These remarks are quite likely more about how I personally react to photographs and art than the average posting on this blog. Still, perhaps you might find something to take away.
Our theme for today is the uncanny valley, a concept from animation and robotics. Robots and animations fall into the uncanny valley when they appear almost human, but not quite. These things are rather disturbing to people - we will accept a robot that looks like a robot, and we will accept a robot that is an extremely good imitation human, but we won't accept things that are good-but-not-perfect. There are lots of cases where making something that's neither one thing nor the other produces a result that is simply unlikeable.
Item one: Why is it so hard to take a good photograph of texture or pattern? What I mean by this is, why can't I take a successful photograph of a pattern that looks like a Mondrian, or a frame filled with chaos like a Pollack? Not to denigrate Mondrian and Pollack, but making a painting of that sort that is at any rate recognizable as a painting isn't particularly hard. I can draw some lines in a rectangle, and people will not talk about how there's no subject, and they won't be irritated by not knowing what to look at, and so on. My lines are no Mondrian, but they will at least function as a piece of visual art, good or bad. If I take a similar photograph of a non-hierarchical collection of linear objects similarly arranged, probably I will hate it, and probably most people will too. I will in fact be annoyed that there is no subject, no well defined visual center, and a few other annoyances.
I am pretty sure I am not alone in this. Photographic abstracts tend to have visual centers, leading lines, and so on. They are abstract because the objects in the frame are not easily identified, not because they lack the traditional functional elements of a photograph. Of course, there are probably exceptions?
My working theory is that the problematic photographs of this sort occupy a middle ground between an abstract representation of texture, color, line, and a representation of a thing. The viewer is ok with the painting, because we accept that paintings needn't be of anything. A photograph, by its nature is always of something. Surely, our mind declares, the photographer was looking at something. What is that something? Why was the camera pointed in that direction? Show me!
Surely, of course, there are elements of tradition in play here as well. We see almost exclusively photographs of identifiable things. Most of the rest are at any rate of things, even if we cannot identify them. So, photographs of nothing are weird and unpalatable to us.
Item two: One of the easiest ways to ruin a good photo is to treat it as something it is not. If you post-process a candid to look like a formal portrait, the odds are good that everyone will hate it. It is now a bad portrait, and a bad candid, it falls into a zone between the two. It is twice as bad as a bad photograph, since it is two bad photographs. If you print a dramatic and moody piece in tones of entirely pale grey, you will likewise push it into a zone where the image cannot decide if it is a subtle tone poem, or a powerful and dark mood piece. Viewers probably are not going to like it. The cues conflict, we don't know what to make of it.
The best photographs have been allowed to be what the are. Sometimes this is quite difficult. If the image is out of focus and motion blurred, one is tempted to try out massive detail-recovery procedures rather than letting the motion be blurred and letting the photograph be what it is. If the photographer had in mind a formal portrait, but shot a back-stage candid, she might fight the image and ruin it, rather than accepting it as a superb candid.
Pick a side, join the battle. Fight! Fight! Fight!
But first, pick a side and commit.