More adaptation of speech by nonspeech

Area of Science:

• Auditory Perception
• Psychoacoustics
• Speech Science

Background:

• The Samuel and Newport (1979) hypothesis posits that auditory input is sorted by spectral quality: periodic versus aperiodic.
• Understanding this fundamental sorting mechanism is crucial for deciphering auditory processing in both speech and non-speech domains.

Purpose of the Study:

• To empirically test Samuel and Newport's hypothesis regarding spectral quality sorting in auditory perception.
• To investigate whether the perceptual system differentiates input based on periodicity, irrespective of its linguistic relevance.

Main Methods:

• Three experiments utilized adaptation paradigms and similarity rating tasks.
• Stimuli included speech sounds (/ja-za/ continuum) and non-speech sounds (white noise, periodic tones).
• Adaptation involved repeated exposure to periodic or aperiodic stimuli to assess perceptual shifts.

Main Results:

• Aperiodic white noise induced significant labeling shifts on the /ja-za/ continuum, while periodic stimuli did not.
• Non-speech aperiodic adaptors demonstrated substantial adaptation effects, comparable to speech endpoints.
• Aperiodic components were found to dominate adaptation effects for the /ja-za/ continuum.
• Similarity ratings confirmed that /za/ is perceptually grouped with unvoiced fricatives.

Conclusions:

• The perceptual system demonstrably sorts auditory input based on spectral quality (periodic vs. aperiodic).
• This sorting mechanism operates independently of whether the input is speech or non-speech.
• Findings support the hypothesis that spectral quality is a primary organizing principle in auditory perception.