The bat's unique sonar spurs research on human sense perception.

''Chirp . . . chirrp . . . chirrrp.'' James Simmons stands at the front of the room, imitating the sounds made by a bat on a nocturnal hunt.

''I know this seems a little silly, but it is much easier than carrying a tape recorder,'' the professor of biology at the University of Oregon explains unselfconsciously.

Dr. Simmons is one of the nation's leading authorities on the unique sonar system which the bat family (Chiroptera) employs so successfully for nocturnal flying and hunting.

His research is not only shedding light on how this elusive bat performs its amazing feats of echolocation, but it has also led to the invention of electronic devices that may help deaf children learn to speak clearly, improve ultrasound imaging used for medical purposes, and even have military application.

Surprisingly, bats and human beings have a lot in common. Both are thought by biologists to have evolved from shrewlike tree-dwelling ancestors. And bats are the only animal besides human beings that are found in every part of the world except the polar regions. ''Bats are more like us than are rats, dogs, cats, or whales,'' this expert says.

But that isn't the reason why Simmons has been studying these animals that many find repulsive. Rather, the ''bat's sonar system provides a good way to study how the brain handles and represents information,'' he says.

The physics of sonar is well understood. Also, bat sounds represent the phonemes in a simple language whose purpose is known. This makes it a fruitful field for neurological research, the scientist maintains.

Research has shown that bats utter two basic types of sonar signals. Both are well above frequencies within the normal human hearing range. One is a chirping sound that starts at a high frequency and ends on a lower note. Such frequency-modulated signals allow the bat to estimate the distance of a target by the time it takes an echoto return. The second type is a constant frequency tone. From the frequency shift of the echo, a bat can tell how fast an object is moving.

Laboratory experiments have shown that a bat's sonar is very sensitive. It can detect a 3/4-inch sphere at a distance of 5 meters - more than 16 feet. Also , it has been demonstrated that the creature can pick up the fluttering of an insect's wings and can discriminate between a number of objects thrown into the air simultaneously.

All this, says Simmons, ''is proof that the bat has a type of sonar system that we would like to know more about because of its real-time processing capability.''

Detailed analysis of the bat's response to various signals has demonstrated that the bat doesn't look at its sound images the way scientists have. Experts have tended to analyze sound impulses by breaking them down into a frequency spectrum, in part because they assumed that was what the ear - the cochlea, in particular - does.

The bat, however, appears to respond to the time structure of the sound waves it receives (mathematically known as the signal's autocorrelation function), Simmons's research has revealed. This is a much simpler signal to process and helps explain how its brain can rapidly interpret complex sonic signals.

Using this insight, the Oregon researcher and his colleagues have constructed an inexpensive real-time spectrum analyzer out of some computer chips from Radio Shack and a cheap oscilloscope. ''We've got a toy spectrum analyzer which will outperform a million-dollar array processor on a submarine,'' Simmons asserts.

Besides the obvious military applications, this could prove useful for training deaf children to speak, he suggests. Converting the sounds they make into a visual form is a proven technique for teaching the deaf to talk, but conventional spectrum analyzers are either very expensive or very slow.

Another potential use for insights culled from the bat's sonic abilities is to improve the ultrasound imaging systems being used for medical purposes. Companies that currently make ultrasound scanners have made no attempt to produce sound pulses that result in the best images, the scientist says. They simply use whatever frequency the generator produces.

Further in the future, neuroscience research of this sort will be essential to unravel the mysteries of human intelligence, Dr. Simmons is convinced. And this, in turn, is necessary before truly intelligent machines can be constructed: ''Where will the insights come from to build intelligence into computers?'' he asks. ''This is a very complicated subject. We don't even yet know what we are really talking about.'' But the mysterious bat may hold some of the crucial clues.