John,
My point was not that one could not specify how to draw a line
in logic, but that specifying the line does not actually
draw it - for
that you have to translate some bit stream into so motor
action, or to
switch the CRT beam on at a particular point, etc.
True, but...
My point being, that
organizations do not operate computers to make lights on the
front flash
on or off, or as expensive room heaters. The meaning of a computer
system is always the behaviour of the organization that uses it.
... that is rather a stretch. The meaning IS the
behavior?? No, the behavior depends (in part) on
the meaning: but the meaning is what it is even
if nobody acts on the information.
A very basic problem with your point is that it
seems to force us to adopt a process-based or
procedural approach to semantics, which takes us
exactly in the reverse direction of the evolution
from hard-wired specific codings to assertional
ontologies which has got us to the present. This
much progress was hard-won, and it would be a
great shame to be taken back to UML, or maybe
even Fortran, by worries arising from management
theory.