Again you could see the auditory effect as an example of the Gestalt principle
of closure. However another way of looking at it may be more profitable.
Richard Warren has interpreted it as resulting from an auditory mechanism
that compensates for masking.^10 He has shown that the illusion can be obtained
only when the interrupting noise would have masked the signal if it had really
been there. The interrupting noise must be loud enough and have the right
frequency components to do so. Putting that in the context of this chapter, we
see that the illusion is another oblique glance of the auditory scene-analysis
process in action.
We have seen how two types of explanation, one deriving from Gestalt psy-
chology and the other derived from considerations of scene analysis, have been
applicable to both the streaming and continuity effects. They differ in style. The
Gestalt explanation sees the principles of grouping as phenomena in them-
selves, a self-sufficient system whose business it is to organize things. The
scene-analysis approach relates the process more to the environment, or, more
particularly, to the problem that the environment poses to the perceiver as he
or she (or it) tries to build descriptions of environmental situations.
Sequential versus Spectral Organization
Perceptual Decomposition of Complex Sounds
We have looked at two laboratory phenomena in audition that show the activ-
ity of the scene-analysis process: the streaming effect and the illusory continu-
ation of one sound behind another. There is a third phenomenon that deserves
to be mentioned in this introductory chapter. It is introduced here not to dem-
onstrate a parallel between vision and audition, but to show another dimension
of the grouping problem. This is the perceptual decomposition of simultaneous
sounds. It can be illustrated through an experiment by Bregman and Pinker.^11
The sounds used in this experiment are shown in figure 9.16. They consist of
arepeatingcycleformedbyapuretoneAalternatingwithacomplextonethat
has two pure-tone components, B and C. This is inherently an ambiguous
event. For example, it could be created by giving an audio oscillator to each of
two people. The oscillator given to one of them puts out the pure tone A, while
the one given to the other puts out the complex tone BC. The two persons are
Figure 9.16
Stimulus used by Bregman and Pinker (1978). A, B, and C are pure tone components.
The Auditory Scene 235