[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: NIDCD plan

Dear Neal,

In reply to your question to the list, I can only comment on the work
from my lab.

I think it is becoming clearer that while the type of organization that
the Gestalt psychologists described, mainly in vision, can be observed in
auditory perception too, it can also be observed that top-down processes,
which were neglected by the Gestaltists, play an essential role in the
organization of the auditory signal.  The clearest example is in speech
perception, where familiarity with speech allows a listener to penetrate
the camouflage (bottom-up grouping) that is obscuring a message.  I think
it will be important in the next few years to try to discover the
properties of top-down processes and how they interact with the
Gestalt-like bottom-up processes.

Another important development that I anticipate in the next few years is
the description and classification of consistent individual differences
in the use of various cues for perceptual organization.  Informal
observations in the laboratory indicate that some subjects are more
influenced by one cue and than another, and the relative susceptibility
may vary from subject to subject.  We need to know whether the differences
are systematic. Can a profile be formed for a subject showing the relative
importance, for him or her, of various cues in solving the scene analysis
problem?  Our ability to describe hearing-impaired individuals in this
way may be of value in deciding on the best form of auditory prosthesis
for them.

A third trend that I observe is the accelerating development of
computational models of auditory scene analysis.  There have been three
important meetings of computer scientists and engineers on this topic in
the past 15 months.  My hope is that the modellers will attempt to bring a
wider set of organizational phenomena into their models.  My forthcoming
CD of phenomena of auditory organization may help to sensitize modellers
to the range of phenomena that need to be explained, and to set criteria
which such models should meet.

Best wishes,


Albert S. Bregman,  Professor, Dept of Psychology,  McGill University
1205 Docteur Penfield Avenue,   Montreal,  Quebec,  Canada   H3A 1B1.
Phone: +1 514-398-6103 Fax: -4896 Email: bregman@hebb.psych.mcgill.ca