Re: Instruction for subjects (Pierre Divenyi )

Subject: Re: Instruction for subjects
From:    Pierre Divenyi  <pdivenyi(at)EBIRE.ORG>
Date:    Thu, 27 Oct 2005 12:40:06 -0700

Fatima, Bruno:

The continuity effect for glide perception seems to be auditory and 
perhaps, although not necessarily, peripheral. This was shown 
recently by Poppy Crum (UCBerkeley) in her dissertation work. She was 
measuring detectability of probe tones after a noise burst following 
a pure-tone glide and found that detectability increased when the 
frequency of the probe was right on the glide trajectory cut short by 
the noise. You should send your inquiries directly to her: pcrum2(at)bme.jhu.edu.

Pierre

At 08:17 AM 10/27/2005, Bruno Repp wrote:
>Dear Fatima:
>
>I found your observations interesting, although I am not really an 
>expert on this phenomenon. Nevertheless, the task reminds me of some 
>work on phoneme restoration that I did long ago.
>
>I suspect that the reported continuity of the disrupted/masked sound 
>is not really auditory, or only partially so. Rather, it is a 
>figural completion achieved at a higher level of processing. It may 
>be analogous to the completition of partially hidden objects in 
>vision. If part of an object is obscured by another object, people 
>don't really "see" the obscured part, even though they perceive the 
>obscured object to be intact. Similarly, the disrupted/masked sound 
>may be perceived as a continuous auditory "object", even though 
>people really don't "hear" it during the masked portion.
>
>You are asking your subjects to compare a sound which is heard in 
>its entirety to one that is heard only partially, and that may be 
>the reason for the confusion. The empirical question is what is 
>actually heard during the /disrupted/masked segment: Do people only 
>hear the masking noise, or do they also hear the masked sound? You 
>could try to address that question by asking subjects to compare the 
>disrupted test sounds with standards that have their central segment 
>attenuated by varying degrees. Perhaps there is a certain degree of 
>attenuation that best matches what subjects actually hear.
>
>Best,
>
>Bruno