[net.ai] A halting problem: a meaty respo

jbn@wdl1.UUCP (01/31/86)

      If we're going to argue foundations of mathematics, I have to put in
a plug for Bob Boyer and Jay Moore's continuing effort to build up
mathematics from a brutally simple constructive base, using an automatic
theorem prover to do the grunt work.  See Boyer and Moore, ``A 
Computational Logic'', Academic Press, 1979.  Their approach is based
on recursive function theory and uses a pure-Lisp notation.  I've spent
many happy hours trying to convince their prover that something I wanted
to prove is correct.  I'm submitting a paper to JACM in which I show that
the generally accepted axioms for arrays (using "select" and "store")
are consistent, using constructive machine proofs.

				John Nagle
				League of the Militant Constructivists