| The study explores, from both a sociocultural and an information processing viewpoint, the role of vocalization, to the self and to others, in regulating the L2 vocabulary retention of adults studying English for Academic Purposes. I asked eight participants to learn five previously unknown words working alone and five different new words in collaborative dyads. In each condition, I audio-taped them as they studied the words from a text and a dictionary, completed a written crossword puzzle, answered oral questions, and did a stimulated recall (SR). I identified three types of vocalizations: solitary private speech, collaborative private speech, and collaborative social speech. I grouped task completion and SR speech into Vocabulary Related Episodes (VREs), which I then divided into smaller behavioural units called Moves to facilitate analysis of their frequency, meaning-form focus, and processing depth. Three progressively deeper levels of processing were identified: Repetition, Manipulation, and Generation. Tests given one week and one month later assessed participants' retention of the vocabulary they had discussed during task completion.; Data analysis confirmed three of the study's four predictions. First, verbalization during vocabulary learning helped participants orchestrate procedure, release emotion, establish group intersubjectivity, and imitate, monitor, test out, elaborate, recast, reformulate, transform, and create L2 word knowledge. Second, the frequency, meaning-form focus, and processing depth of verbalizations were, as anticipated, influenced by type of speech, prior education, learning style, L2 proficiency, task demands, and group dynamics. Third, participants' written test responses showed that they remembered what they had vocalized during task completion. Recall seemed most evident when the vocalizations featured Manipulation and Generation processing that deployed three elaborative word-learning strategies: the creation of mnemonic devices, the connecting of input with L1/L2 knowledge, and the expression of personal opinions triggered by the new words. There was a significant inverse correlation between Repetition and delayed test scores during both solitary and collaborative study. Manipulation and Generation correlated positively with delayed scores, but for the solitary condition only. The fourth prediction, which anticipated that collaboration would lead to better long-term retention, was not borne out. Both conditions were equally effective in the short and long run. |
