Acoustic correlates of L2 English stress ― Comparison of Japanese English and Korean English* (original) (raw)

Developmental change in English stress manifestation by Japanese speakers

2015

This study investigated the developmental change in the phonetic realization of English stress, using corpus data of Japanese English learners with varying proficiency. Previous studies demonstrated that highly proficient Japanese learners of English can produce native-like English lexical stress in terms of intensity, F0, and vowel duration, but not vowel quality. The results of our study showed that the contrasts in intensity, F0 and duration were manifested by all proficiency groups while spectral contrast was consistently absent. In addition, significant differences in vowel duration were found between native speakers and Japanese speakers of low-to-medium English proficiency level. The results imply that Japanese mora-timed rhythm is an obstacle in manifesting native-like lexical stress. These two findings suggest that it is more difficult to overcome L1 interference of segmental phonology than of suprasegmental phonology.

Is Acquisition of L2 Phonemes Difficult? Production of English Stress by Japanese Speakers

2009

This study examined the production of English lexical stress by Japanese speakers to determine which acoustic features associated with English lexical stress are difficult for Japanese speakers to acquire. Realization of lexical accent differs between languages. English is a stress-accent language where the accent is expressed by a combination of pitch, duration, intensity and vowel quality (Fry 1955, Gay 1978, Kochanski et al. 2005 & Kochanski and Orphanidou, 2008). In contrast, Japanese is a pitch-accent language where the accent is dominantly realized by a fall in the fundamental frequency (F0) from an accented high pitched mora to the following mora, and there is very little use of intensity to mark accent (Fujisaki et al. 1986). Studies on non-native English learners have shown that problems on acquiring English stress vary depending on their first language (L1). For example, Arabic speakers show different use of F0 and formant frequencies in stressed and unstressed vowels compared with native English speakers (Zuraiq & Soreno 2007). Lee et al (2006) also reported that Japanese speakers found it difficult to achieve vowel centralization of English unstressed vowels. This study investigated the effect of Japanese (L1) prosody on English (L2). In particular, the study focused on how Japanese speakers who speak fluent English still show the influence of Japanese phonology, and also examined whether there are any characteristics notably different from native English speakers' utterances. Different factors influence L2 speech rhythm. For example, different languages use different phonological units to keep speech rhythm. Japanese uses the mora as the fundamental unit for its speech rhythm and many studies have reported mora-based timing control for Japanese. The duration of each mora is not necessarily equal, but the duration of a word or phrase is determined by the number of morae in it (Port et al., 1987). There is mora-based segmental elasticity and durational compensation between CV which constitutes a mora rather than between V-C across a mora boundary (Campbell and Sagisaka, 1991; Sato, 1993). The effect of the mora on L2 has been reported in English (Mochizuki-Sudo & Kiritani 1991) and in French (Kondo & Shinohara, 2003 & 2006). Vowel quality is also an important cue for stress in English. Everything being equal, vowels in unstressed syllables are shorter and have centralized quality. Although the order of importance among these acoustic cues of English stress varies from study to study, all these are important cues to differentiate stressed and unstressed syllables in English (Fear et al., 1995). Considering the prosodic differences between English and Japanese, various factors affect L2 speech rhythm. An earlier study (Kondo, 2007) found an L1 influence in English speakers' Japanese utterances. English speakers showed strong influence of lexical accent on vowel duration in their Japanese utterances. They showed two patterns: (i) large F0 increase but little durational increase, and (ii) durational increase instead of any F0 increase. In the present study most of the Japanese subjects spent some years in English speaking environments and spoke relatively good English. However, they still showed some influence of their L1, Japanese, in their English utterances. Among these acoustic features, i.e. vowel duration, F0, intensity and quality, some features are easier to acquire than others and some take longer to acquire during the process of L2 sound acquisition.

Perceived Stress as the “Standard” for Judging Acoustical Correlates of Stress

The Journal of the Acoustical Society of America, 1974

Acoustical correlates of stress can only be evaluated in comparison with some °standard° specifying which syllables are actually stressed. The Standard should be consistent from time to time, and largely independent of talker and listener idiosyncrasies. Three phonetically-trained subjects listened to repeatedly spoken texts and spontaneous sentences until they could categorize each syllable as either stressed, unstressed, or reduced. This procedure was repeated three times for each speech text and listener. Two listeners differed from each other on only 5% of all syllables as to whether they were perceived as stressed or not. Each showed about 5% confusion in decisions about stressed syllables from one trial to another. Unstressed and reduced levels were confused more frequently. The third listener gave less consistent results. Subjects, judgements of stress when given only the written text were cf comparable consistency but did rot correspond well with perceptions with speech, if the speech was spontaneous rather than spoken texts. Stress perceptions consequenAy may be suitable for evaluating acoustical correlates to within a 5% tolerance in overall location scores. Pooling the perceptions from several trials and several listeners may improve the stability of this °standard° for stress assignment.