Philosophy of Computer Science

A Heuristic for Argument Analysis

Last Update: 31 January 2010

Note: NEW or UPDATED material is highlighted



— Consider 2 players, Q and A, in a question-answer game:


  1. Q asks "C?".


  2. A answers "C, because P1 & P2".


  3. Q must analyze or verify A's argument:

    1. Do I [Q] believe (i.e., agree with) P1?

      • Note that this is a recursive step:

        • Q could ask A: "P1?"
        • And A could give an argument for conclusion P1 with new premises P3 & P4, etc.

    2. Do I believe (i.e., agree with) P2?

      • This is similarly recursive.

    3. Does C follow validly from P1 & P2?

      • More generally, does C follow rationally from P1 & P2
        (if not deductively validly, then perhaps inductively? abductively, etc.?)?

      • This is arguably not similarly recursive on pain of infinite regress,
        at least according to one interpretation of a famous philosophy article by Lewis Carroll
        (see "Carroll's paradox").


  4. Then Q must reason in one of the following ways:

    1. If I [Q] agree with P1
      and if I agree with P2
      and if C follows validly (or rationally) from P1 & P2
      then I logically must agree with C.

    2. But if I really don't agree with C,
      then I must reconsider my agreement with P1
      or with P2
      or with the logic of (P1 & P2) → C

    3. If I agree with P1
      and if I agree with P2
      but the argument is invalid,
      is there a missing premise P3 that would validate the argument and that I would agree with?

      • If so, then I can accept C
        else I should not reject C
        but I do need a new argument for C

    4. If I disagree with P1 or with P2
      then this argument is not a reason for me to believe C;
      so I need a new argument for C.



(Figure courtesy of Tom Fadial.)



Copyright © 2010 by William J. Rapaport (rapaport@buffalo.edu)
http://www.cse.buffalo.edu/~rapaport/584/S10/arganal.html-20100131