言语的情绪韵律和情绪语义对听觉去信息掩蔽的作用

Unmasking effects of speech emotional prosody and semantics on auditory informational masking

言语的情绪信息(情绪性韵律和情绪性语义)具有去听觉掩蔽的作用,但其去掩蔽的具体机制还不清楚。本研究通过2个实验,采用主观空间分离范式,通过操纵掩蔽声类型的方式,分别探究言语的情绪韵律和情绪语义去信息掩蔽的机制。结果发现,情绪韵律在知觉信息掩蔽或者在知觉、认知双重信息掩蔽下,均具有去掩蔽的作用。情绪语义在知觉信息掩蔽下不具有去掩蔽的作用,但在知觉、认知双重信息掩蔽下具有去掩蔽的作用。这些结果表明,言语的情绪韵律和情绪语义有着不同的去掩蔽机制。情绪韵律能够优先吸引听者更多的注意,可以克服掩蔽声音在知觉上造成的干扰,但对掩蔽声音在内容上的干扰作用很小。言语的情绪语义能够优先获取听者更多的认知加工资源,具有去认知信息掩蔽的作用,但不具有去知觉信息掩蔽的作用。

In daily communication, a speaker’s voice usually carries a particular emotion. Emotional information is transmitted in two ways: prosody and the semantic content of speech. Previous studies have found that emotional prosody has the effect of releasing auditory masking. The purpose of the present study is a), to test whether the emotional semantic content also has the effect of releasing speech from informational masking, and if so, b) to explore what is the difference between the role of emotional prosody and emotional content in releasing informational masking.This study consisted of two experiments, each divided into two sub-experiments. A perceived spatial separation paradigm was adopted in all experiments to separate the effects of informational masking from that of energetic masking. Experiment 1 explored the mechanism of emotional prosody in the unmasking of informational masking. A complete within-subject design of 2(perceived spatial separation: no, have) × 2(emotional prosody: neutral, happy) × 4(signal-to-noise ratio:-8 dB,-4 dB, 0 dB, 4 dB) was adopted in both sub-experiments. Experiment 1a employed time-reversed sentences with no semantic intelligibility as masking sounds(with presumed only perceptual informational masking). Experiment 1b used syntactically correct nonsense sentences as masking sounds(with both perceptual and cognitive informational masking). Experiment 2 also contained two sub-experiments;it aimed to examine the role of the emotional semantics of speech in releasing informational masking. A complete within-subject design of 2(perceived spatial separation: no, have)× 2(emotional semantics: neutral, positive) × 4(signal-to-noise ratio:-8 dB,-4 dB, 0 dB, 4 dB) was adopted in both sub-experiments. Experiment 2a employed time-reversed sentences with no semantic intelligibility as masking sounds. Experiment 2b used syntactically correct nonsense sentences as masking sounds.Experiment 1a showed that the accuracy of recognition of the target sentence uttered in emotional prosody was significantly higher than that of the target sentence uttered in neutral prosody. Experiment 1b showed that the accuracy of recognition of the target sentence uttered in emotional prosody was significantly higher than that of the target sentence uttered in neutral prosody. There was a marginally significant difference between the results of Experiment 1a and Experiment 1b. Experiment 2a showed no significant difference in recognition accuracy between target sentences with emotional semantics and those with neutral semantics. Experiment 2b showed that the recognition accuracy of target sentences with emotional semantics was significantly higher than that of target sentences with neutral semantics. The study found no significant difference between Experiments 2a and 2b.In conclusion, the results of the present study suggest that the mechanisms of emotional prosody and emotional semantics is different in releasing speech from informational masking. Emotional prosody of speech can preferentially attract more attention from listeners and reduce perceptual informational masking, but it only has a minor effect on releasing cognitive informational masking. The emotional semantics of speech can preferentially occupy more cognitive processing resources of listeners. Hence, it can reduce the cognitive informational masking;however, it fails to release the perceptual informational masking.