When subjects hear this degraded input, however, they do not detect the omission; instead, they report hearing the complete word legislature, accompanied by a burst of noise.
In effect, subjects supply the missing sound on their own. If subjects are asked exactly when the “accompanying” noise occurred (simultaneous with the first syllable? the second?), they often cannot tell (Repp, 1992; Samuel, 1987, 1991; for further discussion of the recognition of speech see another future article). Similar effects have been documented with musical stimuli. In this case, simple melodies or scales are recorded and then single notes are replaced by bursts of noise. Once again. subjects report hearing the notes that are not there and once again. subjects are inaccurate in judging when within the sequence the noise appeared (DeWitt & Samuel, 1990).
These restoration errors are plainly related to the errors we have described in print recognition–for example proof-reading errors. In both cases, the perceiver goes beyond the information actually provided, filling in the information that should be there. In the case of print, we attributed this to well-primed detectors Thanks to priming, these detectors will fire even in response to a weak input; that makes the network robust. Unfortunately, though, we can find circumstances in which the detectors fire in the absence of an input- producing what we called a false alarm response. Apparently, the same is true with auditory stimuli such as speech, or even with non-linguistic stimuli, such as musical phrases.
Did you like this article? You can write articles like this and make money from it. It is free to join and you can make money online as soon as you sign-up. Click on the link to Sign-up with Bukisa.com and starting making some good money on the internet.
More Content on the Web by Spill Guy: