Composition: Some Possible Approaches to Learning Pictorial Design
I'd like to start this post by thanking everyone who mentioned artists and posted composition resources in the comments on the last post, or by email. All the input has given me so much to think about that it's taken me some time to get my thoughts organised enough to follow up with this post. I hope this has been useful to everyone so far. For myself, I've reached some conclusions on how I'm going to approach the problem of getting better at pictorial design which I'll endeavour to explain a little at the end of this post.
Since we ended up with such an interesting list of resources, I'll start today by summarising them here so that we have them all in one place:
Some Composition Resources Suggested By Readers
Landscape painter Edward Seago was mentioned first by Nick as an example of good composition. Although I knew of Seago, I hadn't seen much of his work. A bit of trawling around the web has turned up some really beautiful compositions. This one in particular struck me for its very effective use of negative space:
Jon posted this page on composition on Bruce MacEvoy's Handprint site. Mostly it's about what Bruce calls 'format proportions' and is concerned with fitting compositions to the proportions of the rectangle that bounds the picture. To his credit, Bruce himself states that he's wary of forcing compositions to fit preconceived geometrical devices applied after the fact to old masters. He then proceeds to do it himself, but the results are worth having a good look at and are more convincing than many of the geometrical overlays I've seen. I do think there's some really useful information there and it's an interesting perspective on compositional design.
Loomis - Creative Illustration
Jesus mentioned Creative Illustration by Andrew Loomis which I'll be talking about a little more later in this post. You can download all his books in pdf format here, or you can buy a hard copy of Creative Illustration from Amazon by following the link above. It's incredibly expensive but is also incredibly useful and is unlikely to be republished any time soon. I think it has a lot of excellent information about design and composition, and a little that's not so good.
My good friend Peter Yesis emailed me this geometrical overlay of the Vermeer painting I posted previously. There's some other info at that link about the painting too. It's interesting that Vermeer appears to have broken one of the 'rules' of composition and put an important element, the balance, pretty much on the centre line of the painting.
Stapleton Kearns: Geometrical Composition
Stapleton Kearns blog was mentioned by Jon, and there are a few posts on composition from the geometrical design stand-point here (diagonal composition), here (a square within a rectangle, also treated on Bruce MacEvoy's page above) and here (circular composition.)
James Gurney: Eye Tracking ExperimentsJon also brought up the three posts on Jim Gurneys blog about eye tracking: here, here and here. These were perhaps the most interesting resources to come out of the discussion for me, and call into question the whole notion of leading the eye in compositions. In the course of a short email conversation I had with Jim about this, he had this to say:
"My own findings don't negate any of the traditional compositional practices of leading lines, golden section, etc. They still seem to play a role in making good and pleasing pictures, but what I question are factual assertions about how the eye flows or moves in a picture. The reality turns out to be a lot more interesting than any of us had thought."
Jim also mentioned the Yarbus eye tracking studies (English translation of Yarbus' book in pdf format and a research paper on Yarbus' findings.) This is really fascinating stuff for anyone who wants to delve further into eye tracking experiments and how we look at paintings. Stapleton then posted himself on eye paths, and that post has brought out some interesting comments too, including another insightful one by Mr. Gurney himself.
Composition: Arthur Wesley Dow
Margaret mentioned a book called Composition; a series of exercises in art structure for the use of students and teachers by Arthur Wesley Dow (it can also be downloaded in pdf format). I recently got a copy of this book and it's had more influence on the direction my thinking on composition has taken over the last few days than anything else. More on that anon.
Pierro Della Francesca: Geometrical Composition
Finally, David mentioned a geometrical overlay of the Baptism of Christ by Pierro Della Francesca, and made the very good point that there was more than just design behind the geometry. To Renaissance masters, geometry had philosophical connotations.
Well, that's quite a lot of stuff to look through. Taken together, those resources can provide some deep insights into how we can improve our own compositions.
Breaking It Down: Two Ways of Looking At Composition
After a lot of reading and having given this a lot of thought, I've come to the conclusion that these various perspectives on composition can be divided into two main areas, guiding the viewer and design, each of which can be broken down a little further:
1. Guiding the Viewer - The First Approach
I think this area of composition can be usefully divided into two areas, Focus and Eye Paths:
This was mentioned by a couple of people in the comments, not least by Ted Seth Jacobs. Ted is a world famous artist and educator who has published books on learning to draw and paint and taught many of the leading lights of the contemporary representational painting revival in the US, Jacob Collins and Anthony Ryder among them. And his work is exceptional.
