Modes I (detail) by Eric J. Heller. Here we see "a series of standing waves in a lemon shaped 'billiard' cavity. Each standing wave is a different resonant frequency of the chamber; these are successive modes starting at the upper left. The modes closely follow the patterns of classical ray paths bouncing within the chamber, providing a visual translation between classical and quantum physics. There are four topologically distinct classes of trajectories in this billiard system. One class is chaotic, and three are quasiperiodic, making this a 'mixed' system with quasiperiodic and chaotic orbits coexisting at the same energy, but differing in position and direction of travel."

Sufficiently large and/or massive objects are commonly referred to as "macroscopic." There is a consensus (i) that macroscopic objects obey the laws of classical mechanics at least "for all practical purposes," and (ii) that the border between the macroscopic and the microscopic is notoriously elusive.

As far as I can see, there are two reasons why this border is so hard to pin down. The first is that we are looking at a limit rather than a boundary between two domains: among the world's more or less fuzzy relative positions, it is the least fuzzy that are special.

The second reason is the lack of an adequate definition. So before we continue to refer to "macroscopic" objects or positions, we need to be clear about what we mean.

Let us note, to begin with, that the possibility of obtaining evidence of the departure of an object O from its classically predictable position calls for detectors whose position probability distributions are narrower than O's — detectors that can probe the region over which O's fuzzy position extends. Such detectors evidently do not exist for those objects that have the sharpest positions in existence.

"Having the sharpest positions in existence" is of course not good enough for a definition. What is clear, though, is that for those things that have "the sharpest positions in existence," the probability of obtaining evidence of departures from their classically predictable motion is very low. Hence among these objects there will be many of which the following is true: every one of their indicated positions is consistent with (i) every prediction that can be made on the basis of their previously indicated positions (in principle if not in practice) and (ii) a classical law of motion. These are the objects that deserve the designation "macroscopic."

This definition does not require that the probability of finding a macroscopic object where classically it could not be, is strictly zero. What it requires is that there be no position-indicating event that is inconsistent with predictions that could in principle be made on the basis of a classical law of motion and earlier position-indicating events.