Thursday, July 08, 2004


Not much to look at, but this is what we get from a letter-based attempt to define rules... for any letter what is the likelihood of any given letter following it? It's a very rough prototype, but it does seem to yield reliably "realistic" patterns of letter correspondence...

instha t d realize nd wh An outhIoupe of tatss laci wat nchous, onthathepons Als. fond f Medacarartyomec Athous, Cioremend din Gous Ateneen Are r o trofas tre bily fath S. lp oonstit An Ph iS tinobuso

Con; thergean Aldge ina tithr ils Aiasstat S ounthnceco e ofin junthashaly Ses e An ond Angof thor wind At ty Pessiofuctialinitiofone oy iomar Gamictalerve wesucourelurd of onde Aly oregrsimplac o ioupon Abutsinf toflangure ofor. Peen Pins in Ase taripicenthegrty tilighedet-gl-hamelieingere rs Pre tl Plex Alan. Prysurutene berl SthIch Prt Pprea ntit Cany M Spoun Pre foust Stesofof of Co izag


Post a Comment

<< Home