Walt - I thought it would help to push these out so we might hang our hat on a different peg, if we so choose. These are overlapping, but with different emphasis.
The Reflective Practitioner by Donald Schon - The buzzwords are knowing in action and learning in action. There is emphasis on the indiosyncrasy in the situation to which learning is applied. Thus, even if one "knows the theory" there is still art to applying the theory to the particular situation. Knowledge a la Schon is represented as an unfolding - more and more is brought to bear to match what the practitioner knows about the particular situation. The Schon approach is associated with an apprenticeship model of teaching and learning and where rather than doing experiments students design things as the way to learn. But it has also been applied to fields like psychology, where what is designed is a diagnosis of a patient.
Effortful Study, Anders Ericcson - This is associated with performance as in playing a musical instrument or a game of chess. Effortful study emphasizes doing particular tasks that are not endpoints in themselves but that produce greater understanding and the sequencing of the tasks in a way that produces rapid learning. Effortful study is associated with precocious behavior in youth and approaches to instruction such as the Suzuki Method. There is the question of whether prodigious behavior in one domain can be transferred to other domains but, not doubt, effortful study is tied to what we mean by learning to learn skills.
Process Approach to Writing - The entire idea is to emphasize writing as a means for the writer to learn. This is very much tied up with producing narrative to tie ideas together and see if the narrative "makes sense." There is a discovery process in the writing, discovery in the sense of new realizations by the writer, and that writing is done to improve the writer's understanding.
With each of these the notion of *failure* needs to be understood as should ways in which we validate success or failure. So I hope during the course of our discussion we talk about validation. For now, I'd say that a failure is a complete dead end from which the learner needs to back track to make progress. Otherwise, what we have been referring to as failure may be a useful stepping stone, even if it is not the final product.