Many of the people who regularly record these voices say they are spirits -- voices of dead men, women and children who are trying to communicate from beyond the grave. Since spirits no longer have a body with vocal chords, they can't actually "talk." Instead, the theory goes, they use their energy to electronically manipulate sound into a form that resembles the spoken voice.
Voices are rarely heard during recording -- only during playback. They may be so soft that they can barely be heard or so distorted that they must be listened to over and over again to determine their meaning. The words may be in any language, and they can even be in a combination of languages (called polyglot). Sometimes, the voice answers questions or addresses the researcher directly. It may call the person by name or mention something very personal to the researcher. Sometimes, the voice sounds as though it is singing.
Researchers categorize recordings based on their audibility:
- Class A voices are very clear and easily understandable.
- Class B voices are fairly loud and clear and are sometimes audible without headphones.
- Class C voices are very soft and often indecipherable.
Regardless of how clear the recording is, the voices rarely speak for more than a few seconds at a time. Researchers spend hours listening again and again to decipher the meaning behind just a few seconds of sound.