Composition & Perspective

My drawing has taken another one of those mysterious steps forward.  A big part of this improvement is my recent discovery (while in Europe) that to situate figures (or any object) in a landscape, one must first compose the landscape on the page, and place the figure within it. To make the figure the sole focus of the drawing is to NOT SITUATE the figure in relation to anything else, making the figure a portrait or a study, but probably not a composition.

This last point is becoming increasingly clear to me… the need to make a composition. Certain things arise in making a coherent composition that do not arise in making studies. First, one must consider the relation of the overall space to the more specific objects within that space. Second, one must consider perspective. Not simply linear perspective as usually understood (i.e., making houses and fences look like they recede)… or aerial perspective as usually understood (reducing tone and color saturation as objects recede). Instead, I'm thinking of perspective in the broadest possible way… as in… how does the visual field relate to myself.

The visual can relate two ways… first, in terms of visual perception (how do the objects appear to my vision)… and secondly, in psychological terms (what do these objects mean to me).  These two things are related, as the ability to access psychological meanings is related to the ability to understand how things appear to you visually (visual perspective, so to speak).  This seems intuitively true, as whenever we draw something well, we feel a connection to the drawing. We might think that that success of such a drawing magically derives from out connectedness the subject matter of the drawing….but I think that our connectedness to it derives first from our ability to draw it in the first place.

The kind of visual perspective required to make compositions (vs. studies) is important. To seriously persue visual perspective requires certain basic structures be enforced in a drawing. This might all seem trivial, but I think they are not so much trivial as they are fundamental… and like all fundamentals, they are easily overlooked… and the resulting works become unhinged. The first fundamental is the establishment of a horizon line. Without this, one can easily lose track up what is up and what is down. This is probably not a problem in figure studies and portraits and other such close-up work, where the object is not integrated into a visual space. But if you try to construct larger spaces, and to represent complex spatial relations, you must know what is above your eye line, and what is below.

It is not necessary to literally draw the horizon line (obviously), though it is useful to do so in the beginning of the drawing. I realize that doing so seem silly to a more experienced draftsman, who is (in a sense) beyond such rudimentary techniques, and can probably maintain a horizon line instinctively in their minds. But I have to wonder how many people can actually do this consistently… or how many people can render complex spatial relationships without such techniques.

I suspect many people begin their drawing by focusing on an object of interest and drawing it, and then fitting other things around it. These "other things" end up being the space in which the object is located… i.e., the context of the drawing. Such drawings are (in a sense) done in reverse… where the space is constructed AROUND the object, rather than the object being place WITHIN the space. The difference between these two approaches may not be very apparent in the finished work.

For example, if I draw a bottle sitting on a table top, I can approach it in either of these two ways. I can draw the table top and the surrounding space, and then situate the bottle on top of it. Or, I can draw the bottle, and then draw the table-top and surrounding space. The two drawings might be very similar… BUT… if I draw the space first, I am able to consider MANY possible relationships from the beginning, whereas if I prioritize the bottle by drawing it first, I can  accommodate FAR FEWER spatial relationships… if for no other reason than that simply by choosing a size of the bottle (at the beginning of the drawing) I constrain the size of all the other relationships, and it is possible that I don't leave enough space for other visual elements to be realized. These unrealized relationships represent possible limits on the drawing.

By focusing on the object of interest, I feel like I a hamstring my drawings from the beginning. The drawing becomes a glorified study, rather than a composition… as the object isn't COMPOSED into a space, but the space retro-fitted around an object. It occurs to me that this has a parallel in personal identity… such as… do we exist in the world, or does the world revolve around us. When we become ego maniacal and demand that the world revolve around ourselves, we limit what we can consider. The ego maniac only considers things that directly relate to themselves at the moment, and therefore cut themselves off from maximum experience. Drawings that proceed from the focus on a single object have the same problem… they are egomaniacal drawings, so to speak.

