ and ‘meet a requirement’.  This is indicative of the fact that predicates with light verbs and eventive nouns have no psychological salience for the learners, which was further confirmed by a severe underuse of such constructions in their writing. Also, it can be inferred that collocations involving a verb in a figurative sense as in ‘meet a requirement’, appear to be less psychologically salient for learners.
In the second test the students had to find the English equivalents of the following Bulgarian collocations in a short text: отбелязвам напредък; отправям критика; постигам резултати; сблъсквам се с проблеми; сключвам договор; изготвям доклад. Astonishingly, only four students were able to identify precisely all six English equivalents of these collocations. For six of the students the task seemed to be daunting as they provided the following equivalents, which shows that their sense of collocation salience is absolutely misguided: ‘buoyant economy’ [make progress] for ‘отбелязвам напредък, which is an adjective-noun collocation, and ‘insulted by suggestions’ [experience problems] for ‘сблъсквам се с проблеми; ‘reflect  progress’ for ‘отбелязвам напредък’; ‘against the criticism’ [level criticism against] for отправям критика’.
Another finding is that students tend to lift the phrases from the text without any grammatical clear-up, which means that they are unable to notice and record correctly collocations on their own. For example, ‘progress made’ [make progress] for ‘отбелязвам напредък; ‘against the criticism’ or ‘the criticism leveled’ [level criticism against] for отправям критика; ‘compiling a report’ [compile a report] for изготвям доклад; ‘the problems they were experiencing’ [experience problems]; ‘able to produce results’ [produce results]. The last two examples show that students tend to pick up longer stretches of language, which are incorrectly chunked. Also, four students inserted articles in the collocation ‘make progress’, producing ‘make a/the progress’. Two students failed to identify the collocation ‘experience problems’ for сблъсквам се с проблеми’.
The third test proved the students’ inability to ‘gap’ or match verb-noun collocations when they are over a span of two or more words. Students were given two sets of ten separate sentences. In one of the sets it was the nouns that were underlined and they had to identify the verbs they combine with. The other set contained sentences where the verbs were underlined and students had to find the nouns they collocate with. Only 10% of the verb-noun collocations were correctly matched. Only two students were able to match correctly all the verbs and nouns. Again, the students’ sense of collocation salience proved to be misguided. Here are a few examples:

  1. That was the message I wanted to get across to all art lovers – that dolls are another art form. [The collocation that was identified was ‘get across art’ instead of ‘get across a message’
  2. Crucial to resolving this second question is the definition of autonomy that we adopt.
    [The first collocation ‘resolve a question’ was correctly identified while the second        ‘adopt a definition’ was incorrectly matched and ‘adopt an autonomy’ was     produced instead.]
  3. Any generalizations and predictions we may wish to make can therefore never provide us with the truth in particular instances.
    [The collocation provided was ‘make a provide’ which makes no sense at all. The two collocations in this sentence were completely unnoticeable: ‘make a generalization/ a     prediction.]

Since students appear to have a rather misguided sense of the ’perceptual salience’ (Schmidt, 1990:143) of collocations, as demonstrated in the three tests outlined above, it must be recognized that the starting point for improving learners’ collocation competence is helping them correctly identify collocations in text and develop their awareness of the psychological and syntagmatic salience of collocations.
Learning to identify collocations, and the underlying patterns which individual examples exemplify is one aspect of helping students to obtain maximum benefits from the input to which they are exposed. Therefore noticing collocations has to be a central pedagogical activity. Accurate recording of new collocations aids noticing and maximizes the chance of input becoming intake. As a matter of fact, noticing is very similar to consciousness-raising or as M. Lewis puts it, “consciousness-raising resembles noticing” (Lewis, 1997:52).The crucial role of language awareness and noticing was wonderfully captured by Dave and Jane Willis (1996). They believe that the successful learner is actively involved in looking for regularities in language data and in drawing conclusions from those regularities and teachers have an obligation to encourage this process. If we are successful in this, we will succeed in inculcating learning habits which will pay valuable dividends wherever and whenever the learner encounters language. As far as collocations are concerned, students should be encouraged to search to identify them and then make them part of their collocation repertoire. A primary purpose of developing collocation competence is to help learners make better use (i.e. notice, record and retain), for acquisition purposes, of all the collocations which they meet.
Schmidt (1990) proposes the crucial construct of noticing to start to account for the way in which a) not all input has equal value and b) only that input which is noticed then becomes available for intake and effective processing. Noticing operates as a necessary condition for effective processing to take place. A degree of awareness is important before material can be incorporated into a developing interlanguage system. Schmidt specifies what influences operate upon noticing. He discusses the following influences:

  1. The more frequent a form (a collocation in our case), the more likely it is to be noticed and then become integrated in the interlanguage system. If attentional demands from elsewhere are fluctuating, there will be occasions when a form is not noticed, but because it occurs more often, there will be a greater number of occasions, other things being equal, when it is.
  2. Perceptual salience. It concerns how prominent a form (a collocation) is in input. The more a form stands out in the input stream, the more likely it is that it will be noticed. If attentional resources are variable, forms which call attention to themselves and are perceptually salient will have a greater chance of impinging on consciousness.
  3. Instruction may play an important role. Input contains many alternative features for processing, and the learner’s task is to extract relevant features which can then be focused upon fruitfully. (Schmidt, 1990:143)

Schmidt claims that frequency and salience are clear examples of bottom-up processing – what the learner extracts from input is what is highlighted by its own qualities. Instruction can work in a more complex way by making salient the less obvious aspects of the input, so that it is the learner who does the extraction and focusing, but as a function of how he or she has been prepared. In a sense, learning is still input-driven (since the input is not being transformed) but it is the learner who chooses what to prioritize in the input. What has been unstructured, undifferentiated input becomes noticeable and analyzable, leading to future progress. In this view, the role of instruction is not necessarily therefore in the clarity of explanation it provides, but rather in the way it channels attention and brings into awareness what otherwise would have been missed. Equally, the ensuing noticing, provoked by instruction, does not cause learning to occur necessarily – it simply satisfies a first condition. Noticing has a mediating role between input and the operation of memory systems.
Certain factors make noticing more likely to occur, and therefore have an impact on the elements that a developing interlanguage system has to work with. There are two such factors – one concerned with input properties (frequency and salience) and one concerned with how instruction can influence predispositions to extract particular material from the input. Noticing impacts upon the memory systems in general. Some of the time working memory will be implicated, but on other occasions direct connection with long-term memory will reflect aspects of long-term memory in a state of high activation having an impact on what is currently being noticed.
Schmidt (1990) also discusses three other influences on individual differences (IDs) in processing ability: readiness, and task demands. Individual differences in processing ability concern the learner’s capacity to deal with the range of forms (collocations) in input. This seems to be an individual difference variable, in that some people will be more effective input processors than others, and be more able to notice, for given input, new forms (collocations) which may then be integrated into their language development. This might be because some people have greater working memory and attentional capacity or because the analytic processes within working memory are carried out at greater speed.
Schmidt’s fourth influence on noticing is the current state of the interlanguage system – readiness to notice. Noticing might be a function of what the internal structures or mechanisms are predisposing the learner to be ready to attend to. Schmidt’s claim that noticing also depends upon readiness implies that a prediction can be made about what the learner can profitably notice in the sense that the product of such noticing stands a chance of being incorporated into the interlanguage system, because it is the ‘next’ thing to be acquired. This can be illustrated by teaching a figurative sense of a collocation, which is well-established in the interlanguage. For example, intermediate students will know the collocation ‘arrive at + place’ e.g. ‘arrive at the airport’. When this collocation appears in the input the students are dealing with, the collocational field of ‘arrive at’ can be extended by its figurative senses as in ‘arrive at a conclusion’. Similarly, if students (even elementary students) encounter the collocation ‘meet people’ in the input they are using, the teacher can provide them with ‘deadline’ and ‘requirement’ as collocates of ‘meet’ thus supplying the learners with collocations which have great communicative value as they pack meanings that the learners will most often have to express in real life. In doing so, learners are provided with a shortcut to acquiring the collocation fields of verbs. Otherwise, they will have to wait for quite a while until they meet these collocates of ‘arrive at’ and ‘meet’ in the input in order to learn these collocations.