There are various ways that focus can be created in a picture. Ted lists placement, contrast in either value or colour and distinctness of shape. I'd perhaps add edge handling to that list too as a technique that's often recommended for creating focus.
Interestingly though, one of the findings of Jim Gurney's eye tracking experiments was that the viewer's eye will tend to be drawn by elements which are psychologically interesting, regardless of what the artist tries to do. In many cases these elements will be standard things, figures and faces. But it also strikes me that we all have very different life experiences and may perhaps find different things more psychologically compelling.
I think the conclusion here, and Jim Gurney appears to concur in his posts, is that the traditional techniques of guiding focus will work best when they are used to reinforce psychologically interesting elements. Among these elements, Jim lists objects that we can physically interact with, door handles, roads, food perhaps. So still life and landscape artists would do well to consider this too.
A relevant example here I think is L'Eminence Grise by Jean-Leon Gerome:
I've seen this painting mentioned as a good example of an artist directing the gaze of the viewer towards the lone figure descending the stairs. It's interesting that the areas of greatest contrast and colour are elsewhere, and many of the diagonals within the painting converge on the grouping of three figures towards the centre of the painting. But attention still seems to fall eventually on the solitary enigmatic figure on the right.
You could of course come up with any number of explanations for why that happens. But is it not simply that most of the other figures in the painting are looking towards the main focus of interest? In his posts on eye tracking, Jim Gurney found that viewers would first scan around a painting to find the context of the story. In that case, the other figures will be registered as directing their attention in one direction. The focus in this painting works because of the narrative of the story (perhaps in conjunction with the title,) and in opposition to the usual compositional devices. It would be very interesting to see an eye tracking experiment on this painting I think.
This is the area that Jim Gurney's experiments really call into question. First let's look at an example of how this is often covered in books on composition. I think Loomis has an awful lot of useful information to offer artists, but the example comes from his book Creative Illustration. There's much more on this subject in the book, but this is a typical example:
This is a good time to bring in a couple of excerpts from the Yarbus experiments I mentioned earlier. The experiment involved a number of test subjects viewing Unexpected Visitor by Repin whilst having their eye movements tracked. This is the painting:
Have a look at that painting and try to be conscious of the movements of your eyes. Look also for where we might be able to say that Repin had used compositional effects to lead the eye through the composition as Loomis recommends we do.
Now here's the results of the eye tracking of several viewers, looking freely at the painting:
I think that speaks for itself. What's particularly interesting here (and this was also borne out by Jim Gurney's experiments) is that the eye activity is markedly different for each viewer.
What's also interesting is that if a viewer is given a particular task regarding the picture, so changing the context in which they view it, the paths their eyes take through the painting will change:
I infer from this that the psychological baggage we personally bring to a painting will affect the paths our eyes take despite the best efforts of the artist to guide them.
I think it's safe to say at this point that 'eye pathways' in paintings are probably a myth and that if we want to make good compositions we'd be better off directing our energies elsewhere. It would be very interesting though, to see experiments like this done on paintings with a wider range of subjects without figures, and also with abstract work. To the best of my knowledge, this hasn't been done as yet.
In terms of guiding the viewer's attention, there is probably more mileage in the various ways of establishing focal areas in a picture than there is in eye paths, but I think we need to be aware that perhaps even these techniques are not as effective as we might have thought, and that if they don't correspond with elements that a viewer wants to look at they will be probably be ignored, or at least be of secondary importance.
2: Design - The Second Approach
I'd propose that the second main area of composition is concerned with the effect of the overall design. Under this heading would come balance, repetition, spacing and the use of negative space and other basic principles of design. I'm inclined to give this area much more credence and think that this is where we should be directing our energies. It's where I intend to direct mine.
In general, there seem to be two main approaches to this, geometrical and intuitive. I don't think that these two approaches are mutually exclusive and in fact they may work very well in concert with each other. But people tend to fall into one camp or the other I think.
There are many links above that point to the geometrical approach to design. It can be fairly informal as Stapleton Kearns describes its use on his blog, or quite mathematical. The golden section is perhaps the most popular incarnation of the geometrical approach.
