I tried asking a similar question earlier, but based on the answers I got I wasn't able to convey myself well. Here's Feynman's argument (at least as I understand it) for why, from a Quantum perspective, light (roughly) travels in straight lines from points A to B: 1) the photon will travel every path from A to B; with every path, there is an associated amplitude, with a phase which depends on the time of travel and frequency alone. The total probability amplitude for a photon traveling from A to B is given by the sum of the individual amplitudes for all the paths 2) the phases of the paths close to the path of least time will be very close to one another; hence, they will all be in roughly in phase and interfere constructively. 3) as such, the greatest contributor to the total amplitude comes from those paths close to the path of least time. 4) Hence, if the paths near the path of least time or blocked, the total amplitude will drop dramatically; blocking other paths will have negligible effects on the total amplitude. 5) The path of least time between A and B is a straight line. But this doesn't seem like a complete explanation. Suppose we had a laser at A, and a photon detector at B. This explains why the photon counting rate would decrease or go to zero if a barrier were put in the way of the line AB. However, this appears to be unable to explain other features of light-traveling-in-a-straight-line that we commonly observe. For example, if there's some dust, we'll see the lightbeam traveling in straight line. And if the photon detector is our eye-ball instead, we'll see the where the beam is coming from: it traveled from A to B. How can these be explained, from a Quantum perspective?