Four levels of native-likeness : a new method of assessing learner lexicons

This is a cross-sectional study that focuses on the assessment of the semantic network in the mental lexicon of second language learners. Generally, the number of words a learner claims to know is a reflection of achievement and advances in his vocabulary knowledge. However, the size of a learner's vocabulary knowledge only reveals half of the story about his lexical competence. An understanding of the quality of word knowledge allows for a clearer interpretation of the organisation and integration of lexical items in the mental lexicon. Typically, vocabulary researchers have utilised word association tests to measure vocabulary depth. However, there are several problems with the traditional methodology of this test, which has resulted in some inaccurate claims about a non-native speaker's lexical ability, such as the dichotomous native-like versus non-native-like categorisation of lexical ability. The aim of this thesis study was to test an alternative word association test that claims to capture partial native-like lexical knowledge in non-native speakers. Up to this point, this method had only been developed and tested on native speakers of English. The results of the analysis in this study reveal that this test does capture a range of native-like levels of lexical ability in non-native speakers of English. The findings suggest that this alternate word association measure of vocabulary depth could be used in conjunction with other language and vocabulary measures for both diagnostic and placement purposes. The information gained from this study provides a new approach to and understanding of vocabulary acquisition research and assessment