Electronic Thesis and Dissertation Repository

Thesis Format

Integrated Article


Doctor of Philosophy




Webb, Stuart

2nd Supervisor

Saito, Kazuya


University College London



This dissertation investigates the effects of three input factors—frequency of exposure, acoustic variability, mode of input— on learning productive knowledge of spoken forms of second language (L2) words. This thesis takes an integrated article format and is organized into (a) introduction (Chapter 1), (b) three main studies (Chapters 2, 3, & 4), and (c) conclusion (Chapter 5). The three studies involved Japanese university students learning 40 unfamiliar English words through encountering their spoken forms (and written forms in Study 3) while viewing pictures that conveyed their meanings. A picture-naming test was administered before, immediately after, and approximately one week after the treatment and the elicited speech samples were assessed for pronunciation and form-meaning connection. Study 1 (Chapter 2) investigated the effects of repetition. Seventy-five participants were randomly assigned to one of three conditions (1, 3, and 6 encounters), and their performance was assessed for pronunciation (accentedness, comprehensibility, processing time) and form-meaning connection (spoken form recall). Results showed that the number of exposures positively affected measures of form-meaning connection and pronunciation. Measurable learning gains occurred for comprehensibility after three encounters, while six encounters were necessary for foreign accent to be significantly reduced. Study 2 (Chapter 3) investigated acoustic variability and frequency of exposure. Eighty participants were randomly assigned to one of four conditions (3 encounters, 6 encounters 3 encounters with talker variability, and 6 encounters with talker variability). Spoken form recall and word stress accuracy were assessed. Results suggested that frequency of exposure promoted form-meaning mapping to a greater extent than talker variability, whereas talker variability had a stronger influence on word stress accuracy than frequency effects. Study 3 (Chapter 4) investigated input modality. Seventy-five participants were randomly assigned to one of three conditions (reading-while-listening, reading-only, listening-only). The elicited speech was assessed for spoken form recall, accentedness, and comprehensibility. Results showed that the reading-while-listening group outperformed the listening-only group in form recall. The reading-while-listening and listening-only groups sounded more nativelike and comprehensible compared to the reading-only group. This dissertation concludes with implications for researching and teaching L2 vocabulary as well as suggestions for future studies.

Summary for Lay Audience

Learners are likely to pick up second language (L2) words through seeing or hearing the forms of new words while reading books, watching television, and listening to songs. Encountering words in speech and writing therefore is an important source of input for learners to build L2 word knowledge. This dissertation explores the best ways to optimize input for enhancing productive knowledge of spoken forms (i.e., pronunciation) of unknown L2 words. It consists of three empirical studies focusing on Japanese university students studying 40 unfamiliar English words while viewing their meanings conveyed through their pictures. Before, immediately after, and approximately one week after the treatment, participants completed a word production test, and the elicited speech was evaluated for pronunciation accuracy. The word learning format (i.e., paired-associate learning) and test format (i.e., picture-naming test) were identical across the three studies but different in the input participants encountered during the treatment (i.e., input repetition, input variability, input modality). Study 1 investigated the extent to which learners can improve pronunciation of L2 words after hearing the spoken forms of target words repeatedly (1, 3, and 6 encounters). Study 2 investigated the extent to which learners benefit from hearing L2 words produced by different speakers (3 and 6 encounters with speaker variability) versus a single speaker (3 and 6 encounters without speaker variability). Study 3 investigated how mode of input (reading, listening, reading while listening) affects pronunciation of L2 words. Results showed that learners tended to be more accurate at pronouncing L2 words after hearing words repeatedly, and the effect of repetition was enhanced when learners heard the spoken forms produced by multiple speakers compared to a single speaker. Learners encountering spoken input were also better able to pronounce the words than learners encountering only written input. These findings suggest the importance of input repetition, input variability, and input modality for developing productive knowledge of spoken forms of L2 words. To conclude, I discuss several implications for researching and teaching the spoken forms of L2 words.