At any rate, I suppose both approaches are valid, but I know for sure that taking the OVERALL approach is necessary for making the kinds of large narrative type paintings one might admire. When I see the David paintings, the Gericault paintings, and the Delacroix paintings in the Louvre… I know for a fact that I am looking at visual constructions of high complexity. I think to myself… "How does one get to that level of composing ideas?". It's tempting to think that their paintings result from some SUPER DRAWING SKILL and SUPER PAINTING SKILL, wherein they make things LOOK REAL. And while they do have high level drawing and painting skills, that is not what their paintings are fundamentally about.

Their paintings are fundamentally about composed elements… and those composed elements are realized through an appropriate level of drawing and painting skills. The more that one can draw and paint, the more things can be coherently incorporated into the composition. Consider Holbein's "The Ambassadors", with two richly clad figures standing on either side of a table that contains many symbolically important objects. Such a composition is not possible without the ability to draw and paint the many detailed elements of the painting. If Holbein's skills were less richly detailed, he would have to do a different composition. This is not to say that detailed painting is required to do great compositions… but some level of proficiency is required, otherwise visual elements cannot be rendered, leaving one fewer and fewer things to work with. After all, you need to have actual things you can represent and relate.  At the very least, one's drawing and painting skills need to be appropriate to the ambition of the composition.

Without content, one is hard pressed to created meanings through composed associations. That's actually a good definition of composition… COMPOSED ASSOCIATIONS. You compose the associations of visual elements in order to convey some intended meanings. Without content (visual elements), you end up on the far abstract end of the spectrum, like Rothko, where you compose in color blocks. One can argue that Rothko's compositions are beautiful and meaningful… but the beauty and meaning of such works derives NOT from an expression of the intellect as it relates to the specifics of the visual world, but to unconscious associations between formal elements of line, color, shape, etc. Such highly abstracted relationships are not invalid, but I do not think they serve to purpose of expressing meanings directly, in the way that representational compositions can.

It is not necessary that one work in the styles of Gericault, or David, or Delacroix… or to use their subject matter. The issue is more fundamental than that. The issue is… how can one construct visual meaning in representational work. That meaning can be narrative, it can be symbolic, it can be allegorical, it can be whatever… but whatever form of meaning one pursues, it must be presented inside of the space of the painting in terms of composed elements.

I was talking earlier of the need to establish a horizon line, for the simple purposes of keeping track of what is above and below the eye level. But I've noticed something more than that… that the horizon line is necessary for projecting a ground plane… and for coherently establishing major shapes upon that ground plane. Imagine Gericault's "Raft of the Medusa". There are figures above the horizon line and below it, and they are pitched and tilted in complex relations… all of which are themselves related to the plane of the raft, which has it's own perspective. The point of this example isn't to exalt Gericault's draftsmanship, but to point out that the composition (composed elements) require that VISUAL PERSPECTIVE be established and maintained consistently throughout the painting. Without such discipline, not only couldn't all the complex figures be rendered.. BUT MORE IMPORTANTLY… the composition could not be realized. A figure draftsman might be able to draw each of the figures individually, but to have them all composed together into a coherent space is a different issue.

This is my big point… that composition relies on perspective. We tend to think that perspective is simply a rudimentary drawing device… and it is. But it is also a compositional device, and it is a compositional device first. This is where drawing and composition blend together. To talk about drawing and composition as distinct things is something of an artificial distinction. The term composition is usually considered to be some highly abstract idea of the painting, whereas drawing is considered to be the way in which such a high abstraction is realized. But composition is really the arrangement of space and the visual elements in that space. It might begin in the imagination and in abstract notions, but it takes shapes by a consideration of the rectangular canvas, and of the projection of space within that rectangle. In other words, it takes shapes through the establishment of a visual perspective system that is unique to each painting. Such a system sets up and establishes the drawing, so much so that I would call it the meta-drawing. Once established, the drawing takes place within that compositional structure.

