Re: Generic thanks
From: Qakare (qakare_at_hotmail.com)
Date: 09/17/04
- Next message: Reinhold (Rey) Aman: "Re: More stupidities"
- Previous message: Jarel: "Re: another Polish loanword from Chinese"
- In reply to: grapheus: "Re: Generic thanks"
- Next in thread: grapheus: "Re: Generic thanks"
- Reply: grapheus: "Re: Generic thanks"
- Messages sorted by: [ date ] [ thread ]
Date: 17 Sep 2004 00:44:49 -0700
grapheus@www.com (grapheus) wrote in message news:<337ae51f.0409161036.2d22988d@posting.google.com>...
> qakare@hotmail.com (Qakare) wrote in message news:<dad05270.0409160302.45436674@posting.google.com>...
> > "Bobby D. Bryant" <bdbryant@mail.utexas.edu> wrote in message news:<pan.2004.09.16.06.31.09.343764@mail.utexas.edu>...
> > > On Wed, 15 Sep 2004 22:39:26 +0000, r wrote:
> > >
> > > > Note that this is all probabilistic. It is not impossible that the
> > > > Phaistos disc will ever be interpreted without further data, just very,
> > > > very improbable.
> > >
> > > How many possible mappings between strings of symbols and meanings are
> > > there, and how many of those are ruled out by what little we know about
> > > the Phaistos disk?
> >
> > We have more then 60 signs in a syllabic script like Linear B. This
> > means if the script of the Phaistos Disc is a syllabic script there
> > are more then 10^69 possibilities to read the Phaistos Disc! This means
> > it is impossible to found the right solution by chance.
>
> Surely !..
>
> >
> > >
> > > Are we talking about thousands, billions, or an infinite number of
> > > possibilities?
> > >
> > > If someone did offer the correct decipherment, would we actually be able
> > > to distinguish it from all the other possible decipherments?
> >
> > Yes, because the only way to decipher the script is to proof every
> > step! Therefore it's possible to check a decipherment by checking it
> > step by step.
>
> This is IMPOSSIBLE !.. Because the diverse steps are the choice of
> the type of script - the choice of the language - the choice of a
> "starting hypothesis" - etc. It's impossible to "check" these choices
> one by one.
That something is impossible does not mean that it is not necessary.
> The only thing which can be done is taking for EACH choice
> "the most PROBABLE solution". For instance, for the "type of script",
> the most probable choice is a "Syllabic Script".
> But, ONCE the decipherment has been reached
The problem is that with the wrong steps you will reach a wrong
decipherment without noticing it.
>, a GLOBAL VERIFICATION -if
> decisive!- is enough for demonstrating that EVERY CHOICE has been
> correct...
This is obviously wrong. Without knowing the reading direction, the
writing system, the language, the text structure ... how will you
verificate the solution? But even if you have a good criteria and you
are able to show at the end that your solution is right or wrong,
nothing is won. How will you know in the 10^69 - 1 cases when you are
wrong which of your choices is the first wrong one? You need to know
that, because you have not the time to check all the possible choices
for every step. And to found that out the only way is to give proof
for every step you are going.
> >
> > > If someone happened to get it right by means of an unverifiable guess,
> > > would it actually count as a 'decipherment'?
> >
> > An unverifiable guess must be wrong in a universe of 10^69
> > possibilities.
>
> No. It is not "necessary wrong". It has one chance out of 10exp.69 !..
> But EXTERNAL VERIFICATION - if possible!- is the only way to show
> wheter it is right or wrong...
Grapheus, obviously you didn't understand the meaning of 10^69. For
instance the number of atoms on earth is only 10^50 (see
http://education.jlab.org/qa/mathatom_05.html).
Let us assume that you have a easy to calculate criteria to see if you
found the right solution by chance (Such a criteria does not exist,
because there are millions of meaningful solutions there!). And you
are able to check with that criteria with one computer 1 000 000
solutions in a second. Under this circumstances if you are using 1 000
000 computers for 1 000 000 years you have only checked 10^25
solutions (1 000 000solutions * 1 000 000 computers * 1 000 000years
*365days * 24h * 60min * 60s). And 10^69 - 10^25 is still 10^69!
qakare
>
> grapheus
- Next message: Reinhold (Rey) Aman: "Re: More stupidities"
- Previous message: Jarel: "Re: another Polish loanword from Chinese"
- In reply to: grapheus: "Re: Generic thanks"
- Next in thread: grapheus: "Re: Generic thanks"
- Reply: grapheus: "Re: Generic thanks"
- Messages sorted by: [ date ] [ thread ]
Relevant Pages
|