why cannot assign to function call

[Hendrik van Rooyen]

 "Aaron Brady" <cas.. at gmail.com> wrote:

>No; tuples are composite.  If I flip one bit in a byte somewhere, is
>it the same byte?

Yes and No and No and Others:

Yes it is the byte at the same somewhere in memory.
No it is not the same as it was a moment ago,
because the bit has flipped.
No it is one of the bytes recently corrupted by a 
passing cosmic ray.
[Insert other points of view here]

And of course, depending on what we are doing, 
we take the viewpoint that is useful to the task in
hand.  Taking all at the same time leads to a room 
with soft walls and a jacket with funny sleeves.

Simplifying the porch example:

Here is a family heirloom - it is my grandfather's
axe.  My father fixed it up by replacing the handle,
and I replaced the head because it was so worn.
I still think of it as my grandfather's axe, though.

Natural language seems to have very little to do
with logic, or even common sense, as constructs like
"my grandfather's axe" can be encountered in the wild, 
without causing excessive cognitive dissonance.

I think the root trouble for computers when we want them
to make decisions about stuff like this is that they are too
rigid - their binary or boolean nature forces them into taking
the first branch that looks true at the time.  We probably need
something softer based on majority logic like neural nets, or 
something entirely new if we want to use them to explore this 
kind of stuff. (yes I know we can write a simulation, but a 
simulation is not an interpreter, is not a compiler, is not an 
assembler is not an executable - and its at the executable level 
that the rigidity is most apparent)

- Hendrik