I kind of see what you are inferring to, but "There is only one known method for creating knowledge/complexity (evolution)" is not a strictly true statement. I

I kind of see what you are inferring to, but "There is only one known method for creating knowledge/complexity (evolution)" is not a strictly true statement. If "evolution" implies some sort of selection among competing candidates of "knowledge," then the above statement is strictly false. Say, a -> b, b -> c |=> a -> c. The consequence ( a -> c ) is newly "created" knowledge, because you may not have known that fact when the two premises were stated, yet the consequence required zero competition. Point shortened, logical reasoning requires no competition, hence logical inferences, which can be new knowledge, are not evolved. They are "merely" deduced. I'm just attacking that one statement of yours and not discounting your criticism of the Montessori method. It is a worthwhile clarification though, because you are talking about "serious theories of learning." I have not thought much about it, but I do think there are merits, at least theoretically, to the Montessori method.