| dc.description.abstract | The topic of the present thesis is the English progressive construction BE + V-ing; more specifically, as used by young L1 Norwegian learners of L2 English and same-age nativespeakers. The study is form-based and thus explores the developmental path from non-targetlike use of the -ing form to (more) target-like use of the progressive as a complete – and meaningful – construction. To this end, 165 learner narratives were collected using a website constructed to this end; 89 written by Norwegian L2 learners and 76 by L1 speakers of English. The learners were selected from two age groups, around 11 (N=83) and 15 (N=82), and texts written by the L2 learners were assessed to be at proficiency levels A1 (N=12), A2 (N=51), B1 (N=25) and B2 (N=1) of the Common European Framework of Reference for Languages (CEFR). The thesis is theoretically grounded in the framework of Cognitive Linguistics (e.g. Langacker 1987, 1991, 1999, 2008a, 2009, Taylor 2002, Croft and Cruse 2004, Radden and Dirven 2007), which comprises construction grammar (e.g. Goldberg 1995) and related learning theories (e.g. Tomasello 2003a, Robinson and Ellis 2008). In particular, this work draws on Slobin’s (1987b, 1996) thinking-for-speaking (TFS) theory, which states that language users select ways to represent a situation based on the linguistic tools available in their first language. Such TFS patterns are difficult to restructure in a second language, but it is possible, as found in studies and reference works that draw on Slobin’s theory (e.g. Jarvis and Pavlenko 2008, Pavlenko 2011b, Athanasopoulos 2011). In addition to more general language learning theory, one hypothesis specifically targeting the learning of tense/aspect morphology is addressed in the present work. This is known as the Aspect Hypothesis (AH), which posits a strong tie between the semantic category activities and progressive aspect: both L1 and L2 learners have been shown to display sensitivity to this connection early on, and only later expand the use of the progressive to the categories accomplishments and achievements, while refraining from use with states altogether. The data in the present study are analyzed with respect to these semantic categories, as well as the temporal category TENSE, the non-temporal category ING (the -ing form in non-finite contexts), and the extralinguistic categories L1 (Norwegian or English), age, gender and proficiency level. The analysis is quantitative and aims to find if differences in these variables lead to different frequencies in the use of the progressive; statistical significance is tested by means of a multivariate regression analysis. Finally, a more detailed semantic analysis is performed, to find out whether various previous claims about the semantics of the progressive have any explanatory power when it comes to learner usage in the dataset in this study. Traits that are said to be characteristic of verb types used with the progressive are duration, atelicity, agentivity, and the fact that they refer to observable situations. Such visual salience is seen as a very important characteristic in Durst-Andersen (2000), who claims that situations are perceived as still or moving pictures by language learners and that the progressive favours the latter. The above speaks to a universal learning path when it comes to progressive aspect, as do findings and theories related to the frequency and salience of the progressive (e.g. Ellis 2002, Goldschneider and DeKeyser 2005), as well as well-known studies on acquisition order (e.g. Brown 1973, Dulay and Burt 1974a). However, recent years have seen a rising interest in the role of L2 learners’ first language (e.g. Jarvis and Pavlenko 2008) and even in studies that mostly point toward a universal learning path, there is evidence of L1 influence on a more detailed level (e.g. Collins 2002, 2004b, Rocca 2007, Rohde 1996). A chapter section is therefore devoted to a partial contrastive analysis of Norwegian and English, in order to establish potential candidates for transfer. The analysis also seeks to detect any such transfer, based on the methodological criteria proposed by Jarvis (2000). The main findings of the present study can be summarized as follows: the single greatest predictor of frequent use of the progressive in a learner narrative is the presence of activity verb phrases. Other significant factors are L1, age, and proficiency level and tense, but none of these matches activities in explanatory power. Moreover, learners at proficiency level B1 and above master the use of the progressive construction on a level comparable to that of same-age native speakers; this includes the relative frequencies of the progressive and use of -ing in other constructions than the progressive. Their usage also bears witness of a gradual spread from more to less progressive-like constructions, while less proficient learners still struggle to use auxiliary BE. Finally, the semantic perceptions change with age and proficiency level, from an initial focus on observable or otherwise experienced events – what I call referential salience – via the expression of spatial as much as temporal extension, to giving telicity and agentivity more prominence. A chapter that summarizes and discusses the findings against the theories presented in the first part of the work follows the analysis, before the thesis ends with some concluding remarks. Among other things, suggestions are made as to how the findings in this study may inform pedagogical practice, and which areas of further research should receive attention. | en_US |