I'm sure this sounds very longwinded and tortured and hopelessly idealistic. But I don't think it is. It might be longwinded to explain, but the realization of these ideas is elegant in actual practice. How else can meaningful representational paintings be created?

Abstract Logic vs. Representational Logic

The topic (Abstract Logic vs. Representational Logic) arises when we design using the armature of the rectangle, and it has to do with how we perceive abstract shapes and relations, vs. how we perceive representational shapes and relations.

This issue came up as I was trying to compose a picture based on a very quick and sketchy line drawing I had done of four guys sitting on the beach in Barcelona. I liked the drawing a lot, and was trying to find a way to compose a painting from it. The difficulty was that it was so sketchy and linear, that I thought it wouldn't benefit from being painted. There was little form, value or color in the sketch. It seemed a poor candidate for a painting.

So I scanned the drawing and started playing with it in Illustrator. I wanted to find a way to place it within a rectangle that might bring out some qualities I thought were interesting. I was able to change the size of the drawing in relation to the overall rectangle, and to play with it's position. I was also able to superimpose an armature, and to note relations between the various major linear divisions of space and the drawing itself. So anyway, I settled on a composition that I felt good about.

But a funny thing happened when I actually started making the painting... these abstract design qualities that I thought I had zeroed in on, never came through in actual fact. It was obvious that the elements of the drawing were so sparse, that there was no interest in the image. I was hoping that the design of the placement and the scale of the space would make for an interesting abstract quality... but it didn't happen.

This failure revealed something very interesting... which is that there is a fundamental difference between the way we consider abstract things, and the way we consider "real" things. If you look at the images here, there are two that have the drawing in it, and two that have rectangles as abstract stand-ins for the figures. It's interesting to observe how symmetrically correct the figures are in the image beach-real-armature.jpg. But when you remove the armature (in beach-real-no-armature.jpg) the sense of the figures being locked into the armature disappears. Instead the figures just seem to be floating inside a lot of white-space.

This "floating in white space" was intentional... I wanted to compose them figures to highlight the isolation they had when I drew them. After all, they had been a small cluster of figures on a huge expanse of beach, with nothing around them. The analytical "composing" I did was to try and create some abstract design logic that underlay they simplicity of the drawing. But it never came through.

I think this is due to the fact that we just don't perceive representational objects as pure abstraction. If you look at the other pair of images (beach-abstract-armature and beach-abstract-no-armature), you can see that even with the armature removed, the abstract rectangles "read" as being composed geometrically. Abstract things can be read abstractly.

But "representational" things are read primarily as representations. It's as if we need to understand the meaning of the things represented... how they relate to each other, to the things around them.. to the entire "real" meaning of the space they're in. But abstract things have NO real relations, and so we ONLY see abstract relations... and that is the PRIMARY way we see them.

It makes me wonder about the nature of "abstract" composition.... that is... composition that uses geometry to divide space into linear and area relationships. I am convinced by the many examples I've seen that it must be true, that geometric placement matters. However, as my example above shows... it is not sufficient in the case of representation (and is the ONLY thing that matters with abstract content).

Usually the value of geometric composition is explained as some kind of "subconscious" force... a subconscious "sense" of formal placement. Even if we allow that this is true, and recognize that it is secondary to the "representational logic" of the things depicted.... it is still not clear what the relation between geometry and the things represented.

Again, geometric composition is seen as an underlying FORMAL concern... but I fear that this kind of categorization produces a representation/abstraction dichotomy... or a dichotomy between "representational logic" and "geometric logic". But I think that geometry has to be related to representation in some way. How? I'm not sure how to characterize the relationship... but I know it has to exist.

When I consider how I approached my drawing, I can see that I did NOT correspond the geometric design with the figures being represented... other than to place them within a large space in a symmetrical way. However, the representational logic of the figures isn't really about the space "around" the figures, so much as it is about the space "within" the grouping. This represents a disconnect between composition and the figures. And so on.