The last of Schmidt’s influences on noticing is task demands. Particular tasks may, through their characteristics, make certain language forms (collocations) salient. A focus on a particular form may be associated with the nature of a particular task which, as a result, makes targeted noticing more likely to occur.
In Schmidt’s terms noticing has a mediating role. It is not simply a function of input, but is also affected by learner factors which influence how input is processed. In this sense, noticing is the result of existing knowledge systems and processing capacities (top-down processing) constraining what the learner can attend to effectively.
I strongly support the view that the conscious noticing of features of the language (collocations in particular)  that learners meet does facilitate acquisition. Krashen’s claim (1985) that we acquire language in one and only one way, by understanding messages, provides a clear starting point from which to examine our presuppositions about how learners acquire language. His position is that a learner’s interlanguage is modified by meeting new language which lies on the edge of what the learner already knows in such a way that it is incorporated into the learner’s interlanguage so that it is available for spontaneous use. In other words, acquisition involves taking in new material and incorporating it into the knowledge or skills learners already have. Expanding learners’ language resources involves intergrating new language into their intergrammar and mental lexicon.
Although I support the idea of independent learning, if we want to turn the language learners meet – input – into language they acquire and have access to for spontaneous use – intake – they almost certainly need to “notice the linguistic wrapping in which the message is delivered” (M. Lewis, 2000:159). Exactly what noticing might involve, and what may help or hinder input becoming intake, is perhaps the most important of all methodological questions. At least, ”focusing learners’ attention explicitly on some aspect of the linguistic form of the input (collocations) is helpful in accelerating the acquisition process. “ (M. Lewis, 2000:160). The phenomenon of noticing can vividly be explained by ‘the journey-to-work-metaphor’: We all make a daily journey from home to our place of work. The route is completely familiar, and we could give someone else directions for the journey. However, we may not know the names of all the streets we drive or walk along. We have not noticed the names of some of the streets – they are irrelevant when we can achieve our global purpose without attending to such details. The global purpose of language is the communicating of messages; but the medium for doing it is language items – words and phrases – which may need to be noticed if they are to be acquired.
In normal language use, we are usually so predisposed to focus on the message, that the language in which it is delivered is frequently ignored, or, if presented in writing, transparent to the point of being invisible. Therefore, students might fail to notice unless the teacher provided guidance. It is my observation that even advanced and motivated learners often do not see the difference between their  inaccurate or unnatural collocations and a similar correct, natural version which expresses exactly the same content. For example, once a student argued with me that the awkward collocation he used in his essay ‘fall into a traffic congestion’ is as natural and effective as ‘get stuck in a traffic jam’.
Learners frequently do not notice the precise way an idea is expressed by a collocation, unless their attention is explicitly drawn to it. I will illustrate this by a sentence from an open-the-brackets-exercise for Upper-Intermediate students: ‘The Stonechurch estate’, answered Paul, aware of the reputation for crime it _______________ (acquire) in recent years. (New Headway Upper-intermediate) If I did not draw the learners’ attention to the collocation ‘acquire a reputation’ and did not elicit the Bulgarian equivalent спечелвам си репутация from the students, they would never notice it for themselves. All they care about is to get the verb form right. The sentence seems to be empty of lexical substance for them. Learners are usually pre-occupied with the grammatical features of an utterance and fail to capture the ‘lexical filling’, especially collocations, which usually express a common idea in a language-specific form. In doing so, they miss many opportunities for expanding their repertoire of collocations and other items of phrasal nature. Some training in the sorts of collocations and other chunks which make up the texts they read or hear increases the chance of them noticing useful language. As M. Lewis warns us, “Do not assume students are noticing collocations and recording them for themselves. They won’t unless you train them to” (M. Lewis, 2000:163).
In his article The Role of Consciousness in Second Language Learning (Applied Linguistics, 1990, Vol.11 Nr2) Schmidt points out the crucial difference between information that is perceived and information that is noticed:

“When reading, for example, we are normally aware of the content of what we are reading, rather than the syntactic peculiarities of the writer’s style, the style of type in which the text is set, music playing on the radio in the next room… However, we still perceive these competing stimuli and may pay attention to them if we choose.”
(Schmidt, 1990:139)

After a long discussion of the role of consciousness he concludes that “intake is what learners consciously notice” (Schmidt, 1990:142).
Teachers need to be proactive in guiding learners toward the input language which is important from an acquisition point of view. If learners do not notice the differences between the collocation they used to express something and the correct natural version expressing the same content, then that input cannot contribute to intake. Explicit noticing is a necessary condition to ensure that input becomes intake.
It is likely to be helpful to make learners explicitly aware of the lexical nature of language. This means helping learners develop an understanding of the kinds of chunks, collocations and prefabricated groups of words which are the prerequisite of fluency. This is one part of the teacher’s task in encouraging learners not to break the language they meet down too far. Discussing the value of instruction, of which noticing is a part, Diane Larsen-Freeman comments:

“Several researchers have pointed out that explicit grammar instruction will not likely result in immediate mastery of specific grammatical items, but suggest nevertheless that explicit instruction does have a value, namely, facilitating input.”
(Larsen-Freeman, 1977:60)

Although her comments relate specifically to grammar instruction, they surely apply equally to instruction which ensures learners notice any kind of patterning in the input they meet.
Noting multi-word vocabulary in exactly the form it is found in text, recording it, and trying to remember it in that form for re-use later has been on the periphery of language teaching, when in fact it deserves a central role.
To appeal to lexical collocation is not to appeal to something recently discovered. One is simply getting students to look at words in a way new to them, new even in their own language. It is a kind of  a ‘recollection’ of what they know already. Yet, although we appeal to what they already know, it often seems to them to be a new and exciting phenomenon. The newer it seems and the more exciting it is, the better, since what one wishes to achieve is a strategy that encourages motivation. One way to achieve it is through a word game proposed by Brown (1994:25):

Subject Verb Adjective Noun
Preposition Object

The object of the game is to surround the chosen noun with those linguistic elements that will constitute a collocationally correct sentence, e.g. The police are conducting a thorough investigation into the explosion, or, The company is making a careful investigation into consumer complaints. The point is to help students to see that learning more vocabulary is not simply a matter of learning new words, but of learning old words in new relations. Thus they can learn that both investigations and research are conducted but investigations are made while research is done.
Students can be taught to read to see how words hang together. They should be advised to read in manageable chunks, analyzing sentences, noticing how words co-exist with others. And all this is part and parcel of teaching them to look not simply for new words, but at words they know already; not simply at the words they know already, but at these in relation to other words, many of which they will also know already. This type of reading strategy was called ‘pedagogical chunking’ by M. Lewis (1997:52)
Brown (1994) suggests that students should be asked to adopt a Green Cross Code of Reading:

“When you see a word, even in particular a word with which you are already familiar, STOP, LOOK LEFT, LOOK RIGHT, LOOK LEFT AGAIN, AND, WHEN SATISFIED, PROCEED! This is a reading strategy which is, for natural reasons, unnatural – because it seems much more natural to pause in one’s reading when one does not know the meaning of a word; one stops to look up an unfamiliar word in a dictionary, or to give it a dirty look, or simply to pretend it’s not really there. One does not normally stop to take a wider look round a place that is already familiar. Familiarity breeds contempt. But contempt gets in the way. After all, it might strike one that what seemed too familiar to be worthy of note is not all that familiar at all – in just the sense that every “advanced” learner will know what investigation means but will not know how to use the word with the correct collocation, i.e. will not know how to use it correctly. The reading strategy now recommended entails the refusal to take things for granted.”
(Brown, 1994:26)

Brown goes on to say that “some students imagine they discern the cold light of structure and form wherever they look, and seek to apply the logical “must”, a notion of stark necessity, even to a language not their own” (Brown, 1994:27). He argues that it is not a question of belittling this type of student but of trying to illustrate the sociology of language language (any language), of bringing out the patterns where they may be discerned rather than of assuming that they can be constructed a priori. Brown concludes that “it seems perverse to undervalue the potential collocation offers as a means of accessibility to the multifarious uses of language” (Brown, 1994:27).
It is thus necessary to have access to two types of data: contrastive data and learner data, which will allow the selection of the most useful collocations for teaching purposes. To accomplish this purpose four major categories of data are necessary:

  1. Detailed descriptions of English collocations.
  2. Good descriptions of collocations in the learners’ mother tongue.
    These are necessary to assess the potential influence of the mother tongue and consequently to produce the appropriate pedagogical aids for specific mother-tongue groups.
  3. Bilingual computer corpora.
  4. Good descriptions of learner use of collocations.

On the whole, the starting point for improving learners’ collocation competence is getting students to notice and correctly record collocations, constantly revising them and encouraging them to use the collocations they have acquired in their output.


Test 1
Underline eight verb-noun collocations in this text.

