A Quora question asks: How can we say that light tends to travel in straight lines, but when we squeeze it to travel through a very narrow slit, it spreads out? Doesn’t this mean the light is not traveling in a straight line anymore?
The key word is “tends”. For the most part that tendency is what dominates our experience – as can be seen for example by constructing a pin-hole camera.
But there are various situations where it does not apply. Most familiar is the change of direction when light encounters a change of refractive index, but there is also a slight tendency to bend around any obstacle. The wave theory that predicts this was demonstrated by Thomas Young’s two slit experiment but an even more impressive demonstration was perhaps the spot of light directly behind a circular barrier that was predicted by Poisson (as a supposedly ridiculous consequence of the wave theory) and then actually observed in a public demonstration by Arago (having actually perhaps been noted much earlier by Maraldi).
The reason we see don’t see these effects more often is because the wavelength of light is very small and only the part of the beam within a wavelength or so of the barrier experiences any noticeable bending, so it requires a very bright source and a special setup to actually see it. In the case of the circular barrier, it is the rotational symmetry that gives constructive interference between the bent beams from different locations on the boundary. But you can actually see a similar effect from periodic symmetry by looking at the sun through a black woven umbrella (if your twirl the umbrella the bright spots just rotate around around the line to the sun rather than the axis of the umbrella so you can see that they aren’t keeping exactly in line with the actual gaps in the weave).