Re: R.J. Lipton and K.W. Regan • You Think We Have Problems

I used to think about the heap problem a lot when I was programming and I decided the heap quits being a heap as soon as you remove one grain because then it becomes two heaps.

The Pascal sorting of the sorites played on moves between heaps and stacks, but I’ve forgotten the details of that epiphany.  The whole-systems point is clear enough though — the system as a whole makes a discrete transition from one state of organization to another.