selfAdjoint said:
I am pretty sure a sandpile exists as a thing, not just as a collection of grains. Tha angle of slope of the pile is an emergent quality, that cannot be attributed to any particular combination of grains, but only to the whole. Even one grain extra can cause it to slump.
Philosophers differ though, about human society. Libertarians tend to say with Mrs. Thatcher, there are no classes, there are just individuals.
But this ignores the fact thet mob behavior, for example, is different from individual human behavior. Emergence does exist in human society too.
Hey, selfAdjoint. Long time

.
I agree with your quote, and add only that it is the realization of such a synergy (if you'll excuse the term) that allows for some of the more important discoveries of the last century (viz. Nash's "Governing Dynamics", Rorty's "Epistemological Behaviorism").
Pit2,
Your question is one of the oldest and (IMHO) most interesting ones of philosophy. Parmenides put forward the idea that all things were "one" and static; change and plurality are mistaken terms. The Eleatics followed suit (Zeno, being the most famous among them).
After David Hume, empiricist Solipsism could be taken as a different form of the same concept, in that all things could be the "impressions" and "ideas" of a singular mind. Kant added the difference between "concept" and "intuition" in order to escape that, but it's still a valid question (that is, if one follows the philosophical paradigm-shift of Descartes and Locke, and thus has a representationalist bias
a priori).
As for my own opinion on the matter, I'd say that plurality does exist physically, but not always linguistically. Take your own example: if a human body is sliced in two, is it now two "bodies"? Well, no, but it is two separate entities, physically speaking. In terms of the linguistic puzzle, I think it's better to simply say that there has ceased being a "human body" at all, and, in its place, are now two blobs of flesh that, for reasons of social and linguistic convention, cannot accurately be called "human".