Summary: A groundbreaking discovery by researchers using an artificial neural network model suggests that the human brain may naturally exhibit a musical instinct. Through an analysis of various natural sounds via Google’s AudioSet, the study revealed that specific neurons in the network responded selectively to music, mirroring the behavior of the auditory cortex in real brains.
This spontaneous development of music-selective neurons indicates that our capacity to comprehend music might be an intrinsic cognitive function, evolved to better process natural sounds.
Key Points:
- The research employed an artificial neural network to illustrate that music-selective neurons can emerge spontaneously without direct exposure to music education.
- These neurons exhibited behaviors akin to those found in the human auditory cortex, responding selectively to diverse music genres.
- The study suggests that musical aptitude could be an instinctive brain function, evolved to enhance the processing of natural sounds.
Source: KAIST
Musical Instinct and Human Brain Development
Is “musical instinct” a shared trait across diverse cultures despite significant environmental variations? On January 16, a team at KAIST, led by Professor Hawoong Jung from the Department of Physics, unveiled the mechanism through which musical instincts arise in the human brain without explicit training, using an artificial neural network model.
Previously, researchers have explored the similarities and distinctions in music across various cultures to understand its universality. A study published in Science in 2019 highlighted the presence of music in all culturally distinct societies, emphasizing commonalities in beats and tunes. It is known that the auditory cortex in the human brain plays a crucial role in processing musical information.
Professor Jung’s team demonstrated that cognitive functions related to music can spontaneously develop as a result of processing natural auditory stimuli, even without formal music instruction.
By leveraging Google’s AudioSet, a vast repository of sound data, the team trained the artificial neural network to recognize different sounds. Intriguingly, certain neurons within the network exhibited selective responses to music, indicating the spontaneous emergence of music-selective neurons.
These neurons displayed minimal reactions to non-musical sounds like animal noises or mechanical sounds but exhibited heightened responses to various music genres, both instrumental and vocal.
The behavior of these artificial neurons closely resembled that of neurons in the human auditory cortex. Notably, the music-selective neurons encoded the temporal structure of music, transcending specific music genres and encompassing a wide range of styles.
Moreover, inhibiting the activity of music-selective neurons significantly impaired the cognitive processing of other natural sounds, suggesting that the neural mechanisms involved in processing music also aid in processing other environmental sounds. This implies that “musical ability” may have evolved as an adaptive trait to enhance the processing of natural sounds.
Professor Jung highlighted the evolutionary aspect of the study, indicating that the universal foundation for processing musical information across cultures might have evolved under evolutionary pressures.
The research, led by Dr. Gwangsu Kim and Dr. Dong-Kyum Kim, was published in Nature Communications under the title “Spontaneous emergence of rudimentary music detectors in deep neural networks.”
Funding: This research received support from the National Research Foundation of Korea.
About the Study
Author: Yoonju Hong
Source: KAIST
Contact: Yoonju Hong – KAIST
Image: Image credited to Neuroscience News
Original Research: Open access.
“Spontaneous emergence of rudimentary music detectors in deep neural networks” by Hawoong Jeong et al. Nature Communications
Abstract: The study uncovers how music-selective units can spontaneously emerge in deep neural networks through natural sound detection, shedding light on the origins and functions of innate musical characteristics in humans.
