Brain uses internal 'average voice' prototype to identify who is talking

(Medical Xpress)—The human brain is able to identify individuals' voices by comparing them against an internal 'average voice' prototype, according to neuroscientists.

A study carried out by researchers at the University of Glasgow and reported in the journal Current Biology demonstrates that voice identity is coded in the brain by reference to two internal voice prototypes – one male, one female.

that have the greatest difference from the prototype are perceived as more distinctive and produce greater than voices deemed very similar.

The researchers in the Institute of Neuroscience & Psychology conducted the study by generating a voice prototype through morphing 32 same-gender voices together resulting in a smooth, idealised voice with few irregularities.

They then generated different voices by altering the 'distance-to-mean' of the prototype voice – for example, changing the tone and pitch or morphing two or more voices together.

Using functional Magnetic Resonance Imaging (fMRI), the researchers were able to see increased neural activity the further from the prototype the voices were.

This video is not supported by your browser at this time.

This video is not supported by your browser at this time.

Professor Pascal Belin said: "Like faces, voices can be used to identify a person, yet the neural basis of this ability remains poorly understood. Here we provide the first evidence of a norm-based coding mechanism the brain uses to identify a speaker.

"The research indicates this is a similar process for the identification of faces, where the brain also uses an average face to compare against other faces it encounters in order to establish identity.

"So, rather than having to remember each single voice it hears every day for a lifetime, the facilitates the task of identification by remembering only the differences from the prototype it stores.

"It leads to a range of interesting and important questions, such as whether the prototypes are innate, stored templates or whether they are subject to environmental and cultural influences. Could the prototype consist of an average of all voices experiences during one's life?"

The study 'Norm-based Coding of Voice Identity in Human Auditory Cortex' was supported by the Biotechnology and Biological Sciences Research Council (BBSRC).

More information: Latinus et al.,'Norm-based Coding of Voice Identity in Human Auditory Cortex', Current Biology (2013) dx.doi.org/10.1016/j.cub.2013.04.055

Related Stories

Body size conveyed by voice determines vocal attractiveness

Apr 24, 2013

Deep male voices and high-pitched female voices are perceived as more attractive because listeners gauge the speaker's body size from the frequency of their voice, according to research published April 24 in the open access ...

How do we combine faces and voices?

Mar 09, 2011

Human social interactions are shaped by our ability to recognise people. Faces and voices are known to be some of the key features that enable us to identify individual people, and they are rich in information such as gender, ...

Recommended for you

Know the brain, and its axons, by the clothes they wear

Apr 18, 2014

(Medical Xpress)—It is widely know that the grey matter of the brain is grey because it is dense with cell bodies and capillaries. The white matter is almost entirely composed of lipid-based myelin, but ...

Turning off depression in the brain

Apr 17, 2014

Scientists have traced vulnerability to depression-like behaviors in mice to out-of-balance electrical activity inside neurons of the brain's reward circuit and experimentally reversed it – but there's ...

Rapid whole-brain imaging with single cell resolution

Apr 17, 2014

A major challenge of systems biology is understanding how phenomena at the cellular scale correlate with activity at the organism level. A concerted effort has been made especially in the brain, as scientists are aiming to ...

User comments

Adjust slider to filter visible comments by rank

Display comments: newest first

beleg
1 / 5 (1) May 24, 2013
A gender specific auditory prototype does not exist.
A gender specific musical octave does not exist.

Innate in the human auditory cortex is what is labeled musically 'the octave' - arising from auditory tonotopy. The human auditory tonotopy is genetically 'innate'.

In other words the tonotopy is the prototype you seek.

The answer to your question:
"Could the prototype consist of an average of all voices experiences during one's life?"
is no.
gwrede
3 / 5 (2) May 24, 2013
"It leads to a range of interesting and important questions, such as whether the prototypes are innate, stored templates or whether they are subject to environmental and cultural influences. Could the prototype consist of an average of all voices experiences during one's life?"

The concept of storing an average and for the individuals only the difference is pervasive, and not limited to recognizing humans. This way much less information has to be stored for each target.

And the "average" is not a lifetime thing, it's basically "what you're used to, for now".

For example, hang around with a basketball team for a day and then see a friend, the friend will seem short. Talk with fashion models all day and your girlfriend will look ugly.

A white witness saw "a black man". But after a year in Africa you wouldn't see the blackness, and you could describe him very clearly, simply because you now have the "normal" against which to store your memory of him.
beleg
1 / 5 (1) May 24, 2013
That is conditional bias.
Auditory tonotopy is inherent, innate, and a genetically unconditional bias.
A unchanging bias is a standard reference to establish an average for a prototype needed in auditory processes.

Changing the brain specific estrogen levels will change the outcomes and results these researchers established for a voice averaged prototype. This establishes that a voice average prototype is not an experience-based prototype. The prototype is biochemical in nature, tonotopological in shape and lifelong.