| The purpose of this study was to test empirical sustainability of three major positions about the nature of input processing that is associated with the enhancement of second language learners' oral proficiency, seen from the implicit-explicit dimension. These three positions were formulated based on Ellis' (1993) distinction between no interface, strong interface, and weak interface positions. Each of the three positions makes a distinct prediction about the relationship between oral proficiency and the use of two kinds of language learning strategies (i.e., cognitive strategies and functional-use strategies). By using these predictions as research hypotheses, this study tested the sustainability of the three positions.; To test the hypotheses, 175 ESL speakers were randomly selected from East Asian international graduate students at The University of Tennessee, Knoxville. To measure the subjects' use of the two kinds of strategies, 31 items were selected from the Strategy Inventory for Language Learning (SILL), Version for English Speakers Learning a New Language (Oxford, 1990). To measure the subjects' English oral proficiency, a 20-item oral proficiency scale was created for this study. These two tools constituted a questionnaire together with several demographic items. Data were collected by mail survey. Four subjects turned out not to belong to the population. The return rate was 72.5%.; By factor analyzing the respondents' answers on the 31 SILL strategy items, nine strategy categories emerged. Cognitive strategies and functional-use strategies were defined using two and three of these categories respectively. The validity and the reliability of the 20-item oral proficiency scale were checked. Then, multiple R's were computed between oral proficiency and each of these five categories. All three categories that comprised functional-use strategies showed medium correlations to oral proficiency (R's between .564 and .622). Two categories that comprised cognitive strategies showed small correlations to oral proficiency (R's of .300 and .356). The result matched the prediction made by a weak interface position. Therefore, it was concluded that the nature of input processing that is associated with the enhancement of the studied population's oral proficiency was mainly implicit even though explicit processing did play a certain role in oral proficiency acquisition. |
