Comment Re:Define 'observe' (Score 1) 223
Actually, if you would have read past that point, you'd have seen that the process is actually very rigorously defined. It's when whatever particle you use to observe the system interacts with the system. So if you bounce a photon of an electron, that's the observation, not when the photon comes back to hit the photo receptor.
The problem is that this rigorous definition is way past what you want to go into in a beginners guide. A good place to start is looking up quantum decoherence. But the short version is that without an observation, all quantum states are superimposed and we don't know which one the particle or system is in. To get this information, we need to probe it, and since all the possible probes we have are other elementary particles, there is going to be some interaction and the system drops into an Eigenstate with the energy or momentum you wanted to observe (obviously not at the same time, see the uncertainty principle.
In your example, the bat doesn't observe the system directly, the "observation" happens when the photon that hit the bat's eye bounced off whatever it bounced off to get there. Or when the sonar pulse sent by the bat hit whatever it bounced off off.
(and yeah, I know this is so not mathematically rigorous, or correct to the 10th order)