Jason (jcreed) wrote,
Jason
jcreed
One more weird fact about Pfenning-Davies modal logic that I never noticed before: box isn't dual to diamond in the following sense: the sequent

~[]~~A |- ~~<>~A

is not provable. Whereas with other connectives, even intuitionistically, if you stuff in enough negations, you get entailments like this. E.g.

~(∀ x . ~~A(x)) -||- ~~(∃ x . ~A(x))

This means that classical S4 box and diamond are generated by Pfenning-Davies box and its DeMorgan dual --- and Pfenning-Davies diamond is just some other crazy thing with its own classical dual, for all I can tell.
Tags: logic, modal
Subscribe

  • (no subject)

    Playing around with the agda javascript backend, now. Like, my ears are popping from the sudden change of type-theory-pressure.

  • (no subject)

    Trying to understand in general what kind of diagrammatic interactions between degree-three nodes actually read sensibly in the lambda calculus:

  • (no subject)

    Not sure this is the simplest possible inverse (or even that it is correct) but it makes for a fun diagram:

  • Post a new comment

    Error

    Anonymous comments are disabled in this journal

    switch
    default userpic

    Your reply will be screened

    Your IP address will be recorded 

    Help
  • 4 comments