PURPOSE: The present study examined cochlear implant (CI) users' perception of vowels presented concurrently (i.e., double vowels) to further our understanding of auditory grouping in electric hearing. METHOD: Identification of double vowels and single vowels was measured with 10 CI subjects. Fundamental frequencies (F0s) of vowels were either 100 + 100 Hz or 100 + 300 Hz. Vowels were presented either synchronously or with a time delay. In "Double" sessions, subjects were given only double vowels. In "Double + Single" sessions, while double and single vowels were presented, subjects reported the number and identity of the vowel(s). In addition to clinical settings, stimuli were delivered via an experimental method that interleaved pulse streams of two vowels. RESULTS: Although the time delay between vowels had a large effect on identification, the effect of change in fundamental frequency (ΔF0) was modest. Enumeration was poor in general, and identification of synchronous vowels was above chance in only the Double sessions with a priori knowledge about presentation. Interleaved presentation of vowel streams provided no benefit for identification and a marginal benefit for enumeration. CONCLUSIONS: The results demonstrate the importance of episodic context for CI users. Unreliable perception of multiplicity observed in the present results suggests that auditory grouping in CIs may be driven by a schema-based process.
PURPOSE: The present study examined cochlear implant (CI) users' perception of vowels presented concurrently (i.e., double vowels) to further our understanding of auditory grouping in electric hearing. METHOD: Identification of double vowels and single vowels was measured with 10 CI subjects. Fundamental frequencies (F0s) of vowels were either 100 + 100 Hz or 100 + 300 Hz. Vowels were presented either synchronously or with a time delay. In "Double" sessions, subjects were given only double vowels. In "Double + Single" sessions, while double and single vowels were presented, subjects reported the number and identity of the vowel(s). In addition to clinical settings, stimuli were delivered via an experimental method that interleaved pulse streams of two vowels. RESULTS: Although the time delay between vowels had a large effect on identification, the effect of change in fundamental frequency (ΔF0) was modest. Enumeration was poor in general, and identification of synchronous vowels was above chance in only the Double sessions with a priori knowledge about presentation. Interleaved presentation of vowel streams provided no benefit for identification and a marginal benefit for enumeration. CONCLUSIONS: The results demonstrate the importance of episodic context for CI users. Unreliable perception of multiplicity observed in the present results suggests that auditory grouping in CIs may be driven by a schema-based process.