The Tuesday Composition: A Few More Quick thoughts on Direction

Curves and Lines, Swanton Road.  Reading direction, in additition to other factors, influence most Western viewers to "read" the curvy tree first, the straight trees second.
Curves and Lines, Swanton Road. Reading direction, in additition to other factors, influences most Western viewers to "read" the curvy tree first, the straight trees second.

If you like this article, you can now get the book! Joe has expanded the “Tuesday Composition” series into an inspiring new ebook on composition, especially for nature photography. Check it out: The Tuesday Composition.

Last week, we talked about direction in composition and how it relates to movement. There are several other themes that occasionally play into direction in composition, today we’ll briefly talk about a couple of them in no particular order. None of them are relevant to most images, but each of them seems to occasionally come into play when I think about how I’d like to compose an image. Perhaps a few of these ideas will be helpful to you as well.

While I’ve had trouble finding a good reference to the history of the idea, it has been understood for some time that reading direction has a cultural influence on how we look at images. In much of Western culture, we’re more likely to parse images from left to right. If there is an implied horizontal sequence in an image, we’ll probably read the leftmost object as coming first–or at least read it as the more primary object. This is far from an absolute. And in nature photography we’ll rarely have the opportunity to invert the world (at least outside of the digital darkroom), but occasionally I’ve given that idea consideration in composing an image. It’s interesting to take an image and look at how different it feels when flipped horizontally, I recommend trying it on a few of your own images.

A related idea comes from the world of graphic design, when you’re putting together something that mixes images and text, the piece will feel more harmonious if any motion in the image moves in the same direction as the text.

Sometimes we have choices about whether to place an object higher or lower in an image as well.

While it seems to me that photographs that read left-to-right are sometimes more effective (to folks who read left-to-right) than the other way around, I think it’s less often true that reading direction plays into top-to-bottom. In landscape photography, the bottom is usually “near”, and many photographs seem to end up being read more “from near to far” than in any particular direction.

Pogonip and Lake Shore, Mono Lake.  Placing the tree "lower" shows more landscape than if I'd framed the image with the tree at top.
Pogonip and Lake Shore, Mono Lake. Placing the tree "lower" shows more landscape than if I'd framed the image with the tree at top.

Perhaps the simplest observation I’d make is, looking through my images, more often when I’ve had the choice to place an object “high” in an image or “low”, most often it has felt better “low”. I think this is a matter of weight, if you have a smaller “thing” set against a flat background, placing the thing high in the image can feel top-heavy and imbalanced. If it’s an object on a flat ground or seascape, a lower placement in the image will result in a greater sense of space. In  Pogonip and Lake Shore, Mono Lake, if I’d framed the photograph to put the tree high in the image I would have had far less “room” in the image just because of perspective.

As with any of these ideas, exceptions are legion. While a bottom placement of a climber will give her or him space to move into, it robs the image of any sense of the peril to the climber. There’s something about a small climber at the top of a very big wall that really emphasizes the sense of danger and vertigo in a climbing image.

This concludes my discussion of direction for now. Over the next few weeks I have a few posts that talk about different shapes that seem to often come up when we talk about composition, and tie those ideas into what I’ve already said about lines, symmetry, edges, and yes, direction.

This Post Has 2 Comments

  1. Joe,

    Very interesting article. I had wondered about the origins of the reading left to right idea. Too bad you couldn’t locate the source. It left me wondering if there are examples/studies of art in cultures that read right to left that support the idea.

    It seems vertical flipping is far different from horizontal. Right to left versus left to right seems a cultural distinction with no historical/evolutionary importance before the widespread utilization of writing. Presumably, throughout our specie’s history we’ve all perceived the world with up up and down down. So inverting an image to put up down and down up violates more than our culture based predilections. Seeing this way may even be hardwired into our optic/brain system as a consequence of our eyes being located on a vertical rather than horizontal plane. After all we perceive images reflected in mirrors as flipped left-right, but not up-down, even though the physical properties of mirrors make no such distinctions.

    On a less theoretical bent, I have looked at some prints of my more abstract images upside down and also rotated landscape versus portrait. I think my introduction to this exercise was almost two years ago. I made an image of bare trees and a pink//lavender sky reflected in a pond half covered with new ice, the ice (textured rather than smooth) on my side and the open water (textured by a gentle breeze) on the tree side. After I printed it I noticed that I preferred it upside down. There is a tension that plays with perceptual expectations, adding to the obfuscation of the trees and sky caused by the variously textured reflective surfaces. The trees and sky are recognizable as such and so the photo reads as right side up when viewed upside down. But one can also read the plane of the ice receding from near to far. Upside down this makes the top of the image appear closer, contrary to expectations. But the tension seems to increase a mysteriousness latent in the image.

    Well, enough about me. I really like Curves and Lines, Swanton Road. As I read your thoughts about reading left to right I naturally thought about the other factors that drew my eye to the curves (other than the title), namely near to far (the curves tree is obviously in front of the lines trees) and more-so high contrast to less contrast perceptual preferences. The high contrast preference is strong enough that about a quarter of the image horizontally has enough weight to balance the rest. This got me to wondering how this image would work if one pitted the reading left to right preference against the draw to high contrast areas from low contrast areas. So I downloaded the small version of the image presented here, copied it and flipped the copy horizontally.

    The new version doesn’t work as well for me. Sort of contrary to my suspicions, the curves carry so much weight they tip the balance too far. Perhaps they become the “destination”: I read toward them and then become stuck there, whereas in the original my eye is drawn back to the high contrast left, completing a circle that adds to the balance.

    Just for fun I rotated the original and studied it as a portrait, curves at bottom versus on top. It doesn’t quite work either way, but is less off with the curves up. At first I attributed this to that tension of the top of the image feeling closer and the bottom receding that seems to work in my image. But it isn’t balanced enough, too weighty on the top. Then I rotated the copy, which I had previously flipped horizontally. With the curves on top it seems even more out of kilter, but actually kind of works with the curves at the bottom.

    One works better with high contrast curves up, the other vice versa. My first theory here is that as we read the trees left to right any presentation with the trunk end at the left and limb end right appears more organic/natural. Then I noticed this also presents the curvy tree so that the eye, reading right to left, follows the curve of the first big branch up/down into the rest of the image, away from the high contrast, weighty bottom, starting that circle that adds balance to the version you presented. Reading against that curve gets your eye stuck.

    Two further observations about the portrait versions. In looking at the lines section this way there I noticed lots of small horizontal branches that catch the light and seem to balance out the more shadowy trunks. This seems to heighten the abstractness the image would seem to need to work shown vertically. Also, cropping out the lit up green patch helps both versions when presented vertically. I’m not exactly sure why, but it may be that this also makes the image more abstract, less tied to our ordinary perceptions, and thus not so jarring, more acceptable.

    I set out to write a much more straightforward comment of moderate length. Obviously your image and article spurred lots of thoughts. I’m not sure many will plod along through all this theoreticizing, but I learned a lot in the writing.

  2. @Walter: some great observations, I love it that you tried some of the flipping, amazing how much of a difference it makes, isn’t it?

    (Funny story: some years back I had a pretty abstract image win a contest, and get included in an exhibition at the Smithsonian as a result. The contest jury got the image orientation right, but the image got hung wrong. Here’s the original: , the museum hung the piece 90 degrees clockwise.)

Leave a Reply

Close Menu