Science: Talking about a revolution

How much thought do we give to the way we use language? Neurologists have uncovered new links between brain-mapping and speech - and this could spell good news for dyslexics. By Rita Carter

Nobody expected Jacob to have a language problem after his stroke. The damage to his brain was limited to the right hemisphere, and in practically all right-handed people, like Jacob, language faculties are known to be situated on the left. Sure enough, when he came round Jacob was able to talk and understand what was said to him. But when he picked up a newspaper to catch up on the news he discovered something weird: the only way he could make out the text was by turning the paper on its side so that the words ran up and down instead of from side to side.

The doctors who studied Jacob (not his real name) at Hadassah Medical School in Jerusalem, concluded that his reading dysfunction demonstrated the existence of a hitherto undiscovered language area of the brain - one specifically concerned with turning visual patterns into words. This, and a number of other recent discoveries, are forcing a subtle redrawing of the language map of the brain. The shifting contours have implications for the diagnosis and possibly the treatment of disorders such as dyslexia (characterised by abnormal difficulties in reading or comprehension) and stroke-induced aphasia (a speech disorder resulting from brain damage).

The parts of the brain that process language were the first functional cerebral "modules" to be charted, and they have stood for nearly 100 years as solid landmarks in what has otherwise largely been mysterious territory. The two main regions concerned with language are named after the neurologists who discovered them: Broca's area is a chunk of tissue that lies at the side of and behind the frontal lobe (beneath the temple) and Wernicke's area is further back, above and behind the ear.

In right-handed adults these modules are nearly always found in the left hemisphere only, but in left-handers they are sometimes found in the right or (very occasionally) in both hemispheres.

Broca and Wernicke discovered the areas by dissecting the brains of people who had suffered language problems following brain injury, and seeing which types of problem matched which areas of damage. Broca, for example, had a prize patient who was called Tan because, although he could understand words, this single syllable was all he could utter. Wernicke's patients could talk fluently but nonsensically, and could not comprehend what was said to them. The scientists demonstrated that these two distinct language problems - the inability to articulate and the inability to comprehend - arose from damage to their respective brain modules, and the disorders became known as Broca's aphasia and Wernicke's aphasia. Reasonably enough, it has since been widely assumed that Broca's area is the part of the brain that turns words into utterances, and Wernicke's area is the part that gives words meaning.

Recent research work suggests that language exercises parts of the brain that were previously thought to have nothing to do with it, and the idea of Broca's and Wernicke's areas as functional islands of language is being replaced by that of a widespread, multi-centred language system extending throughout the brain.

One study suggests that Broca's area is not the seat of articulation at all. Scientists at Hammersmith Hospital in London used positron emission tomography (PET) to chart the pattern of electrical activity in the brains of normal people while they carried out speaking exercises. According to the traditional brain-map such tasks should have set Broca's area alight, but instead an area behind Broca's, the anterior insula, became active.

The anterior insula is the front edge of a deep cleft in the side of the brain that divides the frontal lobes from the hindmost parts. Most of the insula cortex is hidden within its folds, and its functions remain largely unknown, although taste, disgust and sexual excitement are known to excite bits of it. The connection between these things and language articulation may be that the insula verges on the area concerned with the mouth, which, one way or another, plays a part in all these activities.

The Hammersmith study is not the first to implicate the insula in language production. Another brain-imaging study suggests that a fault in this part of the brain may be one of the causes of dyslexia. A joint team of neurologists from the Medical Research Council and the Wellcome Trust found that reading text normally activates the insula, as well as Broca's and Wernicke's areas, but in dyslexics the insula remains quiet. Professor Uta Frith, one of the researchers, concluded that the insula acts in normal people like a musical conductor, causing the language areas to work in concert, creating simultaneous recognition, comprehension and articulation of words. Without it the spoken and written words are not automatically matched up, so dyslexics have to think about each word they see and translate it, laboriously, into speech.

If this theory is correct, it may one day be possible to "switch on" the insula language area using some kind of electrical implant, rather like the cerebral "pace-makers" already used to stimulate "dead" areas of brain in patients with Parkinson's disease.

Whatever the role of the insula in dyslexia, it is unlikely to act alone. Rod Nicholson, a psychologist at the University of Sheffield, says he has evidence for another region of the brain, the cerebellum, being involved in dyslexia. He found that dyslexics tend to have a "deficit" in the activity of the cerebellum, which is known as the brain's "auto-pilot" because it plays an important function in automatisation of tasks such as driving.

But if articulation is not the main function of Broca's area, what is? Other PET studies show that Broca's area is particularly active when a person is asked to come up with a verb in response to a particular noun - "gallop", for example, in response to "horse"; or "eat" in answer to "meat". So it may be the part of the brain that stitches concepts into meaningful sentences and thus makes fluent speech possible. This involves memory, learning, association and the ability to juggle concepts - more complex than mere articulation. The machine-gun-like, staccato sentences of people with Broca's aphasia may therefore reflect an inability to create sentences rather than - as has been assumed until now - a mere inability to spit them out.

Although language is mainly a left-brain skill, the right hemisphere - which "majors" in emotional perception and expression - is essential for investing words with meaning over and above their literal content. People who have right-hemisphere strokes usually continue to speak coherently, but often their words are stripped of emotional tone; they sound flat and droning. Such patients may fail to recognise emotional content in others' speech, becoming deaf to sarcasm, irony and verbal teasing.

Right-hemisphere activation even seems to improve some types of dyslexia; one study found that dyslexic children were able to read anxiety-laden words such as "hit", "bully" and "hurt" more easily than neutral ones - presumably because such words excite the emotion-sensitive right brain. And Jacob's case suggests that right-hemisphere functions such as discerning patterns are also essential for normal reading.

The widespread infiltration of language throughout the brain emphasises that this peculiarly human skill is not just a bolt-on extra, but something that permeates every act of thinking, feeling and remembering. Using language involves not just making sense of the words, but also apprehending the meaning in the spaces between them.