If you were asking me this, my answer is:
Well…honestly, I would have rather preferred a picture of water than air, but same difference overall for your question I suppose.
In answer, because a photon is a wave in the same way that a wave is a wave in the water.
Instead of thinking of it from the top of the water, like we think of more often when we diagram it as a wave, think of it inside of the water where it isn’t a thing separate from the water, but instead is a displacement in the water.
Same goes for a gust of air.
They aren’t separate things, but displacements moving along within.
A photon is being shown the same way.
It is being shown as not a separate thing from space, but a displacement of it rippling through it in a wave that is more or less eye shaped length wise.
Just like the wave in the water; the wave is just a force of displacement in the water…not the water itself.
I know is sounds like I just countered myself, “It’s the same as the water, but not the water”…what I mean is that in being the force of dispacement inside of space/water/air, then you have it trapped (for lack of a better term) from being separate from space/water/air, but it isn’t the space/water/air itself, but the force of a single “particle” of displacement (the photon) in space.
When it is split, we know what we get.
We get the electron and the positron.
And we know that the spin of an electron is opposite spin of the positron spin.
This kind of starts to fall into place on it’s own with that last bit…you started with one thing A moving along and then it is “hit” by something B and as a result you have A1 and A2 off-shoots and when you have A1 and A2, you find that A1 rotates one direction and A2 rotates the opposite direction.
Then you are told that neither A1 or A2 can be tangibly seen, and that as a result of this A1 and A2 not directly being “held” or seen that an entire set of physics exists to explain that a string in another dimension is bouncing and the tip of that string in this dimension is what gives us A1 all over the place randomly seeming instead of in one direct spot, but still with it’s rotation intact. =D>
Or…A is a wave displacement that split when hitting B length-wise, and that A1 and A2 are spinning in opposing directions because they are on opposite sides of B width-wise and the remainders of A when it hits B.
And the way I showed that is pretty simplistic form of it…water, or air, are actually more alive and less predictable as that nice and tidy diagram, and particles are absolutely no different.
Very fluid.