Adults with ASD respond similarly to human, artificial singing

Adults with autism spectrum disorders (ASD) do not detect differences in “humanness” between human and artificial singing, a new study from Japan reports. 

Shinji Kuriki and colleagues asked 14 adults with high-functioning ASD and 14 adults without ASD to listen to 24 song segments presented in random order. Half of the song segments were sung by a human voice, while the other half were sung by an artificial voice created by software. 

The researchers report that for people without ASD, the songs sung by a human voice evoked more impressions of “humanness” and more positive feelings than those sung by the artificial voice. “In contrast,” they say, “people with ASD had similar impressions of humanness and positive feelings for the songs sung by the human and artificial voices.” The two groups did not differ in their perception of musical characteristics such as complexity, regularity, and brightness. 

The researchers conclude, “The present study demonstrates that human beings seem to have a unique sense of humanness in auditory processing and that this sense may be impaired in individuals with ASD. This could be a reasonable explanation for some characteristics typical of ASD, such as the low orientation to social stimuli and the tendency to prefer robots to humans.”

Citations

“Similar impressions of humanness for human and artificial singing voices in autism spectrum disorders,” Shinji Kuriki, Yuri Tamura, Miki Igarashi, Nobumasa Kato, and Tamami Nakano, Cognition, Vol. 153, 2016, 1-5. Address: Tamami Nakano, Graduate School of Frontiers Biosciences, Osaka University, 1-3, Yamadaoka, Suita, Osaka 565-0871, Japan, [email protected]