Some people lack the ability to recognize familiar voices, according to a recent study from Yale and the Max Planck Institute for Human Cognitive and Brain Sciences, a research center in Germany.

In a study of over a thousand people who had expressed concerns that they might have this condition, researchers confirmed two cases of participants with the disorder, called phonagnosia. Scientists have long known of a similar condition called prosopagnosia — characterized by the inability to recognize faces. With many more known cases, prosopagnosia serves as a model for phonagnosia research. Researchers remain unsure about what causes either condition.

“It’s really unclear why this happens, because this condition is so rare,” Samuel Mathias, a psychiatry postdoctoral associate who was involved in the study. “That’s research for the future.”

The researchers reached out to thousands of self-reported potential phonagnosics, and put them through various stages of testing, including tests on voice discrimination, musical ability and pitch perception. The two confirmed phonagnosics were over two standard deviations below average at identifying specific voices, but their speech intelligibility and musical ability were entirely normal, the study found.

The paper is the first to confirm the existence of the condition with comprehensive testing and a solid research design, Mathias said.

Both of the phonagnosics were aware of their condition before the study, he added. One participant realized something was wrong when she became a mother.

“When her daughter was playing with another kid, she couldn’t tell which of those two voices came from her daughter,” Mathias explained.

The other participant was not aware of any problem until a few years ago, when he was watching an English TV show dubbed in German with a friend. When the voice actor for one of the characters changed, the friend was shocked, but the participant did not bat an eye.

According to Mathias, both participants were normal in every other aspect of auditory perception ­— they only failed on the voice recognition tasks. To the researchers, this suggests that voice recognition might be a completely separate process, both cognitively and anatomically, which can be selectively impaired.

“It’s known that the brain’s recognition abilities are part of a very modular system,” he said. “So if [voice recognition] is modular and distinct, then you can knock it out without it affecting anything else.”

The anterior portion of the superior temporal sulcus is responsible for voice recognition in the brain. Yet, according to the researchers, neither case of phonagnosia involved lesions in that brain area, providing no evidence for the physical determinants of the condition.

The visual counterpart to phonagnosia — prosopagnosia — is significantly more common, said Dartmouth psychology professor Brad Duchaine, who was not involved with the study. Some estimates put the prevalence of prosopagnosia as high as one in 50, he added.

According to Duchaine, the causes of the conditions, whatever they may be, are likely similar.

“There’s evidence that identity processing for faces goes on in different regions of the brain,” Duchaine said. “We don’t know all that much about voice processing, but you can imagine there’s a similar sort of division of labor.”

“Phonagnosia” comes from the Latin roots “phone” and “agnosia,” meaning “voice” and “no knowledge,” respectively. The study will be published in Current Biology.