The term phonagnosia refers to an inablity to recognize familiar voices or to discriminate between unfamiliar ones. This is a rare condition that is usually associated with brain damage: the ability to recognize familiar voices is impaired by damage to several regions of the right parietal lobe, and impaired voice discrimination is associated with damage to the temporal lobe in both hemispheres.
Researchers from UCL now report the first known case of developmental phonagnosia. In the journal Neuropsychologia, they document the case of a 60-year-old woman known as K.H., who says that she has been unable to recognize familiar voices for as long as she can remember.
K.H., an active and professional woman who works as a management consultant, told the researchers that she has always had severe difficulties recognizing familiar voices, even those of close relatives such as her daughter. Consequently, she avoids answering the telephone, and for many years has only answered "booked calls": friends and co-workers pre-arrange their calls, so that she knows who to expect when she answers the phone at a certain time. K.H. also reported that in her job during the 1980s, she introduced herself on the telephone using a different form of her first name, so that she would know work-related calls when she answered them.
After spending much of her life struggling with this inability, K.H. finally began to realize the nature of her problem after reading an article in a popular science magazine. The article was about prosopagnosia, a neurological condition characterized by the inability to recognize faces. Although she had never experienced difficulties recognizing faces, K.H. realized that she might have a vocal equivalent of prosopagnosia. It was then that she approached the researchers at UCL's Institute of Cognitive Neuroscience.
The first thing that the researchers did was to investigate K.H.'s impairment under laboratory conditions. In one task, she was presented with a series of 7-second samples of famous peoples' voices, taken from television interviews, and her performance compared to that of 8 age-matched controls. The controls recognized an average of 17 of the 96 voice samples, whereas K.H. correctly identified only one (that of Sean Connery). Likewise, her ability to learn the voices of 6 unfamiliar young female speakers and to pair them up with names was very poor, as was her ability to determine if pairs of unfamiliar voice samples were spoken by the same person. However, neuroimaging revealed no structural abnormalities in her brain.
One hypothesis of how we recognize a person's identity from their voice states that speech perception and the recognition of emotion in a voice involve distinct mechanisms. Previous studies have shown that listeners can extract information about age, sex, pronunciation and even body size from a voice, all of which are important aspects of the speaker's identity. It is generally believed, however, that recognition involves more than the perception of these characteristics; it is still unclear exactly which properties of a voice are required for recognition, and it may be the case that different voices are recognized by different characteristics, or combinations of them.
The case of K.H. sheds some light on the cognitive processes underlying voice recognition and auditory perception in general. Although K.H. could not recognize famous voices or discriminate familiar ones in the laboratory tests, she was able to perceive vocal expressions of emotions. Despite her voice recognition deficits, K.H. was still able to determine the sex of the speakers in the voice samples presented to her; furthermore, she performed as well as the controls on music perception tasks, and reported that she recognizes familiar songs (interestingly though, she usually cannot recognize specific singers). These findings therefore support the model of separate mechanisms for the perception of speech and emotion, and suggest that speech recognition and musical perception also involve distinct mechanisms, at least in part.
This study adds to a growing number of cases which suggest that specific cognitive skills can fail to develop properly while leaving others unimpaired. Last month, for example, researchers from the University of British Columbia reported the first known case of developmental topographical disorientation; this patient's ability to form cognitive maps was severely impaired, but various other cognitive skills remained unaffected. Similarly, K.H.'s memory for voices is severly impaired, but this deficit does not extend to other classes of auditory stimuli (such as music or environmental sounds) or to other ways of identifying people (her ability to recognize faces was unimpaired). Further work may enable the researchers to better understand the exact nature of K.H.'s voice recognition deficit.
Garrido, L. et al (2008). Developmental phonagnosia: A selective deficit of vocal identity recognition. Neuropsychologia DOI: 10.1016/j.neuropsychologia.2008.08.003
- Log in to post comments
Vilayanur Ramachandran delivered a talk on this at TED a while back. I was riveted. I was also delighted, because when he touched on synaesthesia, I was finally able to identify why I taste in colours!
That's quite interesting, especially that the mechanism for recognizing individual pieces music seems completely separate from the mechanism for recognizing voices. I wonder if within a piece of music K.H. can distinguish between instruments, say, between a viola and an oboe.
I doubt if she can differentiate between a tenor and a bass or a soprano and a contralto. A tenor/baritone or soprano /mezzo ??
I've corresponded with Diana Sidtis, who's been researching this topic for some time, and we agreed briefly that my hearing condition is likely phonagnosia. Until my mid-20s, I thought that identifying yourself when you call on the phone was not good manners - it was necessary to be identified! I had no idea people could recognize voices as distinctive. I thought my own voice was so distinctive that people recognized it, which was why they acted like my self-identification was unnecessary.
Usually I can figure out who's on the phone by context, or if they talk long enough. My wife agreed early in our relationship to tell me who she is; after an embarrassing phone call with my (first) wife, I understand K.H.'s reluctance completely. (Though I just ask who's calling.)
By the way, I enjoy music and play guitar (though I cannot sing on key). I think I've read elsewhere theories that music is processed differently than voices.