We become conscious of the environment through our senses —
and the principal human senses are vision, touch and sound. Sight reveals
colour and great spatial detail while touch makes us aware of heat and texture.
Hearing is both an important medium of perception and in many species, the
chief means of communication.

James A Traer and Josh H McDermott of the Department of
brain and cognitive sciences, Massachusetts Institute of Technology, in their
paper published in the journal, Proceedings of the National Academy of Sciences
of the USA, point out that the signal the ears receive is a mixture of
information from many sources and there is considerable distortion and
duplication of sound waves before they are heard. The kind of distortion,
however, depends on the geometry of the space around us, they say, and their
research is about the mechanics of how biological systems are able to segregate
sources and assess surroundings.

The principal difference in the way the eyes and ears
function is that the eyes need to move and face the source of light but the
ears react equally to sound that comes from all directions. The reason, of
course, is that light, for practical purposes, travels in straight lines while
sound waves, whose wavelength is in metres, passes freely around corners. And
then, sound is also efficiently reflected by most surfaces. The result,
however, is that the eyes are able to focus on specific objects and create a
detailed map of the field of view, which is possible only to some extent for
the ears.

Apart from the eyes being provided with lenses that focus
objects clearly on the retina, which aids estimate of distance, the fact that
we have a pair of eyes enables us to see “in depth” or to make out what objects
are near and which ones are further away. An auditory parallel of such
“stereoscopic” vision is “stereophonic” sound, where the fact that we have two
ears helps us make out where different sounds originate from, so long as we hear
the sounds directly, and not second hand, by echoes or via loudspeakers. In the
case of depth vision, the effect is thanks to the slightly different views that
each eye sees, because they are placed a few inches apart, and the brain
learning to make use of the perspective.

But in the case of sound, the separate ears receive, from
each source, sound waves that are either a little louder or softer, or a little
out of phase, which is to say, the stage of wave motion, depending on the
distance from the ears. When the sounds are equally loud or in phase, of
course, we sense the sound as coming from directly before or behind us. Even
such limited judgment, however, is often obscured by the ears receiving sound
that has been reflected off other objects or surfaces, which obscures the
source and causes “reverberation”.

Reverberation is an effect that can even result in speech
being unintelligible because the echo of each word from the walls and roof a
large hall runs into the sound of the following word.

The study of the MIT duo was regarding the characteristics
of the ever-present reverberation of sounds in places — the ones people usually
inhabit and what part of this people can actually discern — which helps them
isolate the sources. Based on the experience of reverberation in known places,
the study reveals that the brain is able to separate sound into contributions
from the source and the environment. That helps recognition of sounds and also
provides information about the surroundings.

In the case of bat or dolphins, the use of sound is much
more effective and the animals’ ears are the chief organ of navigation. The
difference in these cases is that what the animals listen to are not sounds
created by objects in the surroundings, but the distinct echo of a high pitched
“click” that the animal itself generates.  
There would, of course, be an element of secondary echoes in the return
sound that is heard, and this may contribute to enhancing the information
received, in the manner that the MIT researchers have discovered in the case of
human subjects of their study.

The first step in the study conducted was to see if there
were statistical regularities in the reverberation space of the normal aural
environment of people. If there was such a “normal” pattern, this may explain
the observation that the brain is able to filter out the distortion caused by
reverberation and purify the meaningful signal. The study therefore first
identified places, which could be objectively considered a “normal” environment
by tracking a group of volunteers with the help of random text messages 24
times a day for two weeks. The volunteers were required to respond by stating
the place where they were when they received the message, and this generated a
starting list of 301 locations, in the Boston metropolitan area — in 271 of
them it was possible to conduct further study.

Trials at each of the 271 places resulted in data of the
nature of sound energy received by a listener at these places, particularly the
timing and amplitude of the echoes and the falling off of the loudness, as the
sound died out. The result of the trials, the study says, was to find that
there was a common feature of the way the later part of the sound decayed. This
finding — that the reverberation that arises when sounds are generated in
places which people frequent — has common characteristics, leads to the
possibility that the brain, with experience, learns these patterns and can then
devise a way to filter the distortion out.

The next part of the experiments was to play these sounds,
and also synthetic sounds that had the same characteristics or had been
modified, to human listeners, to see if they could make out the changes in the
nature of reverberation. The result, the study says, was that the subjects were
able to consistently identify the cases where the patterns were different from
the normal. Next, the experimenters tested whether listeners could identify the
part of the sound that came from the original source. Here again, the subjects
gave statistically significant correct answers when the source sound was
accompanied by a “natural” synthetic reverberation but not when the
reverberation added was different from normal.

The study thus shows that the auditory system has a working
method of using past experience to filter out the distortion caused by
reflections from surroundings, which the listener is familiar with. This helps
to both identify the source of the sound and also gives her an idea of “normal”
surroundings, when the nature of distortion by echoes is like what is normally
experienced. The ability, to take experience into account when data is
distorted or insufficient, parallels what happens in the visual field too.

By S Ananthanarayanan

(The writer can be contacted at [email protected])