There's a huge difference in the way young
children deal with
brain injury when compared to older children and
adults. As it ages, the
brain becomes less "
plastic," less able to change its own
operation and
composition to find new ways of processing. The older person's brain seeks out strategies for
coping with the loss of function -- ways to
ignore the problem -- instead of changing itself to
solve it. In comparison, the young child's brain will reorganize itself with ease, putting common processing
structures in places where they aren't normally found. Hence the
Kennard Principle: The earlier in life
brain damage is sustained, the better the chances of recovery.
This observation was made in 1936 by Dr. Margaret Kennard from her studies of language processing. When a child's left hemisphere (the hemisphere primarily responsible for language processing) was damaged very early in the child's life, they could still recover full language ability later on. This was confirmed in 1973 by a study done on people who had prenatal or early postnatal left or right hemisphere damage. In the left hemisphere damaged children, language "moved over" into the right hemisphere, and occupied the normally inactive areas corresponding to Broca's and Wernicke's areas there.
Unfortunately for proponents of the Kennard Principle, the reverse changes didn't happen. That is, no visiospatial processing was moved to the left hemisphere in response to right hemisphere damage, instead that processing was noticeably impaired. In fact, as the children with left hemisphere damage got older, they too showed problems with visiospatial processing, as if the reorganization had forced some of those abilities out of the right hemisphere altogether.
In fact they had, a result known as the crowding hypothesis. According to this idea, when damaged, the young brain reorganizes the best way it can at the time, without regard to any future development. Since language abilities are acquired before visiospatial ones, language recovery goes reasonably well but leaves visiospatial processing permanently retarded. In other words, when a lot of damage is done, earlier developing forms of processing "crowd out" later developing ones.
One corollary to the crowding hypothesis is that it isn't the damage location that matters during early development, but instead the extent of the damage decides how much overall brain power will be lost. Damage position becomes more and more important as the brain becomes less plastic. Another corollary is in direct contradiction to the Kennard Principle. Namely, damage that occurs later in life and destroys a single modality of processing may be easier to overcome (with cognitive strategies) than earlier damage that will leave the victim retarded over multiple modalities.