An evaluation of TTS as a pedagogical tool for pronunciation instruction: the ‘foreign’ language context

Abstract

Despite positive evidence demonstrating the pedagogical benefits of Text-To-Speech (TTS) synthesisers for second/foreign language learning (Liakin, Cardoso, & Liakina, 2017), there is a need for up-to-date formal evaluations, specifically regarding its potential to promote learning. This study evaluates the voice quality of a TTS system in comparison with a human voice, and examines its pedagogical potential for use in an English as a Foreign Language (EFL) setting in terms of speech quality, ability to be understood by L2 users, and potential to focus on a specific language form. EFL learners in Brazil completed four tasks to evaluate the quality of TTS-generated texts. Results suggest that the TTS voice performed equally as well as the human voice in almost every assessment measure, demonstrating a high level of intelligibility and the ability to provide learners with opportunities to notice linguistic forms.

Can TTS help L2 learners develop their phonological awareness?

“…Recently, there have been studies that evaluate the voice quality of TTS systems, corroborating previous hypotheses that current synthesized voices are ready for use in L2 education: they are not only perceived as appropriate (or 'good enough') from a sound perspective (Bione, Grimshaw, & Cardoso, 2017;Cardoso, Smith, & Garcia Fuentes, 2015), but they are also more likely to enhance opaque features of the target language (e.g. regular past tense marking in English is often detected with higher accuracy in synthesized speech; Bione et al, 2017;John & Cardoso, 2017). One area that has not received consideration by researchers is TTS' ability to raise L2 learners' PA.…”

“…in comparison with spontaneous production). These findings suggest that, considering these two measures, TTS constitutes a good source of aural input (as reported Cardoso et al, 2015 andBione et al, 2017), and that it may be used to enhance the opaque alternations found in past -ed marking in English (as hypothesized, based on Bione et al, 2017, andJohn &. We acknowledge some of the limitations of our study, particularly the limited number of participants and lack of statistical rigour in our analysis.…”

“…Based on Cardoso et al (2015) and Bione et al (2017), participants also performed an auditory identification task, in which they heard 16 sentences, 12 of them in the past (four distractors), but without lexical elements that could identify their tense.…”

A Comparison of Synthetic and Human Speech: an Evaluation by English as a Foreign Language Students in a Public Costa Rican University

“…Several studies have found potential benefits from using TTS in language classes or have found no significant differences between using a synthetic or human voice (Bione et al, 2017;Cardoso et al, 2015;Craig & Schroeder, 2019;Hillaire et al, 2019;Kang et al, 2008) and the possibility of more interactive models where people can keep conversations in real time with machines (Ku-mar et al, 2023). In addition, TTS systems may use various techniques to improve the quality and naturalness of the generated speech, such as adjusting the rhythm and intonation to match that of a human speaker or adding natural-sounding pauses and inflections.…”

“…Therefore, the potential application of TTS may impact several language programs. Regardless of the potential benefits of implementing TTS in ESL settings (Bione et al, 2017;Cardoso et al, 2015;Craig & Schroeder, 2019;Hillaire et al, 2019;Kang & chatGPT, 2023), currently, the only discussion about TTS systems in the English major has centered on using them to read abstracts. No other language program in the university where this study took place has considered using TTS synthesis, and no attention has been paid to the use of computer-generated voices in educational settings, especially in ESL settings where listening materials are abundant but sometimes difficult to find or do not fully adapt to the course requirements.…”