Richard R. Fay- rfay@luc.edu, 773-508-2714
Director, Parmly Hearing Institute
Loyola University of Chicago
6525 N. Sheridan Rd.
Chicago, IL 60626
Popular version of paper 2aAB2
Presented Tuesday morning, December 4, 2001
142nd ASA Meeting, Fort Lauderdale, FL
The sense of hearing by goldfish has been investigated in detail using behavioral methods based on Pavlovian conditioning. We have been using a type of experiment known as "stimulus generalization." In these experiments, an animal is conditioned, pairing a long-duration complex or simple sound with a mild electric shock that normally causes respiratory mouth movements to slow down briefly. After 10-20 conditioning trials, the animal begins to slow down respiration when the sound comes on, indicating that the animal heard the sound. After 40 of these trials, the animal is presented with novel sounds, and the degree of respiratory slowdown is measured. This is an estimate of the degree to which the novel sounds are similar in the animal's perception to the original, conditioning sound. Using this method, we have determined that goldfish seem to perceive simple and complex sounds similar to the way that human beings and other vertebrate animals that have been tested (primarily, several mammals and birds).
In one experiment, goldfish were conditioned to respond to a complex sound made up of a "click" sound repeated regularly for six seconds. To human listeners, this sound has a timbre determined by the frequency components making up the click, and a pitch or roughness determined by the rate at which the click is repeated. This is analogous to a human listener identifying the musical instrument playing a note (source identification based on timbre), and also recognizing the note value played (pitch). We can identify the source (instrument) regardless of the note played, and a specific sequence of notes can display a melody regardless of the instrument (source) playing. Similar experiments on goldfish reveal that they behave similarly. After conditioning to a particular source and pitch, changes in click repetition rate (pitch) lead to reduced responses, suggesting that the animal is aware of the changes in pitch value. Changes in the source characteristics (its timbre, to a human listener) also cause a reduction of response, suggesting the animal is aware of the timbre changes. Changing both pitch and source values results in a response reduction proportional to both, nearly independently.
These and other experiments of this type have led us to conclude that goldfish
know the sorts of things that we know about complex sounds. Further, we have
concluded that these capacities for perceiving sound qualities exist among all
vertebrate animals, and could thus be termed primitive or shared. Following
from this, we have concluded that these capabilities of human hearing are also
primitive, and possibly inherited from the earliest vertebrate animals.