Manipulation of F2

Formant: The strongest sound of the vowel

F2: The second formant in a vowel that determines how far front or back the sound is

The most classical example of Categorical Perception is when an individual listening to sounds can tell the difference between two speech sounds. The differentiation is only possible when the individidual is aware that the two speech spounds are able to be differentiated as two separate phonetic segments. Many studies trying to test this theory have been created and replicated. They all share the common goal of finding out whether a listener can differentiate between two phonetically different sounds and how great or small the difference can be in order for them to still be able to be differentiated.

In order to discover Categorical Perception, many scientists do studies that have listeners differentiate between sounds that are being manipulated. Examples of such experiments are two experiments performed by the Department of Linguistics at Stockholm University.

Department of Linguistics at Stockholm University Experiment Example 1

For the first experiment, scientists take the sounds “ba”, “da”, and “ga” and manipulate them on a spectrum. They can start by playing the sound “ba” for the listener and the listener will have no question in their mind that they heard the sound “ba”. Scientists will then start manipulating F2 slowly, one level at a time. Until a certain level, the listener will still be relatively sure that they heard the sound “ba” and will continue reporting that it is what they hear. Suddenly, after a certain amount of manipulation of F2, the listener will be sure that they hear the sound “da”. They will continue to hear this sound for a period of time though F2 is still being manipulated. Eventually, the listener will suddenly hear the sound change to “ga”. They will also continue to hear this sound for a period of time even though F2 is being manipulated still through the highest level of manipulation they plan on using. Results of this experiment should show a curve pattern where for a period of time, the curve is in a steady up position, which eventually declines and then steadies again. This pattern is repeated for each of the three sounds.

Department of Linguistics at Stockholm University Experiment Example 2

In the second experiment, the listener is instructed to listen to pairs of sounds and determine whether the sounds were the same or different from one another. Unknowing to the listener, all of the pairs are different. Each sound can range from a large level of manipulation of F2 to only 1 level of manipulation. The sounds with the larger manipulation are more easily differentiated than those with smaller manipulations.

These experiments both show that listeners are only able to detect a certain amount of variation of F2 and if the difference is small enough, they will not be able to perceive it. This is the main basis of what is Categorical Perception.

Typical Experiment Results:

Chinchillas

Chinchillas have also been found to have categorical perception and seem to have a similar perception of VOT as do humans. Scientists have been able to train them to show a different response to the sounds /d/ and /t/ that were being spoken by a human, which proves that they can differentiate and have some sort of Categorical Perception, even though they do not have the ability to produce formal language.

References

Categorical perception. (n.d.). Retrieved from http://www.ling.gu.se/~anders/education/CPstim.html

Categorical perception. (n.d.). Retrieved from http://www.ling.gu.se/~anders/KatPer/Applet/index.eng.html

Ladefoged, P. (2005). Vowels and consonants. Blackwell, MA: Blackwell Publishing.