Mises makes that exact point   
[see Julio Huato's post, http://eh.net/pipermail/hes/2005-May/003038.html ]    
in HUMAN ACTION in his critique of mathematical modeling in economics.  He contrasts the
logic of thought with that of action in arguing that math modeling is inappropriate
theorizing for social theory of the world actually inhabited by humans.  I don't have a
copy at hand or I would cite the chapter and section for those unfamiliar with the book.
  
Sam Bostaph