For decades, vocabulary size tests have been built upon the idea that if a test-taker knows enough words at a given level of frequency based on a list from corpus, they will also know other words of that approximate frequency as well as all words that are more frequent. However, many vocabulary size tests are based on corpora that are as out-of-date as 70 years old and that may be ill-suited for these tests. Based on these potentially problematic areas, the following research questions were asked. First, to what degree would a vocabulary size test based on a large, contemporary corpus be reliable and valid? Second, would it be more reliable and valid than previously designed vocabulary size tests? Third, do words across, 1,000-word frequency bands vary in their item difficulty? In order to answer these research questions, 403 ESL learners took the Vocabulary of American English Size Test (VAST). This test was based on a words list generated from the Corpus of Contemporary American English (COCA). This thesis shows that COCA word list might be better suited for measuring vocabulary size than lists used in previous vocabulary size assessments. As a 450-million-word corpus, it far surpasses any corpus used in previously designed vocabulary size tests in terms of size, balance, and representativeness. The vocabulary size test built from the COCA list was both highly valid and highly reliable according to a Rasch-based analysis. Rasch person reliability and separation was calculated to be 0.96 and 4.62, respectively. However, the most significant finding of this thesis is that frequency ranking in a word list is actually not as good of a predictor of item difficulty in a vocabulary size assessment as perhaps researchers had previously assumed. A Pearson correlation between frequency ranking in the COCA list and item difficulty for 501 items taken from the first 5,000 most frequent words was 0.474 (r^2 = 0.225) meaning that frequency rank only accounted for 22.5% of the variability of item difficulty. The correlation decreased greatly when item difficulty was correlated against bands of 1,000 words to a weak r = 0.306, (r^2 = 0.094) meaning that 1,000-word bands of frequency only accounts for 9.4% of the variance. Because frequency is a not a highly accurate predictor of item difficulty, it is important to reconsider how vocabulary size tests are designed.

L2 (second or foreign language) research indicates that vocabulary knowledge is not only the “single best predictor of text comprehension,” but also a strong indicator of listening, speaking, and writing proficiency (Alderson, 2000, p. 35). Understanding the development of vocabulary knowledge, including both vocabulary size and vocabulary depth, or quality of vocabulary knowledge—is therefore essential to the building of an overall insight into L2 proficiency. This study aims to explore the developmental status of vocabulary depth among postsecondary CFL (Chinese as a foreign language) learners of higher proficiency levels who have studied Chinese for over four years. In particular, it focuses on these learners' identification of two types of word association—synonym and collocational associations and how factors such as association type and target-word frequency impact association identification. The process and strategy use that are involved in the inference of word association are also explored. For these purposes, this study employs a combination of quantitative and qualitative research methods. Cross-sectional data were collected through a paper-and-pencil test of Chinese word associates from seventeen learners within five postsecondary CFL programs in the U.S. Each participant worked on two test booklets assessing synonym and collocational associates respectively for the same 44 adjectives selected from the three word frequency levels of below 1000, between 1000 and 5000, and above 5000. A two-factor within-subjects ANOVA revealed both significant main effects for association type and word frequency on association identification and a significant interaction between the two. Simple effect analysis and pair-wise comparisons further revealed that association identification became increasingly stronger with the increase of word frequency for collocational association, yet remained non-impacted by frequency before reaching the mid- to high-frequency transition for synonym association. Meanwhile, CFL learners' collocational knowledge was significantly higher than synonym knowledge at mainly the medium- and high-frequency levels. These reslts indicate that synonym knowledge seemed to lag behind in development as familiarity with words increased, but began to catch up at higher-frequency levels. Interview data collected from six CFL learners show that they employed a wide variety of knowledge sources, such as radical knowledge, morphological knowledge, contextual clues, sound information, or L1 in inferring word association. Inference success seemed to be influenced not only by their preexisting word knowledge, but also an integrated and flexible use of linguistic and contextual information in the inference process. Implications of these findings are discussed in relevance to curriculum and pedagogical development of CFL teaching and the understanding and definition of CFL proficiency in general. This study fills a gap in CFL vocabulary research by building a tentative measure of vocabulary depth and bringing greater insights into the developmental status of higher-level CFL learners in synonym and collocational association as well as the process that is involved in inference of word association.

Résumé : En raison de la concurrence qu’il y a au Québec entre une norme québécoise et une norme internationale, les questionnements normatifs sont nombreux chez les professionnels de la langue, et tout particulièrement chez les enseignants de français. Dans cette étude, nous avons cherché à voir comment les représentations de la norme lexicale chez les professeurs de français langue d’enseignement au secondaire se manifestent dans leur discours à partir d’usages dont la valeur normative n’est pas clairement définie. Pour ce faire, nous avons mené des entrevues auprès de 20 enseignants et enseignantes de français au deuxième cycle du secondaire au Québec au cours desquelles nous leur avons demandé de corriger de courts textes et de justifier leurs corrections. Puis, nous les avons interrogés sur leur conception de la norme lexicale écrite et sur leurs positions normatives à l’endroit de diverses catégories d’emplois lexicaux. Il en ressort que cette norme est le français québécois standard tel qu’on peut le dégager des dictionnaires les plus usuels ou encore perçu selon l’intuition linguistique des participants. Ce français se distingue du français oral et il exclut les anglicismes critiqués, les mots vulgaires, les impropriétés et les barbarismes. Si certains types d’emplois sont clairement jugés non compatibles avec la norme, d’autres donnent lieu à des prises de position normatives moins tranchées, plus instables. Ainsi, les néologismes, les archaïsmes, les mots familiers et les emprunts peuvent être acceptés, selon la situation de communication et selon des critères comme l’usage et l’absence de synonyme. Le modèle de l’Imaginaire linguistique d’Houdebine nous a servi de grille pour catégoriser et analyser les représentations linguistiques observées dans le discours des participants. Cette analyse révèle que les enseignants s’appuient principalement sur des critères d’ordre prescriptif et constatif, soit le dictionnaire et l’usage, et parfois aussi sur d’autres critères comme la précision, l’intercompréhension, la fréquence ou la valeur identitaire de certains usages lexicaux pour décider de les corriger ou de les accepter lors de la correction de productions écrites. / Abstract : Due to a competition between two linguistic norms in Quebec, a Quebec norm and an international norm, language professionals – especially French teachers – often deal with prescriptive questions. Our study aims at studying how high school French language teachers’ attitudes towards the lexical norm are reflected in their discourse based on examples of uses without a clearly defined prescriptive value. To do so, we conducted individual interviews with 20 Secondary French teachers (Cycle Two) in Quebec. During these interviews, they were asked to correct short texts and to justify their corrections. Then, they were questioned on their conception of the written lexical norm and on their normative positions towards various categories of lexical uses. This research reveals that their norm is the Quebec French standard as it is described in dictionaries or, in many cases, perceived according to the informant’s linguistic intuition. This French variety is distinguished from spoken French and excludes criticized anglicisms, coarse language, language errors and barbarisms. Although some categories of uses are clearly deemed incompatible with the norm, not all categories are given a clear-cut, stable stand. Thus, neologisms, archaisms, colloquial words and loanwords may be accepted, depending on the context and based on certain criteria such as usage and non-existence of synonyms. Houdebine’s model of linguistic imaginary served as a grid to classify and analyze the linguistic representations observed in the participants’ discourse. This analysis shows that teachers rely heavily on prescriptive and constative criteria, that is to say dictionaries and usage, and sometimes also on other criteria such as accuracy, mutual understanding, frequency or identity value of some lexical uses in order to decide whether to correct them or not while assessing written productions.