|
|
On Oct 23, 6:03 pm, "Peter Olcott" <NoS...@xxxxxxxxxxxxx> wrote:
> "R. Srinivasan" <sradh...@xxxxxxxxxx> wrote in
> messagenews:1161586978.148178.180730@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
>
>
>
> > On Oct 22, 7:09 pm, "Peter Olcott" <NoS...@xxxxxxxxxxxxx> wrote:
> >> An ill-formed question is any question that requires an answer from a
> >> solution
> >> set where no correct answer exists within this solution set. One example of
> >> an
> >> ill-formed question is the following: "How tall are you green or blue?".
>
> > At the instant the question is asked, neither YES nor NO is a correct
> > answer that can be made accessible. But *later*, it turns out that
> > there will be an answer that is either YES or NO
> I don't think so. Not for the scenario that I am envisioning.
>
> int WillHalt(string SourceCode, string InputData) {
> if ( TheProgramWillHalt(SourceCode, InputData) )
> return TRUE;
> else
> return FALSE;
>
> }void LoopIfHalts(string SourceCode) {
> if ( WillHalt(SourceCode, SourceCode) == TRUE )
> while (TRUE) // loop forever
> ;
> else
> return;
>
> }LoopIfHalts(LoopIfHalts);
>
> In this case neither TRUE nor FALSE ever provides a correct answer. The result
> of any answer is ALWAYS toggled to make the result incorrect.
>
You say the result is made "incorrect". That means you have already
conceded that there is a determinate"result", It is just that this
result cannot be correctly predicted. If you think of the program
execution as a time-dependent process, what happens is that a certain
prediction is made by WIllHalt and then LoopifHalts proceeds to do the
opposite later on, *after* "seeing" this prediction. This is what I
mean by saying that *after* the prediction has been made, there is a
final outcome, either Halt or not-Halt, that contradicts the prediction
(making it incorrect, in your words). The problem is that classical
logic requires that a determinate truth value, reflecting the final
outcome (either Halt or not-Halt) must exist at *all* times, even
*when* the prediction is being made by Willhalt (in say, any particular
execution). So the paradox is why WIllHalt cannot get it right even by
guesswork -- after all, one of either Halt or not-Halt must be correct
even when the question is being asked of WIllHalt.
The NAFL position is completely different. If the outcome is truly not
predictable (as either Halt or not-Halt), then there is no determinate
truth value till the human mind perceives in some way that one of these
outcomes must be true (maybe by actually conducting a computational
experiment,though this is not necessary). NAFL would take the stand
that prior to this perception by the human mind, the TM neither halts
not non-Halts.,i.e., there is no determinate truth value independent of
the human mind, unlike classical logic. But this would be a
contradiction, for a Turing Machine *must* either halt or not halt *by
definition*. The only way out of this contradiction is that the halting
problem cannot be undecidable, in which case the human mind already
knows the correct outcome (Halt or not-Halt) via a proof in an
appropriate theory. So this is the only way that a determinate truth
value can exist at all times in NAFL. We have no option but to search
for logical reasons (from the NAFL standpoint) for invalidating
Turing's argument.
This is only a crude explanation, there are many subtleties that I
glossed over in the interests of clarity and brevity.
Regards, RS
|
|