I am reading Rudolf Arnheim's book Art and Visual Perception, which is more or less a coherent survey of what was known in 1974 about how we visually perceive the world, and artworks within it. It's pretty interesting.
A core concept is the notion of a percept, that which is perceived rather than merely seen. Consider a photograph of a pair of dice, one a little further away than the other, not visually overlapping.
Each die presents itself as an irregular 6-sided figure, broken up into 4 irregular four sided sub-regions. Each sub-region is a more or less uniform pale grey, with one or more elliptical blotches of a much darker grey. One of the 6-sided figures is smaller than the other. This is the visual impression, this is what lands on the retina.
The percept on the other hand is: a pair of identically sized dice.
Were you shown the photo for only a moment, you would certainly recall that it is a pair of dice. Likely you would not recall what numbers were showing. You do not remember the visual impression, you remember the percept.
Photographers strive to see the visual impression more than other people. Through the finder, we notice whether the dice overlap in the visual field. When the dice do, the visual impression changes radically. Rather than a pair of 6-sided shapes, there is a much more complex shape that represents the two dice. Visually, it is completely different. Perceptually, the difference is negligible unless the forward die more or less completely obscures the hinder one. A slight visual overlap is almost literally discarded by our visual system, we still perceive two dice, one a little further away.
It is said that, when shown an arrangement of pieces on a chess board for a brief study, and then asked to recreate the arrangement, the expert will create a strategically similar arrangement, whereas the tyro will places individual pieces on or near their original locations. The latter arrangement may be more accurate in terms of position, but less accurate strategically. The chess expert perceives the chess board quite differently, compared with the tyro.
Most of us are expert see-ers, in the sense of being able to identify salient features of the world around us. We perceive at a fairly high, fairly abstract, level and must exert effort to see more simply, more directly, more in terms of the direct visual impression.
As a result of this hard-won technical, visual, way of seeing, photographers overrate the visual impression. While maybe nobody would urge you to "fix" your photograph of dice by eliminating a visual overlap, we are constantly reminded not to let poles "grow out of" people's heads, which is precisely the same issue. An accidental visual overlap,
which may or may not be perceived by non-photographers, is a very great evil, to be cloned out or re-shot.
In reality, if the percept is clear, there is a flagpole behind the person, then let it grow out of their head. Who cares?
This brings me around to one of my perennial favorite topics, the lighting of the photographic portrait.
Lighting the face this way or that alters the visual impression radically, in ways that the enthusiastic strobist notes instantly.
The percept does not budge. Not one bit. We are people-recognizing machines, first and foremost. One light, ten lights, it doesn't matter, we're going to recognize that it's a person and read their mood instantly. We're going to perceive Aunt Sally in an instant.
We are told by strobist web sites to use short lighting on fat people, because it makes them look thinner. (Turn the face a little away from the camera to make the face visually asymmetrical, and then light up the narrowed half.) This isn't really true. We know Uncle Bill, and when we look at the picture we're not really even going to see the visual impression the picture leaves, we're going to more or less directly perceive Uncle Bill and his great moon face. Sorry, Bill.
This is not to say that short lighting is useless on Bill. This particular picture of Bill is, maybe, more flattering, even though it in no meaningful way conceals his fleshy magnificence. What it does, maybe, is allow us to recognize Bill without drawing particular attention to the dizzying breadth of his face. We are invited, rather, to dwell upon the twinkling eye and the neat haircut.
At issue here is really what it means for Bill to "look thinner." In one sense, there is no making Bill look thinner any more than there is making a die look like a hexagon. But in another sense, we can certainly emphasize or attenuate this aspect of that of Bill, or of our die. We can draw attention toward one thing and away from another. By lighting the pair of pips on top, and darkening the other visible sides, the die takes on the character of both a die, and of a rolled snake-eyes.
The die remains a die, unequivocally, but, maybe, the rolled two enters our percept of the die. We might recall the number on top.
The painter can actually paint a thinner man with Bill-like features, or can flatten a die out and render its cube-ness versus hexagon-ness ambiguous. The photographer has no such luxury, short of becoming essentially a painter.
Short of optical illusions, or painting, the photographer has to find actual real things to make the picture out of, with the assumption that they will be perceived more or less as they are. You cannot simply take a picture of a boring thing "in a way that makes it interesting," nor an ugly thing beautiful, nor a large thing small.
If the goal is to render the essentially ugly thing beautiful, you cannot really conceal the ugliness. If you yourself do not see beauty in the thing, you ought to give it up as a bad job. If, on the other hand, you see beauty in the ugly thing, then your job is not to conceal the ugliness but to reveal the beauty which you perceive. There may be more to this than simply lighting the shot and picking the viewpoint. Perhaps you need many photographs, perhaps you need words, or sound, or smell. Perhaps you need to circle the important bit with a red pen.
When you look at something, you perceive much more than the visual impression would suggest. The photograph discards most of that, leaving only a spare and limited visual impression. From that spare and limited visual impression, your audience constructs anew the percept. Your ability to manipulate that perception to match your own is remarkably limited.