Accommodating Talker Variability in Noise With Context Cues: The Case of Cantonese Tones

PURPOSE: Listeners often rely on context cues to manage talker variability in speech and achieve perceptual constancy, a process known as extrinsic normalization. However, everyday communication typically involves both talker variability and noise, and the interaction between these factors is not well understood. This study examined the effects of different noise types and levels on listeners' ability to use contextual cues for adapting to talker variability and additionally explored the role of attentional control in this process. METHOD: Thirty-seven young native Cantonese speakers participated in a speech perception task to identify Cantonese tones from four different talkers using speech contexts provided in either quiet or noisy environments. The study tested various signal-to-noise ratios (SNRs; 10, 5, 0, -5, and -10 dB) and noise types (babble noise [BN] and babble-modulated speech-shaped noise [BMN]). Attentional control was measured using the Stroop Color-Word Test. RESULTS: Listeners were able to use context cues to adapt to talker variability in Cantonese tones at SNRs of 0 dB and above. The effectiveness of using context cues decreased as the SNR lowered. BN created more difficulty for extrinsic normalization than BMN at -5 and -10 dB SNRs. Notably, listeners with lower Stroop interference scores demonstrated better extrinsic normalization in BMN and at 10 and 0 dB SNRs. CONCLUSIONS: Listeners can effectively use context cues to adapt to talker variability in Cantonese tones under low-to-moderate noise conditions. However, high noise levels significantly hinder this ability. BN presents greater challenges than BMN at lower SNRs, likely due to increased informational masking. Attentional control plays a crucial role in facilitating extrinsic normalization in specific noise conditions. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.29621342.