The major task of second language collocation research is to discover what it means for L2 learners to acquire a collocation, how they learnit, and what problems they encounter in acquiring a collocation. Previous research has provided a comprehensive description of how L2 learners use collocations and what problems they encounter in using them. Yet little is known with regard to factors responsible for the stagnant development of collocation knowledge. Hence, this study is intended to investigate factors that are associated with this collocation lag. By examining the factor(s) thatis/are responsible for the lag in collocation knowledge, we can better understand the process of collocation learning. Furthermore, more knowledge on how collocation is acquired can further shed light on how collocations are best learnt and taught.

As regards which factor(s) may be responsible for collocation lag,we hypothesise that vocabulary growth is an inhibiting force in collocation learning. So a general question is asked: is vocabulary growth an inhibiting factor in the learning of collocations by L2 learners? This question needs to be further divided into detailed questions by taking particular types of collocations as examples. For this purpose, one most important and frequent type of collocation — verb + noun (henceforth: VN) collocations is first selected.

The relationship between verb increase and the production of VN collocations is examined among verb + noun collocations produced by different levels of learners. At a macro level, verb increase is measured in terms of the development from delexical to lexical verbs. Verbs are divided into two categories according to the semantic contents they take: delexical verbs (do,make, take, have, give and get) and lexical verbs (acquire, fulfil, perform,etc.). Accordingly, VN collocations are divided into delexical verb + noun (DeLexVN) collocations and lexical verb + noun (LexVN) collocations. At a microlevel, the growth of verbs is measured in synonym sets (Fellbaum, 1998) inspecific VN collocations. The following developmental patterns with regard tothe increase of verbs from delexical to lexical verbs are hypothesised, such that lower levels of L2 learners make more errors with delexical verbs andhigher levels make more errors with lexical verbs in VN collocations.

The hypothesis on the developmental patterns is closely linked with the general hypothesis that verb increase is ahindrance in collocation acquisition. More precisely, at lower stages of L2 development, due to their limited mastery of verbs, learners resort to delexical verbs to collocate with a noun instead of a specific lexical verb. As their verb vocabulary grows, they have more access to lexical verbs and tend to make more collocation errors with lexical verbs, because the increase in synonymous verbs allows more chances of incorrect verb choices. The increase inlexical verbs and the subsequent occurrences of errors with lexical verbs suggest that vocabulary growth impedes collocation acquisition. To test whether the growth of verb vocabulary constitutes an inhibiting force in collocation learning, the relationship between vocabulary growth and collocation development has to be viewed locally in specific VN collocations, through locating the semantic domains of verbs in collocations where there is anincrease in verbs and examining whether the increase in these verbs subsequently leads to collocation errors.

Based on the two hypotheses, the following research questions willbe addressed:

1) What developmental patterns appear in the verb + noun collocations produced by L2 learners, in terms of delexical verb and lexicalverb + noun collocations?

Is there a tendency towards increasing use of lexical verb + noun collocations with rising proficiency?

Is there a tendency towards increasing errors with lexical verb +noun collocations and decreasing errors with delexical verb + noun collocations with rising proficiency?

2) Within specific semantic domains of the verbs in verb + noun collocations used by all levels of learners, is there a tendency for these verbs, as they increasingly occur at the higher levels, to be associated with collocation errors?

Research questions 1) and 2) are interrelated as they bear therelation of the whole and a part. They are both concerned with the increase of lexical verbs and the production of verb + noun collocations at the higher levels. Research Question 1) addresses the relationship between the overall increase of lexical verbs at the higher levels on the whole and the increasing/decreasing trend of verb + noun collocation errors associated with lexical verbs; the scope of research concerning verb increase and collocation errors is further narrowed down in Research Question 2), which is aimed at aconcrete investigation of the increase of lexical verbs within particular semantic domains. Through confining the verb increase into semantic domains,the study sets out to examine if verb increase in a semantic domain is a factor associated with the lag in verb + noun collocational knowledge. The particular focus on verb increase in semantic domains in Research Question 2) is built on the belief that learners may be confused with semantically related words (e.g.acquire and obtain) rather than words falling in different semantic domains(e.g. acquire and change) in producing verb + noun collocations.

Furthermore, two other common types of collocations — adjective +noun and noun + noun collocations — will be examined, so as to compare there sults with verb + noun collocations. Similar research questions on verb +noun collocations will be addressed with regard to adjective + noun and noun +noun collocations. Specifically, we focus on the following questions:

3) Are adjective + noun and noun + noun collocations produced by Chinese L2 learners at the same accuracy level as verb + noun collocations? If not, what patterns do they follow?

4) Within specific semantic domains of the adjectives in adjective +noun collocations and nouns in noun + noun collocations used by all levels of learners, is there a tendency for these adjectives/nouns, as they increasingly occur at the higher levels, to be associated with collocation errors?

