Faculty Learning Community: Agenda and Notes (November 30, 2009)
Mark H. Johnson and Yuko Munakata,"Processes of change in brain and cognitive development." TRENDS in Cognitive Sciences
9, 3 (March 2005): 152-158.
Lunch served in Dorothy Vernon Room of Haffner Dining Hall
Agenda (from Mike Sears):
I'd like to use this week's discussion to explore whether discourse in other disciplines regarding learning might be useful for our endeavors. Specifically, cognitive science and computer science has been exploring the roles of different mechanisms of learning in cognitive development and the effectiveness of such mechanisms under various circumstances. I've attached a manuscript that describes some of these mechanisms. If you get a chance, please give the paper a quick look. I'm not suggesting that this paper has direct applications for teaching quantitative reasoning, but it certainly might operate as an effective metaphor when thinking about our teaching approaches. Regardless if you have time to read the paper or not, please do join the discussion on Monday.
A meeting summary (Paul)
The cognitive science/neural network distinction between "supervised" and "unsupervised" learning generated a lot of interesting conversation about classroom objectives and practices. "Unsupervised", in the technical literature, is equivilent to "without feedback," and "supervised" to a teacher knowing the "right" answer and providing feedback to students about their proximity to that answer. Each has its virtues and problems. Among the latter for unsupervised learining is "inefficiency" and for supervised learning is a narrow domain specificity of acquired knowledge.
The extremes of the technical distinction led to considerable discussion and clarification of peoples' preferred classroom objectives and atmosphere. Unsupervised learning involves changes reflecting internal dynamics in response to whatever input occurs. This is always taking place and probably needs to be paid more widely recognized as playing a role in classroom activity but is harder to think about in terms of current educational practice. What emerged as a perhaps useful description of spaces between the two technical extremes was three levels of organization:
- structured classrooms (dominated by unsupervised learning but with deliberate attention to the input being used)
- structured classrooms with feedback about how students are trying to learn
- structured classrooms with feedback about the "right" answer
This in turn generated extensive discussion about the role/desirability of various kinds of feedback and the extent to which one wanted students to become "autonomous" as opposed to "social" learners. Feedback about the likely success of particular directions of change risks the possibility of again producing knowledge of restricted domain specificity. And encouraging "social" learning may play into the inclination of individuals to "validate" their own current understandings rather than reach for new ones.
An alternative, perhaps, is to think of both internal dynamics and social learning as processes similarly involving "feedback" about the adequacy of current understandings in new contexts. Such feedback is not about "errors" in either performance or method of approach but is instead about differences between what is currently being made sense of and what is yet to be made sense of. In this context, the point of "structure" is not to provide a guide to particular outcomes conceived in advance but rather to assure that all participants in classrooms are effectively challenged to themselves achieve new and wider understandings, challenged not only by their internal dynamics but by their interactions with the differing understandings of others. To achieve learning atmospheres of this sort would mean not only giving up the ideas of "errors," but also having as an overriding objective the facilitation of inquiry skills that would serve in a wide variety of contexts, including unanticipated ones.