Or an elephant?

There are lots of things that we know when we see them, but the boundaries are indistinct. Take the colour blue for example. We all know the colour blue when we see it, but the actual boundary with green is indistinct, which is why you occasionally get people arguing whether a particular item is blue or green.

I'm sure the astronomers who voted for the criteria 'cleared the neighbourhood of its orbit' will know an example when they see it - which is why they were so adamant that Mercury, Venus.....Neptune had done so and Pluto hadn't. But they didn't really have much time to think of a more formal definition. I expect this bit of the planet definition will be firmed up in the near future. (The astronomers who opposed the definition didn't principally do so because this criteria was unclear, but they disagreed as to whether it should be a criteria at all.)

As a planet would clear its neighbourhood through its gravitational influence, what will be ignored is any object in those few circumstances where a planet's gravity cannot clear it (no matter how big the planet is). This includes any object in orbit around the planet; any object in the Trojan points, and any object in a synchronous orbit.

But there will be boundary issues as with any definition of this nature. They'll deal with them as and when they get an actual physical example. (There's no point doing so beforehand, as there may be some unknown physical process at work that means this particular boundary will never be an issue.)