This posting is meant as a follow-up to the discussion Clark Aldrich and I had going in his post about Dead Reckoning. I'm posting it here because I think it leads the discussion in a new direction.

Most of us are now used to the slogan "learning by doing" and the implicit opposition between "knowing" (brain-based, or even more restrictedly, mind-based) and "doing" which involves a range of associated skills, from perception to reasoning.

I'd like to suggest a way of going beyond the simple dichotomy "knowing/doing" and, in a valiant attempt to rescue knowledge from oblivion (or the perils of rational management, as the first term of KM), propose a list of various approaches to building (rather than just acquiring) knowledge:

. knowing from sensing (perception with its multiple facets),

. knowing from being told by someone who, we believe, already knows,

. knowing from doing,

. knowing from playing by the explicit rules,

. knowing from playing by the implicit rules,

. knowing from playing beyond the rules (explicit or implicit),

. knowing from being expected to know or being recognized as someone who knows (or is in a position to know).

The second one is primarily what we associate with the old school and traditional pedagogy. But focusing on this alone may be doing an injustice to the notion of knowledge, which need not be static. The third one -- much bandied about these days -- encompasses all the playing modes that follow. The last one is the most mysterious, but I think much more common than we suppose. I think it deserves some special attention.

In the perspective of these types of "coming to know", knowing and doing are not opposed. In fact, they are closely linked. The essential ways of building knowledge (doing and being, rather than the passive "being told") are extremely varied.

My question to Clark is, "can these distinctions have any meaning or potential development in simulation"? I expect that they can. The last three seem to be fairly specific to forms of informal (i.e. unplanned and unprogrammed) learning. Knowledge gained in this way may of course subsequently be formalized, but it could be argued that deep learning takes place only through the conversion of the informal into the formal. (I'm not saying I'm ready to argue that, only to entertain the idea and therefore open the debate).

Any takers?