Here's another of my more or less arbitrary and somewhat forced deconstructions of how a photograph gets made. There are three elements, because there's always three elements.
In this case: Idea, Subject, and Rendering.
I try not to think of subject in the usual sense, it's not just the thing in the middle of the frame. What I intend here is the collection of everything the camera is pointed at. This might be a single obvious object, or a person's face, but it might also be a chaos of things with no clear single object in the fore. Subject for my purposes here means simply all the stuff that you, the photographer, have chosen to point the camera at.
Rendering is often inextricable from subject, but I choose to separate it out here. It is everything about how you the photographer have chosen to show the subject. The framing, the point of view, the use of color, the use of post-processing effects, and so on.
By idea I mean the meaning to be conveyed, in the usual broad sense I employ. The feeling, the emotion, the message, the whatever-it-is beyond a simple representation of the subject that you have in mind for this picture.
The subject and the rendering together make a picture. You choose where to point the camera, you choose how to make the print (or whatever the final output is) and there you have a picture. It may or may not have an idea associated with it. The point of photography as art is to get an idea attached to the picture.
Here's a simple example where we have all three elements, working together perhaps as well as they ever have:
Weston's Pepper #30. The idea is something about sensuality, something about sex. Just like every single picture Weston ever took, as far as I can tell. The subject is a pepper. The rendering ties the two together, a lot of darkness, a lot of accentuated curves and skin, a carefully selected point of view. A very very carefully selected point of view. The rendering of this subject is such a clear illustration of the idea as to border on smut, as well as on parody.
What happens when we have in hand less than all three elements?
I venture to suggest that most photographers never really get past subject. They see a thing, a scene, a juxtaposition of objects, a subject in the sense defined above. They point the camera, click, and they're done. Some minimalist rendering occurs, often just whatever the camera produces, sometimes with a little fiddling to make the subject "more clear" or something. There's probably a hint of an idea here as well, "this object is pretty" or "these things are cool." Still, it's mostly about subject.
What if all we have is a rendering? The internet is awash in these things as well. HDR techniques applied to nothing at all. Instagram filters applied to pictures of nothing much, or of randomly chosen objects. Lomography. Black and white pictures of.. stuff. Again, there's a little subject, maybe a little idea, but it's mostly about rendering.
If all you have is an idea, there's no picture at all here, and very little to talk about.
What if you have two of the three? Now things start to get interesting.
Suppose you have an idea and a subject. This is, essentially, the pre-visualization problem. Here is this tree, or mountain, or model, or instant in time on the street, a subject. I have an idea in mind. The grandeur of nature, the isolation of man, or more likely something I can't even put in to words. But it's an idea, something I want the picture to mean, something I want the viewer to feel, to experience. What is lacking is a rendering. How shall I make a picture of that subject which conveys my idea?
Ansel Adams writes about this sort of thing a lot, driving past the same view day after day, gradually refining his ideas of season and light until the day came when he simply stopped the car and took the picture, because the environmental aspects of the rendering were right at just that moment. I have written some ideas about how to actually perform this task of finding a pre-visualization, a rendering to suit, in this essay here and some of the following material.
Suppose you have a subject and a rendering but no particular idea? You're on the hunt for an idea. You can go the Modern Fine Art approach, and simply write an idea down in International Art English. Almost any idea will do. Something political is best. It should be loosely connected to the pictures, but only a faint connection is really required.
This isn't really what I am interested in, though. What I like is is a strong connection between the work and the idea. To get this, you're going to have to tinker with your subject and rendering, and work toward some inspiration about an idea, and then complete the circle by making some pictures which are about the idea (based on your original notions of subject and rendering). I wrote about this problem here and here.
For these two situations, I have proposed approaches that reduce to trying to simply invoke inspiration in the one case, and in trying to iteratively sneak up on inspiration by a process of tinkering and trying things out. Both are basically methods for persuading ones unconscious mind to find the missing pieces of the puzzle, and both are essentially built on possession of a large and broad photographic vocabulary. Some mixture of the approaches is surely just as good, if not better. Tinker, try things, then have a nap, a shower. Leave it alone. Tinker some more. Use whatever ratio of tinkering to napping suits you.
Finally, consider the third option, a combination of an idea and a rendering without a subject. This is a lot like the pre-visualization problem, really, except stood on its head. Generally you're going to have some idea, and a rendering concept that supports the idea. If that pair is any good, the rendering already supports the idea just fine. You could be using erotic art tropes for rendering, and an idea of sensuality and sex. You could use shadow-eliminating HDR techniques and some ideas about modernity and technology, perhaps.
This isn't a situation in which I have found myself, to be honest. I'm not sure how common it really is, but I suppose it might turn up now and then. At the moment I do have a minor version of it, as I seek to extend a portfolio with new subject matter. I am currently testing out subjects that share certain things with the previously used subjects, and are wildly different in other ways.
In general, It seems reasonable to seek subjects that support the idea as well, but possibly obliquely. Weston's pepper is a brilliant example of lateral thinking here. His idea is always the same, so we may assume that he had it a priori. This approach to rendering appears quite frequently. One can imagine, at least, that his choice of pepper was a subject-last situation, and he saw, somehow, the erotic possibilities of the vegetable.
And there you have it. That's how to Art. It's how I Art, at any rate.