University classes are returning face-to-face from the emergent remote teaching for the COVID-19 infection control. However, some content for remote teaching can be applicable to face-to-face/hybrid teaching. This paper describes experiential exercises and experimental video content for university acoustic classes in a speech-language-hearing therapist program. The 30 classes were designed to learn relationships between acoustics and speech-language-hearing therapy and induce interest in academic inquiry. The approximate number of minutes for experiential exercises students’ interests is also reported.

Speech intelligibility is also an index of severity in dysarthia and its evaluation is important. However it is judged only by the auditory impression evaluation of the degree to which a listener understands the speech. The purpose of this study is to provide a method of assessing articulation and speech intelligibility more objectively not only in read speech but also in spontaneous speech. We proposes "non-segmented mora boundary ratio" (an appearance frequency of mora boundaries that are not segmented clearly on the spectrograms). This analysis method, which focuses on the segmentation of mora boundaries, can quantify the imprecise consonants and speech intelligibility in spontaneous speech.

Speech features have been found to affect the attitude we have towards the interlocutors. One of the speech perception research that has been gaining scholarly attention is the impact of listeners’ ability to correctly identify varieties they are asked to evaluate on their attitude (Carrie and McKenzie, 2018; Dragojevic et al., 2018; McKenzie, 2015). According to McKenzie (2015), listeners firstly categorise speech as either native or non-native speakers of the language, prior to a more detailed judgement. To gain more understanding of this mechanism and the social and educational impact of speech categorization, the present study attempts to investigate the role of (mis)categorisation of second language accents in an EFL context. Speech samples were elicited from 44 Japanese learners of English of different English proficiency and the samples were subsequently rated by 12 Japanese listeners on a 6-point semantic differential scale. Following the rating task, the listeners were invited to provide comments explaining their ratings. According to a series of statistical analyses, who “erroneously” judged the Japanese-accented speech as produced by a native speaker of English gave a more positive attitude evaluation overall. This study highlights the importance of raising awareness of linguistic stereotyping (Ferri and Magne, 2020) that appears contingent on further understanding of rater behaviour.

In this study, we compared the productions of three Japanese native speakers with five advanced level Chinese learners of Japanese of vowels that precede syllabic nasal /N/ with vowels that do not precede /N/. The 1st and 2nd formants of each vowel were measured. The formant values of the /a, i, o/ vowels were similar in all conditions. However, in the case of /u/ and /e/ for the Chinese learners of Japanese, the formants were remarkably different between vowels that precede /N/and those that do not. In addition, the articulatory movement of two Japanese native speakers and two Chinese learners were observed using ultrasound imaging. The tongue shape of /a,i,o/ by learners was similar to that of native speakers; however, the tongue movement of learners for /u, e/ showed differences from those of native speakers. Specifically, for Chinese speakers, /u/ was produced more frontward or backward depends on the initial consonants, and /e/ was pronounced with the more backward tongue position. And the tongue shape of pre-/N/ vowels were different from those of the single vowel condition. Japanese native speakers, on the other hand, showed stable tongue position during both when /N/ followed the vowels and when it did not. This study revealed that even advanced learners showed different articulatory movement for producing vowels, especially if the vowel precedes /N/. Our results indicate that the native language of the speaker affects tongue articulation of vowels, especially before syllabic /N/.