Children, Reading, and Internal Representations
The "Ideas and Trends Section" of the Sunday NY Times (9/6/98, "Week in Review," p. 4) describes a controversy over the value of reading to young children rooted in the tensions among public belief about how children's minds develop, beliefs about children from popular non-academic psychology, and the views from technical academic psychology. In the popular mind, parents' reading to young children early and often increases the children's verbal abilities and thus gives them a jump start on school. But a non-academic psychologist (Judith Harris) has recently published a book (The Nurture Assumption) which argues that peer pressure, not parental input, is more important to children's development (and children and their peers don't read to each other, one wonders about the effectiveness of parents' reading to children); academic psychologists have long argued while early exposure to any input has impact on developing brain-minds, the brain-mind also develops at its own pace, and the real effects of early intense reading are simply unknown (because reading involves the coordination of visual and linguistic information). The lesson? Go ahead and read as much as you want to children, but don't expect miracles since a lot more than mere exposure (form whatever source) is going on.
This controversy invokes a number of principles of cognitive science. For example, it shows the value of cautious empirical inquiry into the causes and effects of input on developing minds. Caution about predictions is also in order -- does the "mere exposure theory" correspond to the facts? Does the "peer pressure theory" cohere with other developmental explanations? What is the best explanation, an internalist or externalist one? This controversy further shows that the relationship between input (exposure to reading) and output (verbal and school performance) is not transparent: input and output are mediated by internal representations, which themselves are not mere copies of input. Intentional representation-using systems actively work on input and transform it according to the structure of the internal mental code. This makes claims about straightforward input-output correlations (reading to young minds causes good later performance) claims necessitating careful scrutiny rather than self-evident truths.
Now, what about other domains of knowledge? Is your knowledge of faces a function of faces you've been exposed to? Is your knowledge of space and objects a transparent representation of your experience of space and objects?