XPost: alt.privacy, sci.crypt   
   From: nobody@goldmark.org   
      
   On 11-02-11 12:49 AM, VD wrote:   
   > On Fri, 11 Feb 2011 00:09:10 -0600, Jeffrey Goldberg wrote:   
      
   >> I think Tom Riddle's maternal grandfather used it [to keep secrets] when   
   talking to   
   >> his family when a visitor arrived.   
   >   
   > Marvolo Gaunt. Very good! All the Gaunt family were Parselmouths.   
      
   Thanks! I forgot his name and was too lazy to take the key strokes to   
   look it up.   
      
   >> Yet it is still possible for someone who doesn't have the gift of   
   >> Parseltongue to fake it well enough to authenticate, as Ron did in   
   >> Deathly Hallows.   
   >   
   > To open the Chamber of Secrets.   
   >   
   >> So it can, to some extent, be learned by outsiders.   
   >   
   > One can mimic any language without learning it.   
      
   In this case, mimicking the language allowed a successful replay attack.   
      
   I am not saying that Parseltongue is like a human language in general.   
   Of course it isn't.  What I am saying is that with respect to whether it   
   should be considered cryptography it is like natural languages.  The   
   things that make Parseltongue unlike human languages don't seem to play   
   a role in deciding whether it can be consider cryptography.   
      
   If it were impossible for a non-Parsalmouth to authenticate this way,   
   then it would be different from human languages with respect to   
   something that matters to the question of cryptography. I would have   
   preferred for it to have worked that way, but the Deathly Hallows was   
   already long enough without having to write an episode about retrieving   
   the basilisk fang that made more sense.   
      
   [Note regarding spoilers: I figure everyone here has read all the books   
   or doesn't care enough to be bothered by spoilers.]   
      
   > Again, I vote "Yes" because PT has the same underlying reason for use   
   > as cryptography. Hide and exchange data only between those who have   
   > the keys.   
      
   Again, I won't vote because I haven't settled on a definition of   
   "cryptography" that I'm fully happy with and would be decisive. So as I   
   said up top, the question as asked isn't interesting, but the question   
   of whether Parsaltongue differs from natural languages in its   
   cryptographic properties does seem meaningful. To that question, I think   
   that there is no relevant difference.   
      
   Cheers,   
      
   -j   
      
      
   --   
   Jeffrey Goldberg          http://goldmark.org/jeff/   
   I rarely read HTML or poorly quoting posts   
   Reply-To address is valid   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)