The Brain That Changes Itself
By Norman Doig
In 1996 Merzenich, Paula Tallal, Bill Jenkin, and one of Tallal’s colleagues, psychologist Steve Miller, formed the nucleus of a company, Scientific Learning, that is wholly devoted to using Neuroplastic research to help people rewire their brains.
The Scientific Learning staff includes child psychologists, plasticity researchers, experts in human motivation, speech pathologists, engineers, programmers, and animators.
Fast ForWord is the name of the training program they developed for language-impaired and learning-disabled children. The program exercises every basic brain function involved in language from decoding sounds up to comprehension – a kind of cerebral cross-training.
The program offers seven brain exercises. One teaches the children to improve their ability to distinguish short sounds from long sounds. The child has to catch the cow with the computer cursor and hold it by depressing the mouse button. Then suddenly the length of the moo sound changes subtly. At this point the child must release the cow and let it fly away. A child who releases it just after the sound changes scores points. In another game children learn to identify easily confused consonant-vowel combinations, such as “ba” and “da”, first at slower speeds than they occur in normal language, and then at increasingly faster speeds. Another game teaches the children to hear faster and faster frequency glides (sounds like “whoooop” that sweep up). Another teaches them to remember and match sounds. The “fast parts of speech” are used throughout the exercises but have been slowed down with the help of computers, soothe language-disabled children can hear them and develop clear maps for them; then gradually, over the course of the exercises, they are sped up. Whenever a goal is achieved, something funny happens: the character in the animation eats the answer, gets indigestion, gets a funny look on its face, or makes some slapstick move that is unexpected enough to keep the child attentive. This “reward” is a crucial feature of the program, because each time the child is rewarded, his/her brain secretes such neurotransmitters as dopamine and acetylcholine, which help consolidate the map changes he/she has just made. (Dopamine reinforces the reward, and acetylcholine helps the brain “tune in” and sharpen memories.)
Children with milder difficulties typically work at Fast ForWord for an hour and forty minutes a day, five days a week for several weeks, and those with more severe difficulties work for eight to twelve weeks.
The first study results, reported in the journal Science in January 1996, were remarkable. Children with language impairments were divided into two groups, one that did Fast ForWord and a control group that did a computer game that was similar but didn’t train temporal processing or use modified speech. The two groups were matched for age, IQ, and language-processing skills. The children who did Fast ForWord made significant progress on standard speech, language, and auditory-processing tests, ended up with normal or better-than-normal language scores, and kept their gains when retested six weeks after training. They improved far more that children in the control group.
Further study followed five hundred children at thirty-five sites – hospitals, homes, and clinics [in USA]. All were given standardized language tests before and after Fast ForWord training. The study showed that most children’s ability to understand language normalized after Fast ForWord. In many cases, their comprehension rose above normal. The average child who took the program moved ahead 1.8 years of language development in six weeks, remarkably fast progress. A Stanford group did brain scans of twenty dyslexic children, before and after Fast ForWord. The opening scans showed that the children used different parts of their brains for reading than normal children do. After Fast ForWord new scans showed that their brains had begun to normalize. (For instance, they developed increased activity, on average, in the left temporal-parietal cortex, and their scans began to show patterns that were similar to those of children who have no reading problems.)