It seems to me like there’s a twenty-questions style rule-based game waiting to be developed here. That is, rather than asking twenty questions about an object (“is it an animal?”) you ask them about a concept.
There are a few challenges- the first is that natural divisions about concepts are unfamiliar, they’re probably fuzzy instead of binary, and you need fairly deep conceptual knowledge of the thing in question to find it. For example, if the concept is “Athenian direct democracy,” and the first question is “does it relate to history?” the answer is yes but that could easily lead down the wrong trail.
So maybe use a guided search progress. Show the students pre-generated questions and answers, and have them come up with both a general rule that will fit the category and an example of that category, for both the section just ruled in and ruled out. For example, we start off with a blank slate, and see the question “Does it relate to history?” with the answer “Yes.” I say that the “in” category is close to the rule “any concept associated with history”, an example is “the movement of troops during the American Civil War,” and the “out” category is “any concept not associated with history”, an example is “the 3 by 3 identity matrix.”
A second question is asked- “Does it relate to economics?” and the answer is “No.” My “in” category is now “non-economic concepts associated with history,” an example is “universal suffrage,” and my “out” category is “economic history,” and an example is “the exchange rates of private currencies in the United States in 1850.”
Having both in and out categories seems like it helps develop imagination in multiple directions at once, and have the descriptions building on each other (with the attendant change in concepts and examples) seems like it will simulate an organic conversation better.
I don’t know how reliant this method is on a good question-answer trajectory. It seems like it’ll make a big difference, but it also seems like decent trajectories should be fairly easy to generate.
It seems to me like there’s a twenty-questions style rule-based game waiting to be developed here. That is, rather than asking twenty questions about an object (“is it an animal?”) you ask them about a concept.
There are a few challenges- the first is that natural divisions about concepts are unfamiliar, they’re probably fuzzy instead of binary, and you need fairly deep conceptual knowledge of the thing in question to find it. For example, if the concept is “Athenian direct democracy,” and the first question is “does it relate to history?” the answer is yes but that could easily lead down the wrong trail.
So maybe use a guided search progress. Show the students pre-generated questions and answers, and have them come up with both a general rule that will fit the category and an example of that category, for both the section just ruled in and ruled out. For example, we start off with a blank slate, and see the question “Does it relate to history?” with the answer “Yes.” I say that the “in” category is close to the rule “any concept associated with history”, an example is “the movement of troops during the American Civil War,” and the “out” category is “any concept not associated with history”, an example is “the 3 by 3 identity matrix.”
A second question is asked- “Does it relate to economics?” and the answer is “No.” My “in” category is now “non-economic concepts associated with history,” an example is “universal suffrage,” and my “out” category is “economic history,” and an example is “the exchange rates of private currencies in the United States in 1850.”
Having both in and out categories seems like it helps develop imagination in multiple directions at once, and have the descriptions building on each other (with the attendant change in concepts and examples) seems like it will simulate an organic conversation better.
I don’t know how reliant this method is on a good question-answer trajectory. It seems like it’ll make a big difference, but it also seems like decent trajectories should be fairly easy to generate.