There's a detailed treatment on the geometrical approach to pictorial composition in Classical Painting Atelier: A Contemporary Guide to Traditional Studio Practice by Juliet Aristides. It's a fascinating chapter, one of the most interesting in the book. Aristides is firmly in the geometrical planning camp. Whilst recognising the value of intuition, Aristides believes that an intuitive approach alone is not enough to make good compositions and relegates it to personal style, stating the following:
"Artistic intuition and sensitivity to order are vital elements that contribute to the style of an artist, yet these elements alone are not enough if one is to achieve a consistent level of compositional mastery"
Personally, I think that's a somewhat limited way of looking at composition. If we're only talking about the western art tradition and that only previous to the 20th century (which perhaps she is, being a Classical Realist) then the statement makes more sense.
But I think it ignores in particular the pervading influence of eastern art on western artists which began in the late 1800's and is still being felt today. Chinese and Japanese art both represent rich, vibrant - and very long - traditions which I think are particularly strong in the area of composition, and it's this that has influenced western artists the most. Because of that, I have a problem with absolute statements of fact like the Aristides quote above. I think the picture is more nuanced.
I do think the Aristides book is excellent though, I'd recommend it, and the chapter on composition is extremely interesting. I don't think anyone could seriously doubt that many Renaissance artists used geometry in their composition, and Aristides gives some excellent examples of quite convincing geometrical overlays. I find some of the examples in the book much less convincing though, and it seems that if an important element of the picture is just close to a line or point of geometric divisional harmony, that's taken as proof positive that the artist in question based the composition on the provided geometrical overlay. Correlation isn't necessarily causation, especially when it's just somewhere near and not spot on.
To my mind, conclusive proof of a particular mathematical design being used in a particular painting would be a preparatory study by the artist showing how it was worked out. Without such supporting evidence I think we're firmly into the realm of conjecture and should employ liberal amounts of circumspection.
All that said, I thought it would be interesting to try out some of the ideas in the book in a still life painting. This was the result:
The composition for this painting was based on the geometry on the right, taken directly from Aristide's book. The proportion of the picture is built on two golden section rectangles, one above the other, highlighted in yellow. I've placed the centre of three of the peaches right on what are supposed to be harmonic intervals, circled in red. In addition, the diagonal line formed by the top edge of the jug follows one of the main diagonal divisions, which also bisects a plum on the far right, and the handle joins the body of the jug at the intersection with another of the main diagonals.
Of course, one clumsy experiment like this isn't proof of the effectiveness or otherwise of this approach, and I don't take it as such. I do feel that there's something nice about the composition. Although the objects appear fairly randomly placed if you haven't seen the geometry, the way they line up does seem to satisfy some kind of inner preference we have for order.
But the painting is still a bit of a dud if you ask me. As you can see I didn't finish it because I thought it was too rubbish, and having invested days and days worth of work working out the composition, doing thumb nails of the value balance, etc. I became disheartened at the way it was going and abandoned it as an interesting experiment that didn't quite come off.
The (tentative) conclusion I drew from this experiment is pretty much the opposite to Juliet Aristides' statement quoted above. It's this: that hanging a composition on geometrical design can be very effective, but unless it's combined with an intuitive sense of design, colour and good spacing, it's not enough on it's own to make a good picture.
Which brings me nicely to the intuitive approach to design.
Let me clarify a little what I mean by intuitive. I feel I have to since the word is used pejoratively by detractors, and sometimes indiscriminately by supporters.
I'm sure we've all had an 'aha' moment now and again, when something which seemed difficult to understand suddenly seems to fit together. Perhaps you've had the experience of going to bed worrying about a problem, and found that when you woke up the next day you seemed to have found a solution.
These are quite common experiences. What I think is happening there (and this is based on a fair amount of reading on recent findings in neuroscience) is that our brains process much more information unconsciously than they do consciously. I don't think intuitions are a gift from the Gods, I don't think there's anything mystical about them. I don't think that they're necessarily irrational either. They seem to be the results of processing carried out by our brains outside of normal every day consciousness. Our brains process information in a number of different ways, and we're not always aware of it happening. Anyone interested in further reading about this could do a lot worse than The Developing Mind by Daniel Siegel.
