I recently finished reading The Children's Machine by Semour Papert (1993). A very interesting read yet somewhat narrow in its scope. While he does talk about education in general, Papert talks mostly about mathematics education and even calls learning mathetics. While he gives justification in his book for using this term, I don't think it will catch on. Some of the thoughts about education are brilliant though.
On page 165 he talks about a student, Debbie whose problem is not a lack of bits of knowledge, but a lack of connections between the knowledge that she has. Papert asserts that perhaps this is a very common problem (1993). most computer aided instruction assumes that a lack of knowledge is present and tries to give instruction to fill in this gap. As I have looked at many different types of CAI, I have seen much of the same thing, the computer tells or shows you something. It gives you information assuming that this information will change your conceptual understanding or behavior about a certain subject. Then it asks you to regurgitate that information to find out if you know it.
But to me building conceptual understanding or changing behavior is much bigger than that. It involves much more than being told something, such as application and experience. The rise of the Internet has given us information that was unimaginable in the past and has provided instant access to that information. The assumption should not be that people do not have the information or knowledge they need, it should be that they don't know how to connect it to other relevant information and experience to make it applicable. If we knew exactly what to do with the information and how to organize it, then anyone who has read about rock climbing should be able to rock climb well, and anyone who has looked up information about a fixing a car should be able to fix it.
In the business world, Performance technologists have been talking about this for years. They explain that most performance problems cannot be fixed with training. Training assumes a lack of knowledge.
This does not mean that the information on the internet is useless, certainly it is far from it. According to Bloom's Taxonomy you must have knowledge before you can comprehend it apply it, synthesize it or evaluate it. Knowledge is necessary but not sufficient. I think that the real trick is not to figure out what information someone needs to know, but how to help learners connect new information to existing information in their mental models of a domain. So far things that have been suggested include microworlds, authentic tasks, experiential learning, cognitive tools, situated learning. Perhaps there are more effective ways not yet discovered....
Papert, S. (1993). The Children's Machine: Rethinking School in the Age of the Computer. Seymour Papert. New York: BasicBooks.