Chances are, you are reading this first sentence and hearing your own voice talking in your head.
According to a new study, internal speech makes use of a system that is mostly employed for processing external speech, which is why we can "hear" our inner voice.
The study comes from the Department of Linguistics at the University of British Columbia, and is led by researcher Mark Scott, who analyzed a brain signal known as "corollary discharge" - a signal that separates sensory experiences we produce ourselves from experiences that are external. This signal helps explain why we are unable to tickle ourselves: it predicts our own movements and omits the sensation of feeling ticklish.
According to the study, this prediction usually filters out self-made sounds so we don't hear them externally, but rather internally. The corollary discharge therefore prevents the sensory confusion that would otherwise arise.
Until now, the phenomenon of internal speech has been mostly unexamined. But through two experiments, Scott saw evidence that corollary discharge is an important component of our internal speech experiences:
"We spend a lot of time speaking and that can swamp our auditory system, making it difficult for us to hear other sounds when we are speaking.
By attenuating the impact our own voice has on our hearing - using the 'corollary discharge' prediction - our hearing can remain sensitive to other sounds."
Our inner voice is actually a prediction
Scott theorized that copies of our internal voices produced by the predictive brain signal can be created even when there is no external sound. In effect, our inner voices are the result of our brain internally predicting our own voice's sound.
If his theory was true, Scott knew that when external sensory information matched the internal copy our brains produced, that external information would be edited out. The results of his experiment confirmed his hypothesis.
When participants said certain syllables just in their heads - such as "al" or "ar" sounds - that matched an external sound, the impact of that external sound was greatly minimized. However, when the internal syllable did not match the external sound, their own perception of both sounds did not diminish.
For his main experiment, Scott used 24 male participants. He chose subjects of the same sex so their voices would match the gender of the voice that produced the external sounds.
A diagram from the paper illustrates the three conditions of Scott's main experiment. After each sound was played twice, an ambiguous target sound "da-ga" was played.
During the experiment, there were three conditions:
- Hearing - participants listened to a sound without engaging in any speech imagery
- Matching - participants thought about the same external sound they were hearing
- Contrasting - participants imagined a different sound than what they were hearing (for example, if they heard "ar," they imagined "al").
The overall results show that inner speech weakens the impact of external sounds when the two are the same.
The findings provide strong evidence that a system involved in processing external speech also works on internal speech, which may help with mental conditions.
Scott notes: "This work is important because this theory of internal speech is closely related to theories of the auditory hallucinations associated with schizophrenia."