A model of the way the brain produces and analyses spoken language (at least originally; it can also be analogised to written language with some modifications). Initially proposed by Carl Wernicke in the 1870s, it was formalized and expanded by Norman Geschwind about 90 years later.

Wernicke had noted that patients with damage to a particular part of the brain (located in the temporal lobe and now called Wernicke's area) suffered from receptive aphasia - they could no longer understand speech, and although they could articulate words as fluently as before, their speech was usually meaningless (for examples see Wernicke's aphasia). He proposed that this area of the brain was reponsible for the comprehension of language.

Slightly earlier in the 19th century, Paul Broca had noted that patients with damage to another part of the brain - Broca's area, in the frontal cortex - suffered from expressive aphasia. They found speaking almost impossible (see Broca's aphasia), although they still appeared capable of understanding the speech of others. It was proposed that Broca's area was responsible for the production of language.

According to the Wernicke-Geschwind model, Wernicke's area contains the auditory codes for words (the information about what they sound like), and also contains the meanings of words. Broca's area contains the articulatory codes for words - the motor commands that tell the mouth and larynx how to articulate each word.

When a word is spoken, its auditory code is first activated in Wernicke's area. This is transmitted to Broca's area, which activates the articulatory code for the word and sends it to the motor cortex, the part of the brain which actually gives the order for the word to be spoken.

When a spoken word is heard, the sound is received by the primary auditory cortex, which transmits the information to Wernicke's area. The sound is matched to the auditory code of the word, which in turn activates its meaning.

This explains the aphasias reported by Broca and Wernicke. Broca's patients could not produce speech because their Broca's area was damaged, meaning that the articulatory codes for the words they wanted to say were damaged or destroyed. However they could still comprehend speech because their Wernicke's area was intact. Wernicke's patients, on the other hand, had no problem producing speech, but the damage to Wernicke's area meant that the meaning of the speech of others was improperly understood, and their own speech had virtually no meaning.

Being a simple model with clear, testable predictions, the theory has generated a great deal of research. Some of this has supported it, and the theory retains some credibility, but it has been criticised on a number of grounds. Its practical basis comes to a great extent from the original investigations by Broca and Wernicke. A series of case studies from the 19th century is not conclusive proof of a model, and it has been claimed that, firstly, "pure" forms of expressive and receptive aphasia are vanishingly rare, and secondly, damage to the cortex is rarely as localised as Broca and Wernicke assumed.

Although some later research has supported it, some has also gone against it. Penfield and Roberts (1959)1 studied the effects of the surgical removal of language areas from the W-G model and found no lasting deficits consistent with the model's predictions. Hecaen and Angelergues (1964)2 came to the same conclusion with 215 patients suffering brain damage in the left hemisphere. Ojemann (1983)3 assessed participants on linguistic tasks such as short-term verbal memory while areas of the cortex were being electrically stimulated. He found that linguistic capability could be disrupted by stimulation of areas outside the boundary of the W-G model, and that participants in any case varied greatly in the organisation of language in the cortex.

Although the model has some validity, it is clearly imperfect. It is generally believed today that the processing of language is likely to involve some parallel component, and is not merely serial as in the Wernicke-Geschwind model.

See also: Broca's area, Wernicke's area, Broca's aphasia, Wernicke's aphasia.
1 - Penfield, W. and Roberts, L. (1959) Speech and Brain Mechanisms, Princeton, NJ: Princeton University Press.
2 - Hecaen H, and Angelergues R. (1964) "Localisation of symptoms in aphasia", (?)
3 - Ojemann, G. (1983) "Brain organization for language from the perspective of electrical stimulation mapping", Behavioral and Brain Sciences, 6, pp.189-230.