杏吧原创

We have misunderstood how the brain processes language, says a controversial new analysis. The brain regions that have been identified as specialised language processing centres are actually hubs that coordinate the processing of language across multiple brain regions, argue at University College London and his colleagues.

Their idea runs counter to most models of language in the brain and has been met with scepticism.

Since the late 1800s, there has聽been growing evidence that聽specific brain areas process language. For example, Broca鈥檚 area in the frontal cortex is thought to be involved in speech. People with damage to this area understand what is said to them, but can struggle to speak. Brain imaging studies consistently show activity in this and other areas when performing language tasks.

Skipper doesn鈥檛 dispute that聽regions like Broca鈥檚 area are involved in language processing, but he thinks much of the rest of聽the brain also plays a role. 鈥淭here鈥檚 a ton of processes that go聽into language,鈥� he says. 鈥淚t鈥檚 the聽most complicated thing we聽do聽as human beings.鈥�

To demonstrate this, he and his colleagues conducted a two-part study. In the first part, they reanalysed data from existing meta-analyses of previous studies of brain imaging data. They confirmed that research looking at language processing showed activation was localised to regions like Broca鈥檚 area.

However, when they looked at more specific forms of language processing, for example of nouns versus verbs, they saw many more brain regions coming online.

Free newsletter

Sign up to The Daily

Sign up to newsletter

Skipper argues that this is because different language tasks use different parts of the brain, but聽brain imaging studies work by聽taking averages across multiple tasks. As a result, a region that is activated by the word 鈥渆lephant鈥� but not 鈥渃hicken鈥� wouldn鈥檛 show up in the averages, even though it聽is actually involved.

鈥淭hose language regions that survive, when you average, are actually connectivity hubs,鈥� says Skipper. 鈥淭hose are the regions that coordinate lots of other regions in the brain.鈥�
The researchers also scanned people鈥檚 brains as they watched movies. The films had been annotated so the team knew when characters were speaking and what they were saying, so they could see changes in brain activity as people processed language. Again, activation wasn鈥檛 confined to regions like Broca鈥檚 area, but encompassed up to 67 per cent of the brain.

at the Massachusetts Institute of Technology disputes the work鈥檚 conclusions. She has studied a network of brain regions known as the multiple demand network, which she says is sometimes activated during language processing, but that . Instead, she says these areas 鈥渁re active during diverse demanding tasks鈥�, and that is what the team鈥檚 meta-analysis picked up.

Skipper, however, says he and his colleagues don鈥檛 believe that such a fixed multiple demand network could exist anatomically. Rather, they believe that nearly any part of the brain can kick in as needed.

It is difficult to reconcile ideas of the whole brain being involved in language with evidence from brain lesions in people who have had a stroke, says at Harvard Medical School. 鈥淥ne lesion in one particular region of the brain and people lose their ability to speak.鈥�

She does think there is some kind of distributed network, though. While regions like Broca鈥檚 area are crucial for language, many other parts of the brain are believed to be involved, she says.

How language evolved: A new idea suggests it鈥檚 all just a game

Our mastery of language presents many mysteries, not least where grammar comes from and how children learn to speak so effortlessly. Now researchers argue that it all makes sense if you think of language as a game of charades

For example, Simonyan and her聽colleagues have used brain imaging to find , including in brain structures like the putamen and thalamus that are far removed from regions like Broca鈥檚 area. Simonyan has also been involved in work showing that . For this聽reason, she describes the new聽study as 鈥渃onfirmatory鈥� rather than 鈥渓ife-changing鈥�.

For Fedorenko, the key point about regions like Broca鈥檚 area is that they respond selectively to language. She says other brain regions can be involved, but those regions aren鈥檛 language specific.

But Skipper doesn鈥檛 think that regions like Broca鈥檚 area are only selective for language. Most brain regions participate in many tasks, he says, something called . 鈥淎ny given 鈥榣anguage region鈥� is talking to more than 40 per cent of the rest of the brain at any given moment, which means, by definition, that those other regions are language specific in those moments,鈥� says Skipper.