It seems to me that if we want to develop our intuitive sense of design then we need to feed our brains with a lot of raw material for them to chew over whilst we get on with other things, the washing up and stuff like that. When we process a new piece of information, or practice a new skill, our brains create representations of the information or action. These representations are built from links between neurons, called axons. We develop these links throughout our lives, and the more often that particular patterns of connections are 'fired,' the stronger they become. The phrase used by neuroscientists to describe this process is "neurons that fire together, wire together."
These connections are physical, not metaphorical. Learning new skills and processing new information actually physically changes the structure of the brain, a process known as neuroplasticity, and with repeated use these connections eventually become coated in a substance that makes them even more permanent and further facilitates their firing - they become almost 'hard wired.' Thus are habits formed, and this is also why they're so difficult to break. If we stop using the connections, they die away after a while. Use it or lose it. A better way to say "you are what you do" is to say "you are what you repeatedly do."
So it follows that we ought to be able to develop and strengthen the areas of our brains that are used in design through the right kind of repetitive use. The more we practice, the more natural and effortless this skill will become, and the more we'll be able to use it without having to consciously think about it. It will become intuitive.
I see a practical application of that in the approach put forward by Arthur Wesley Dow in his book Composition, subtitled A Series of Exercises in Art Structure For The Use of Students and Teachers. This is the book that Margaret mentioned in the 'Resources' section above, it's been re-published with a couple of different titles.
First, a few examples of Dow's own compositions from various stages in his career:
The influence of Japanese art can be felt pretty strongly there, especially in the later wood block prints. But the one thing that fairly shines out to me is how lovely the compositions are, how strong the design is in each piece.
Dow's approach to composition is in stark contrast to the planned geometrical approach. In fact, in Composition he comes out quite strongly against it:
"The secret of spacing in Greek art has been looked for in the golden mean...but the finest things were certainly the product of feeling and trained judgement, not of mathematics. Art resists everything that interferes with free choice and personal decision."
The primary idea of Dow's book is that composition is best learned through practice. The book is based around a series of practical exercises split into three sections: line (and spacing,) notan (a Japanese word for dark and light) and colour. Creative Illustration by Loomis has a similar structure. In the Dow book, each series of exercises is intended to be expanded on by the student and repetition, refinement, adjustment and invention are key themes. His overarching concern is the development of an intuitive feeling for proportion and spacing, which Dow expects to develop naturally thorough the course of the exercises.
This approach needs to be seen in historical context I think. He reflects the concerns of the Aesthetic movement of the late 1800s which stressed intuition, abstract design and beauty over the story telling and three dimensional modelling of much Victorian art of the time. Like Whistler and other artists of the Aesthetic movement, Dow was very interested in eastern art, in particular Japanese painting and woodblock printing. The contemporary fashion for 'Japonisme' was also present in the work of the impressionists, post impressionists and symbolists, crossing stylistic, philosophical and cultural boundaries.
I do believe that it would be quite possible to develop this natural sense of design through the course of normal practice, just by drawing and painting without concentrating on it particularly. But I also believe that concentration on a particular aspect of painting and drawing, especially through repetitive practice, allows us to develop that particular aspect more fully and perhaps in less time too.
I think we all have to find the approaches that suit us best, there is no 'one size fits all.' Dow's approach may not be for everyone, but to me it seems an eminently sensible place to start developing an intuitive feel for design and spacing.
The next post on composition will describe the exercises I intend to follow, based on a combination of Dow's exercises and some of what seem to me to be the more useful and practical parts of Loomis' approach. As always, I'll post the results as I go along. If my sense of design improves over time, it should be obvious enough in the work I produce. If not, that should be obvious too!
I don't intend to completely ignore the geometrical approach, however. I think it's important to work with something for a while before any real conclusions can be drawn about whether it suits our way of working or not. Perhaps, with enough practice, a practical balance between the two can be found.
This has been a very long post and has taken some time to write, in fact I've had to re-write it several times. It couldn't have happened without the input that people gave on the last post and a couple of email conversations that it sparked off. I've tried to present everything as clearly as I can, so that it'll be the most useful. It's certainly helped me to get my own thoughts about composition much straighter in my head. I hope that as we go along you'll all benefit from this exploration as much as I have already.
Posted 23rd December 2011