So what exactly is light, anyway?
It’s a tough question. Isaac Newton thought it was composed of streams of microscopic particles he called corpuscles. Really it wasn’t a bad idea. Light rays travel in straight lines just like fast moving projectiles, light bounces off objects in a manner not entirely unlike a ricocheting bullet, and if you try hard enough you can even explain refraction in terms of particles being slowed in matter online the lines of a ball bearing sinking in molasses.
But it’s pretty hard for the particle view to explain diffraction and interference, which are both very characteristic of waves. After Newton, the view of light as a wave began to predominate. Like sound waves in air or ripples in water, light seemed to behave very much like a wave:
Eventually James Clerk Maxwell came along and showed that electric and magnetic fields obeyed the wave equation and that in fact light was nothing more than waves of those fields. Case closed.
Well, almost. Water waves and sound waves need something material to “wave”. Physicists assumed there was a thin and invisible medium called luminiferous aether permeating the universe, and that light waves were oscillations of this substance. But in 1887, Albert Michelson and Edward Morley were able to do a very careful experiment to measure the speed of light as the earth moved through the aether as it orbited in space. Their experiment came up negative. There just wasn’t any aether.
Well that wasn’t necessarily so bad. Maybe some waves just don’t need a medium, and when Einstein came along with special relativity, it was immediately clear that light would always travel at the speed of light no matter what frame of reference you happened to be measuring it from. Light was a wave, if a special one that didn’t require a medium to wave.
Around this time scientists were also fiddling around with the photoelectric effect. When light was shined on certain objects, those objects would release electrons. The wave theory could explain that just fine, since waves both carry and transfer energy. But it wasn’t so good at explaining the details. If you increase the intensity of the light, more electrons are emitted – but the energy of each released electron doesn’t change. If you increase the frequency of light (ie, shift it toward the blue end of the spectrum), the energy of each released electron does increase. Even worse, if you shine a very dim light the electrons are emitted at a low rate. But though the rate is small, the emission of electrons begins immediately – even if the light hasn’t been shining long enough to have transferred enough energy.
So Einstein proposed that the wave theory wasn’t quite right. Light came in discrete units called photons, and each had a certain specific energy dependent on the frequency of the light. Photons would hit the electrons in the material and eject them with an energy equal to the energy of the photon minus the energy required to pop them free of the material.
This explained the photoelectric effect pretty well. Higher light intensity meant more photons, not more energy per photon. Thus more electrons generated, but not at a higher energy per electron. In fact, Einstein got his Nobel Prize for his work on the photoelectric effect rather than his work on relativity.
So is the particle theory right after all? If so, what about all the wave behavior of light? The answer is, perhaps unexpectedly, that the particle theory is still wrong. Despite my casual particle-like description of photons hitting electrons, photons aren’t particles. They don’t have quantum mechanical wavefunctions, and they don’t have positions even in the fuzzy quantum mechanical sense that (say) a proton has a position. Now pretty clearly it’s possible to define a function of position that determines the likelihood of a particular atom to interact with a photon, but that’s not quite the same thing. It’s more accurate to say that a photon is a quantum mechanical object that isn’t a wave or a particle. It’s something much more mathematically complicated that has certain particle-like or wave-like properties. It wasn’t until Feynman and colleagues developed the full quantum electrodynamic description of light that this was all figured out. And that just for now; it’s always possible that some new observation may disagree with QED, and in that case the search will be on for a more comprehensive theory. For now though, QED remains possibly the most accurately tested theory in all of physics, with no observational disagreements thus far.
This sort of situation is not entirely satisfying. It would be easier if it were possible to explain light in terms of things we’re familiar with in everyday experience, like waves and particles. But to the extent that we do so, we lose some of the more elusive but real properties of light. Most of the time this isn’t a problem. Light as a wave can explain pretty much every macroscopic property of light, and light as a particle can explain many of the microscopic properties of light. Still, you should always keep when I or anyone else talks about light as a wave or particle, it’s an approximation. Nature is more subtle.