and the low one as
L2–L1–L2–L1–L2–L1–....Again each stream is composed of two alternating tones. In fact, if the infant
lost track of which one of the pair of tones started the sequence, the two streams
wouldbeconsideredtobeexactlythesameastheywerewiththeoriginalorder
of tones. Suppose, however, that the infant does not segregate the high from
the low tones. In this case the forward and the backward orders of tones are
quite different from one another and remain so even if the infant forgets which
tone started the sequence.
To summarize, the segregated streams are quite similar for the forward and
backward sequences whereas the unsegregated sequences are quite different.
Using the habituation/dishabituation method, Demany tried to determine
whether the infants considered the forward and backward sequences the same
or different. The results showed that they were reacted to as being the same.
This implied that stream segregation had occurred. In addition, Demany showed
that this result was not due to the fact that the infants were incapable in general
of distinguishing the order of tonal sequences. Pairs of sequences whose se-
gregated substreams did not sound similar to an adult were not reacted to
as being the same by infants. In general, the infant results paralleled those
of adult perception and the older and younger infants did not differ in their
reactions.
Undoubtedly more such research is required. After all, the infants were not
newborns; they had had some weeks of exposure to the world of sound. But
after this pioneering study, the burden of proof shifts to those who would ar-
gue that the basic patterns of auditory organization are learned. Unfortunately,
working with very young infants is difficult and the amount of data collected
per experiment is small.
The unlearned constraints on organization can clearly not be the only ones.
We know that a trained musician, for example, can hear the component sounds
in a mixture that is impenetrable to the rest of us. I have also noticed that when
researchers in my laboratory prepare studies on perceptual organization, they
must listen to their own stimuli repeatedly. Gradually their intuitions about
how easy it is to hear the stimulus in a particular way comes to be less and less
like the performance of the untrained listeners who are to serve as the subjects
of the experiment.
Undoubtedly there are learned rules that affect the perceptual organization
ofsound.Ishallrefertotheeffectsoftheserulesas‘‘schema-basedintegration’’
(a schema is a mental representation of some regularity in our experience).
Schema-based analysis probably involves the learned control of attention and is
very powerful indeed. The learning is based on the encounter of individuals
with certain lawful patterns of their environments, speech and music being but
two examples. Since different environments contain different languages, musics,
speakers, animals, and so on, the schema-based stream segregation skills of
different individuals will come to have strong differences, although they may
have certain things in common. In later chapters, I will give some examples
of the effects of schema-governed scene analysis in the fields of music and
language, and will discuss a theory of sequential integration of sound, pro-
The Auditory Scene 245