Le 17-juin-05, à 07:19, Russell Standish a écrit :

Hmm - this is really a definition of a universal machine. That such a
machine exists is a theorem. Neither depend on the Church-Turing
thesis, which says that any "effective" computation can be done using
a Turing machine (or recursive function, or equivalent). Of course the
latter statement can be considered a definition, or a formalisation,
of the term "effective computation.

Hmm - I disagree. Once you give a definition of what a turing machine is, or of what a program fortran is, then it is a theorem that universal turing machine exists and that universal fortran program exists. To say that a universal machine exists, computing by definition *all* computable function, without any "turing" or "fortran" qualification, you need Church thesis.

Bruno


http://iridia.ulb.ac.be/~marchal/


Reply via email to