My friend Beth is desperately worried about her son at the moment. He wants to enroll on a course of some sort but just can’t make a decision about what to study. I gave Beth a ring and we had a long chat about it last night. She said he’d like to study for a degree but is afraid he won’t meet the requirements for university entry. Beth thinks he should do a course in Management because he’d like to set up his own business in the future. I agreed that would be a wise choice.
(English Collocations in Use, Michael McCarthy & Felicity O’Dell, 2005, p.9)

Test 2
Read the text and find the English equivalents of the following verb-noun collocations:

  1. отбелязвам напредък – …………………………………………………
  2. отправям критика – ……………………………………………………..
  3. постигам резултати – ………………………………………………….
  4. изправям се пред проблеми – ………………………………………
  5. сключвам договор – ……………………………………………………
  6. изготвям доклад – ………………………………………………………

Exports increased in the last quarter due to the present buoyant economy. Ministers are content with the results, saying that they reflect the progress made in decreased regulation for small businesses.
Tax refunds are on the increase as invalid assessments multiply in the tax office. Tax officers protested against the criticism leveled against them, saying that they were insulted by suggestions that they were not able to produce the correct results. They said they were compiling a report which would present in minute detail the problems they were experiencing since the computer contract had been placed with another company.
Test 3
A: Read the sentences and find the verbs, which usually combine with the underlined nouns. For example: make a decision. Some sentences have more than one collocation.

  1. Crucial to resolving this second question is the definition of autonomy that we adopt.
  2. Any generalizations and predictions we may wish to make can therefore never provide us with the truth in particular instances.
  3. If you manage to persuade her that she is loveable without designer gear, then you will certainly do her a huge service.
  4. You mustn’t give in to temptation.
  5. They like to save their money but are shouldering heavy debts.
  6. The main parties are so much alike that it doesn’t make much difference who is in power.
  7. A lot of social problems never seem to get dealt with properly.
  8. The National Health Service doesn’t seem to work no matter how much money is thrown at it.
  9. People would soon kick up a fuss if they aren’t allowed to vote.
  10. As soon as my application has been processed, I’m going to leave the country and live in Canada.

B: Read the sentences and find the nouns, which usually combine with the underlined verbs. For example: make a decision. Some sentences have more than one collocation.

  1. In both countries, eye contact is avoided as a sign of respect.
  2. In Britain, you might have a business lunch and do business as you eat.
  3. His genius as an artist was soon recognized by many people.
  4. He survived two plane crashes.
  5. We have a bit of a confession to make.
  6. His father had a business, which he ran effectively.
  7. Exercise is important if we want to be healthy and fit, but only when it is performed in moderation.
  8. Pollution is one of the greatest dangers we must face.
  9. This of course, will eventually spell disaster for low-lying areas that will be flooded as sea-levels rise.
  10. The problem of too much trash with no place to put it is being faced by all modern municipalities.


  1. Brown, Phillip R. (1994): “Lexical collocation: A strategy for advanced learners”. Modern English Teacher, 3 (2), 24-27.
  2. Granger, S. (1998): Learner English on Computer. London: Longman.
  3. Hill, J. (2000): Revising priorities: From Grammatical failure to Collocational Success. In: Lewis, M. Teaching Collocation. Thomson & Heinle.
  4. Kaszubski, P. (2000): Selected Aspects of Lexicon, Phraseology and Style in the Writing of Polish AdvancedLearners of English: A Contrastive, Corpus-Based Approach. PhD Thesis.Poznan: Adam Mickiewicz University.
  5. Kjellmer, G. (1991): A Mint of Phrases. In: Cowie, A. P., Phraseology: Theory, Analysis, and Applications. Oxford: OUP.
  6. Krashen, S. (1985): The Input Hypothesis: Issues and Implications, Longman
  7. Larsen-Freeman, D. (1977): Chaos, Complexity, Science and Second Language Acquisition. Applied Linguistics, Vol 18/2.
  8. Lewis, M. (1997): Implementing the Lexical Approach. LTP.
  9. Lewis, M. (2000): Teaching Collocations. Further Developments in the Lexical Approach. Thomson & Heinle.
  10. Meara, P. (1984): The Study of Lexis in Interlanguage. In: Nesselhauf, N., Collocations in a Learner Corpus. (2005) Amsterdam: Benjamins.
  11. McCarthy, M. and F. O’Dell (2005): English Collocations in Use. CUP.
  12. Palmer, F. R. (1981): Semantics. A New Outline. Cambridge: CUP.
  13. Schmidt, R. W. (1990): “The role of consciousness in second language learning”. Applied Linguistics, Vol 11/2 , 129-150.
  14. Sinclair, J. McH. (1991): Corpus, Concordance, Collocations. Oxford: OUP.
  15. Soars, J. and Liz Soars (2003): New Headway – Upper-Intermediate. OUP.
  16. Willis, J. and Willis, D. (1996): Challenge and Change in Language Teaching. In: Lewis, M., Teaching Collocations. Thomson & Heinle.