From PhilPapers forum Philosophy of Physical Science:

2017-01-11
Happy Unspooky New Year
Huygens revisited
I find Huygens' idea of (secondary) waves a very powerful one, even if his model of light does not really convince me. Somehow it seems to explain important facts about light, how it can go around corners without losing its main, straight direction, and also meet other waves when particles going in straight lines would simply ignore each other.
Even if one does not buy into the destructive interference story, some explanation is needed as to how light seems able to do much more than beams of particles can account for.
There is a very familiar property of beams of light which seems to be very often ignored. Light might travel in straight lines, but it certainly does not travel in straightforward lines. In fact, we know light either as an ever expanding sphere of brightness, like a sun or a bulb,  or as an expanding cone of light, as when it comes out of a flash light.
My metaphor of an infinite number of images had already given me a clue as to the importance of the width of a beam in conveying an image through space. My example of the street lamp and how it appeared each time to move with the position of a pinhole, is I think another clue.
Information contained in a small beam of light will still be there when this beam of light gets wider and wider, and vice versa. Also, it seems like rays have a tendency to push each other out of the way, and claim as much space as possible for themselves. That would explain how rays which have to go through a very narrow pinhole again create a cone of light as soon as they leave the hole. How can they do that?
You would expect rays that enter a pinhole obliquely to follow their course in a straight line, creating a narrow beam of light entering also obliquely the dark chamber. How could they possibly, as particles, change direction and act as is they were originating from a source of light situated at the mouth of the pinhole? Especially since there are also examples of light rays entering a space obliquely.

Light rays get compressed when entering a narrower space, and expand as soon as they can. In the second half of Goethe's video we see that light beams from an arrow formed of lighted lamps all go through the same hole, each one, including the first, filling it. They all accommodate each other to go through the opening and then start to expand as soon as they have passed through.

Huygens's genius was to recognize that the model of projectiles was not adequate, that the behavior of light looked indeed more like that of a wave. But he did not think the concept through and got carried away, just like Young, Fresnel and others, by easy analogies.

The first thing we have to explain is how a single source of light can become two distinct sources each following its own paths and creating together the interference patterns.
First thing first, the duplication of the light source. Imagine one source of light behind the middle part separating two holes or two slits, all at the same level. The light will propagate in the direction of the wall, some rays being stopped by it, others will be let through by the slits.
This is a crucial moment and the way we describe the events will determine the kind of theories we will be able to build.
We can think of the light constituted of rays that move in straight lines in every direction. In this case, the only rays that will be able to pass through the slits will be oblique rays, and there is no reason for them to change direction once they have crossed the slits. This way we would get two isolated fields brightly lit at each extremity of the screen put a distance away from the slits, and there will therefore be no interference phenomena at all. 
This is obviously wrong, so we have to abandon the idea of light rays.
Does that mean we have to abandon also the idea of light as made of particles? That is the jump that Huygens did, later followed by everyone, because it seemed the most straightforward, and therefore logical solution.
We cannot answer this question yet, first we have to know what happens to the light between the source and the slits.
The wave theory sounds very plausible. A big wave is split into two smaller ones, and these two waves fill up the space after the slits in the direction of the screen, with crests and troughs interfering sometimes constructively, other times destructively. There is no doubt that the wave theory offers a very plausible model of the behavior of light.
Is there an alternative?
Let us take our inspiration from the wave concept and apply it to particles. Let us say the light source starts as a single point, which then divides itself into 3, 5, 7, 9.... All the while going forward. Arrived at the left slit be elements 13, 14, 15, 16, 17. Let us rename them as elements 1, 2, 3 ,4 ,5. These five elements will be able to expand more quickly now that they are not hindered by the original outer limits, covering over half of the width of the panel separating them from the other slit, where the same process takes place. Even when added together, the elements in both slits will have much more space to expand in than they would have had if they had remained a part of the original light source. If they keep the same division rate, the distance between them will be relatively greater. Wherever we put the screen, we will see blobs of light that were in fact destined to break up in smaller units until each one was composed of a single photon. What the wave theory considers as locations of destructive interference would simply be the empty spaces between blobs created by each successive division.
Also, we must realize that the empty spaces are never completely empty, especially when two light beams are interacting with each other. In fact, what are considered places of destructive interference are probably the locations where fewer photons have landed, coming from a single beam.
This process of quasi-cellular division should not be considered as something mysterious. A ball of light will emit a condensed cloud of elements in all directions, and the distance between the elements of that cloud, or field if you prefer, will become larger and larger. 

A proper understanding of the behavior of particles should make the concepts of field and wave superfluous, even though many, if not most, of their effects would have to be retained.