PERCEPTION OF SPEECH SOUNDS

语音感知

• 批准号: 6116313
• 负责人: PATRICIA K KUHL
• 金额: $ 8.67万
• 依托单位: UNIVERSITY OF WASHINGTON
• 依托单位国家: 美国
• 财政年份: 1999
• 资助国家: 美国
• 起止时间: 1999-05-01 至 2000-04-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6116313
• 关键词: Mammalia; Primates; communication behavior; psychology

Duplex perception is a laboratory phenomenon in which one stimulus, or one stimulus component, simultaneously contributes to two distinct percepts in human listeners. The primary demonstrations of this phenomenon involve altering the quality of a critical component of a speech syllable (/da/ or /ga/) so that the component is heard out as a separate, nonspeech event (a chirp) while also contributing to the perception of the syllable. The important question is whether this demonstration reflects a specialized mode for processing speech information, or a generalized perceptual tendency in the auditory system for attributing stimuli to a particular sound source. Many speech perception phenomena can be replicated in nonhuman primates. The perceptual abilities that monkeys demonstrate cannot be attributed to uniquely human speech processing mechanisms and may be due to more general auditory mechanisms. Conversely, perceptual tasks on which human listeners succeed and monkeys fail are good potential candidates for special-processing mechanisms in humans. In 1997, we conducted Duplex Perception tests with nonhuman primates. These experiments used synthetic /da/ and /ga/ syllables as stimuli. Tests of Duplex Perception are valuable for theory construction in that they provide a critical test of the two opposing models of the initial status of the mechanisms underlying speech processing in humans.

双重感知是一种实验室现象，其中一个刺激或一个刺激成分同时有助于人类听者的两种不同感知。 这种现象的主要表现是改变语音音节中一个关键成分（/da/或/ga/）的质量，使该成分作为一个单独的非语音事件（啁啾）被听到，同时也有助于音节的感知。 重要的问题是，这个演示是否反映了处理语音信息的专门模式，或者是听觉系统中将刺激归因于特定声源的普遍感知倾向。 许多言语感知现象可以在非人灵长类动物中复制。 猴子表现出的感知能力不能归因于人类独特的语音处理机制，而可能是由于更普遍的听觉机制。 相反，人类听者成功而猴子失败的感知任务是人类特殊处理机制的良好潜在候选者。 1997年，我们对非人类灵长类动物进行了双重感知测试。 这些实验使用合成的/da/和/ga/音节作为刺激。 双重知觉的测试对于理论构建是有价值的，因为它们提供了对人类语音处理机制初始状态的两个对立模型的关键测试。