Perception and Production of Plosives in Putonghua as a Third Language by Indonesian Speakers

Abstract

Studies on third-language (L3) acquisition have reported mixed roles of the first language (L1) and second language (L2) on L3 perception and production. The current dissertation examined the effects of phonetic and phonological (dis)similarity in previously-known languages on L3 speech acquisition. This study focuses on the plosive distinction in three languages that all employ a two-way phonological contrast but with varying phonetic realization. The languages involved in the study are Indonesian, English, and Putonghua (Mandarin Chinese). The opposition in plosives lies primarily in voicing in Indonesian, in aspiration in Putonghua, and in both aspects in English. The perception and production of Indonesian, English, and Putonghua plosives by Indonesian speakers were examined, who learned English as an L2 and then Putonghua as an L3 (IEP speakers). Twenty IEP speakers participated in the perception, and ten of them participated in production experiments. They were grouped into high-proficiency and low-proficiency groups according to their proficiency in Putonghua.

In the perception study, the categorical perception experiment was conducted where both IEP groups listened to synthesized continua of plosives in Putonghua, Indonesian, and English that varied in terms of Voice Onset Time (VOT) values. The results of the English perception experiment show that both groups of IEP speakers have formed a distinct perceptual category for English voiceless aspirated plosives in terms of high perceptual accuracy and rapid reaction time. In the perception of the Putonghua experiment, the IEP speakers’ results regarding the category boundaries are similar to those of the L1 Putonghua group. However, the low-proficiency group exhibits broader bandwidths and lower discrimination scores, even though their perceptual judgment is even quicker than that of the high-proficiency group. Moreover, despite the high-proficiency IEP group having similar performance to the L1 Putonghua speaker group’s performance, the high-proficiency group’s maximum discrimination accuracy in distinguishing velar plosives is significantly lower, and their reaction times in perceiving these sounds are slower than those of native speakers. This indicates that, despite their considerable proficiency in the language, high-proficiency IEP speakers have not attained a native-like perception of Putonghua plosives.

In the production study, the IEP speakers read the English target words embedded in the middle of carrier sentences. The results demonstrate that IEP speakers’ production of English voiced plosives is similar to their L1 Indonesian voiced plosives in terms of VOT distribution, closure duration, fundamental frequency (F0), and first formant (F1). The VOT distribution of voiceless plosives is longer than that of L1 Indonesian voiceless plosives, indicating that IEP speakers have established a new category for the voiceless aspirated plosives of English. In the Putonghua production experiment, they imitated a native speaker of production Putonghua words contrasting plosives at syllable-initial positions. Their production shows distinctive VOT values for aspiration opposition and longer closure duration for unaspirated plosives. However, the low-proficiency group produced aspirated plosives with a significantly shorter VOT than those of their high-proficiency peers. The low-proficiency group’s production also shows less robust opposition between the aspirated and unaspirated categories, indicating that there is a difference between the two groups of IEP speakers’ production.

The findings from these two studies reveal the characteristics of IEP speakers in terms of L3 speech perception and production. The existence of similar sounds in acquired languages can facilitate the acquisition of L3 sounds. This result is consistent with the Linguistic Proximity Model (LPM), which posits that facilitative influence is more likely when the target sound shares similar phonetic features with those present in the L3 learner’s existing languages. The findings of the perception and production of L3 plosives demonstrate that L3 learners’ perception and production abilities improve as they gain experience in the target language. Moreover, perception may precede production in the initial stage, and production ability may exceed perception ability in the developmental stage. This is consistent with the hypothesis of the revised speech learning model (SLM-r) that L2 segmental production and perception coevolve without precedence.

Date of Award	11 Sept 2024
Original language	English
Awarding Institution	City University of Hong Kong
Supervisor	Bin LI (Supervisor)