As another field of enquiry, this book will investigate the role ofL1 in the production of congruent and non-congruent L2 collocations. Congruent collocations refer to collocations whose word elements in one language have direct word-for-word translational equivalence in another language; if word elements in one collocation do not share direct word-for-word translational equivalence between two languages, then it is considered as a non-congruent collocation (Nesselhauf, 2005; Wolter and Gyllstad, 2011). Previous studies show that congruent collocations are much easier than non-congruentones (e.g. Bahns, 1993; Nesselhauf, 2005), and non-congruent collocations once acquired, are processed independently of the learners’ L1 (Yamashita and Jiang, 2010; Wolter and Gyllstad, 2011). In light of these findings, we set out to examine Chinese L2 learners’ performance on congruent and non-congruent collocations, with the aim to test whether congruent collocations are easier than non-congruent ones for them, and whether non-congruent collocations, once acquired, are less prone to errors.

The present research is therefore an empirical study of the phraseological performance (verb + noun collocations in particular) of Chinese learners of English across different proficiency levels. This investigation will provide insight into how collocations are acquired by EFL learners, aquestion which has not yet been addressed but is of central importance for acomprehensive understanding of second language collocation learning. The study will shed light on the key question that researchers in SLA attempt to answer, expressed as “What is acquired? What is not acquired? Why so?” by Gries (2008:407).

The objective of the research project was an empirical study of theuse of collocations by Chinese L2 learners. However, in order to describe non-native phraseological competence, it is firstly necessary to establish native-speaker norms in this regard. The native speaker and non-native speaker dichotomy is contentious. Davies (2003: 1) introduces the common-sense concept of native speakers, referring to“people who have a special control over a language, insider knowledge about their language” and are the “models we appeal to for the ‘truth’ about the language”. However, as Davies (2003) acknowledges, he can more easily define what a non-native speaker is than a native speaker, and he even argues that the native speaker is a myth. Those who have two native speaking parents, both preferably monolingual, and are raised in a native speaking community, can still not be definitely defined as native speakers of that language, since other social factors like mobility and the rise of new Englishes are at play(Davies, 2003). As English is becoming a lingua franca, and an increasing number of proficient academics whose first language is not English enter English academia (Hyland, 2006), it is even harder to define what a native speaker is. The theoretical aspects of the native speaker construct will not beaddressed in this study.

Nevertheless, for the investigation and description of learner interlanguage, language learning goals in terms of nativespeaker norms, need to be set. Two widely-used English collocation dictionaries and the British National Corpus, a collection of the texts in British English, were taken as a kind of target norm for L2 English learners. Language forms produced by L2 learners that conform to the norm were regarded as well-formed, and those that deviate from the norm were viewed as erroneous.

The book is divided into ten chapters. Chapters 2 and 3 discuss previous theoretical and empirical studies on (L2) collocations. Chapter 2 highlights the importance of collocation and clarifies the notion of collocation. The significance of collocations is discussed in terms of their prevalence in native-speaker texts and their importance for a fluent and idiomatic control of English for L2 learners. Then the notion of collocation is examined on the basis of previous different approaches, and a definition and classification applied in this study are presented. Chapter 3 reviews previous collocation studies in second language learner English. In this chapter, the methodologies commonly adopted in L2 collocation studies are firstly addressed, with a view to introducing the methodology that has been more and more widely used in the analysis of collocations in learner corpora; then major findings ofprevious L2 collocation studies are discussed. Chapter 4 presents the detailed design of the present cross-sectional study of Chinese EFL learners collocation performance. The learner corpus chosen for such an investigation, the types of collocations targeted, the sources of reference in extracting these collocations, and the procedures for collocation extraction and analyses are introduced. Chapters 5, 6, 7, 8 and 9 contain a detailed analysis of the data. In Chapter 5, the overall picture of Chinese L2 learners’  performance in verb + noun collocations isdepicted, with the main focus on the developmental patterns of collocation production from delexical verb + noun to lexical verb + noun collocations. Chapter 6 is devoted to an investigation of the relationship between verbincrease in specific synonym sets and collocation uses associated with verbs inthese synsets. Moreover, an alternative explanation, i.e. the learning of new nouns in collocation production, is made in order to see whether the acquisition of new nouns is responsible for a lag in collocation. Chapter 7goes on to explore learners’ performance on two other important and frequenttypes of collocations, i.e. adjective + noun and noun + noun collocations. It aims at a corroboration of findings from verb + noun collocations. Chapter 8 presents and compares learners’ performance on verb + noun, adjective + noun and noun + noun collocations. In Chapter 9, cross-linguistic influence in the production and learning of L2 collocations is investigated. It addresses the role of L1 in learners performance on congruent and non-congruent collocations, and the role of L1 in different word-class collocations such as verb + noun and adjective + noun collocations. Chapter 10, finally, concludes the study by summarising its findings, discussing both theoretical and pedagogical implications for effective collocation learning, acknowledging the limitations of the present study and putting forward proposals for future research.