next up previous
Next: Cartesian Counterfactuals via State Up: Detailed Examples Previous: Formalization of the Skiing

Co-ordinate frames for Skiing

With the axiomatization of the above story given below we can now consider the first two counterfactuals. In each case we take an approximate theory, which will be some subset of the consequences of the axioms in Section 5 , and a co-ordinate frame. From the theory and the frame we judge the truth of the counterfactual. We show that choosing different approximate theories, or choosing different frames can lead to different choices.

We give the co-ordinate frame that the two instructors use. They choose what situations were actual, and what fluents held at the situation tex2html_wrap_inline1454 , and the type of Slope4 as their frame. Here tr is a truth value. The co-ordinates in the frame are the following terms, (two of these are propositions, and so take on the values true or false).

As their core approximate theory they take the axioms about the effects of various ski moves, 15, the axiom relating tex2html_wrap_inline1458 and tex2html_wrap_inline1448 , 16, and the unique names and domain closure axioms and the frame axioms 18. As the current world, they choose the values,

as they agree that Junior did fall on a slope with a bump, while skiing Slope4.

With this frame the following counterfactuals are false and true, respectively.

In contrast the counterfactual

is true. We prove the first two claims formally after introducing the necessary machinery in the next section.

John McCarthy
Wed Jul 12 14:10:43 PDT 2000