What is it about a likeness that delivers instant recognition? There can be no quick or easy answer to this question, and portrait artists frequently devote an inordinate amount of effort to making small adjustments to their work in pursuit of this fugitive ideal. Sometimes the accumulation of these adjustments has an effect directly opposite to that sought, and the image seems to move further and further away from a true likeness – as often as not because the focus has been on successive corrections carried out within an overall framework that will never stand up to scrutiny because the original structure lacks anatomical accuracy. My usual way of avoiding this cul-de-sac is to front-load effort into initial drawing so that subsequent adjustments do not have unwanted knock-on effects on the relative positioning of other features. Richard #1 looks like a casual graphite drawing, but it is the product of an obsessive process of measuring and re-measuring. I won’t even think about using paint until I am confident that face and hands have been set up accurately in a way that won’t cause me problems further on.
I have started this piece with an experiment: darkening the corners of the canvas with acrylic aerosol colours, even prior to the initial graphite drawing. It is a purely intuitive move, and I’m not entirely sure why I’ve done it – perhaps to play a little psychological trick on myself, enforcing a stronger focus on the initial drawing of the face and the stages immediately following it. Richard #2 is a straightforward continuation of the drawing process, using paint to work over the graphite, fixing and developing the image in a more definite way. The phrase ‘elusive geometry of recognition’ comes from an excellent little book ‘Painting Portraits’ by Anthony Connolly. Unusually for an artist of his quality, Connolly is a talented writer. At this stage I really am wrestling with the elusive geometry of recognition, picking out the elements of Richard’s facial features that seem to define his character, and trying to give them enough emphasis to capture his individuality. Take this process too far and the portrait lapses into caricature.
Time to start work on the flesh tones now. For Richard #3 I am using a palette of Van Dyck brown, Venetian red, Naples yellow, pale yellow ochre, rose dore, and flake white. As I make small changes it becomes apparent that the area of the mouth is most important in conveying Richard’s personality. His face settles naturally into a quizzical half-smile, as though there is a private joke that he is permanently on the verge of sharing with us. Keeping hold of this impression will be an essential aspect of establishing and retaining the likeness. I should point out here that I don’t really know how objective this assessment is. I have held the view for some time that it is possible to learn most about another person by looking at their mouth, particularly the way that the tiny muscles that control facial expression have shaped this area of the face, etching an emotional history into its lines, depressions and extrusions. Is this the way things really are, or merely my own way of weighing up other people? Whatever the truth, I am convinced that this plays a crucial role in the elusive geometry of recognition.
Every now and then an artist will see a potential subject that just demands to be painted. For me, Paul Dickson came into that category as soon as I saw him. The term ‘full of character’ is something of a cliché, but one that perfectly describes his air of dishevelled bonhomie. If Ivan Kramskoi had received a commission to paint Falstaff, then Paul, I imagined, would have provided the ideal model.
Perhaps influenced by the informality of Paul’s demeanour, I decide to move away from the careful preparatory drawing I have employed for the first three portraits, and start with a much looser approach. The canvas has been prepared with a bright burnt sienna ground, and I work quickly with thin washes of flake white and burnt umber to establish broad areas of light and dark. Next I scrub in a thin background using a mixture of raw umber and earth green. Finally I draw in some of the darkest areas with a less diluted mixture of raw umber and lamp black. At this stage it is simply a rough sketch, and it is more important to convey a sense of vitality than to think in terms of accuracy. There is enough of a likeness in Paul #1 to leave me feeling that I can work on from here.
The next session starts with attention to the flesh tones. I use a thin glaze of Naples yellow and flake white for the highlights, and various mixtures of dark alizarin crimson and burnt umber for the shadow areas. The crimson comes from a box of Ferrario Van Dyck oils that somebody gave me many years ago. Ferrario call it ‘rose madder deep’. These oils have a beautiful consistency and I recommend trying them. I use the same dark crimson and burnt umber for the t-shirt, blending with fan brushes. I dig out some Ferrario sepia for the coat lining, and add a touch of it to some flake white for the woolly outer layer. I darken the green background and take this opportunity to reshape the outline of the hair. Using pale yellow ochre thinned with Sansodor I put some highlights into the beard, and then redraw the eyes and some shadow areas with burnt umber. At this stage it is starting to look like Paul…
Paul #2 needs a couple of days before I can continue, and I spend quite a bit of time in the studio watching paint dry and pondering an issue: it’s the dishevelment question. My overall aim is to get away from conventional images in the public mind of ‘the homeless’, and to represent my subjects in a more aspirational light – one that is much closer to the actuality of what St. Mungo’s Broadway achieves in terms of rescuing its clients from the exigencies of being on the street and sleeping rough. But Paul’s aura is resolutely chaotic and I can’t imagine conveying his character in a cleaned up form. Eventually I come back to thinking about Kramskoi, and how he would have sought to elevate his subject’s humanity by providing a psychological insight, representing a depth of character independent of the superficial elements of physical appearance. With this in mind the adjustments in Paul #3 are minor, but directed towards depicting somebody whose innate sense of humour provides a strong barrier against the damage that life can do.
Starting work again I concentrate at first on the eyes and mouth. I need to make subtle changes that convey a benign sense of mischief. This isn’t something that can be done in a single step, but rather acts as a guiding principle as I proceed. There really is an element of trial and error in this, and with Paul #4 I feel that I am having more success with the eyes than with the mouth. I decide to transfer attention to the beard and moustache, bringing in some highlights, and this is helpful in lightening the mood of the piece. Intuition takes a hand here: something tells me that the flat background of Paul #3 is an obstacle of sorts. I need to focus the viewer’s attention on the minutiae of the facial features. I thin some lamp black with Liquin and then rub it gently into the upper corners with a soft cotton cloth, grading it down towards the hair leaving the illusion of a halo of light around my subject’s head. Satisfied with this effect, I decide that it is time to start thinking about what to do with the hair…
I start by sketching in the hair with burnt and raw umber, mixing in flake white and pale yellow ochre for the highlights. The immediate effect of this is to change the apparent shape of the face, giving a much stronger likeness. I am referring to a number of photographs, and this improvement, checked out rather obsessively, I must admit, increases my confidence level to a marked degree. I work some darker areas into the beard with rigger brushes and then work on the eyes, trying to get a mischievous twinkle into them. I will have to correct their asymmetry, a legacy of the original drawing at Paul #1, but I defer this because I suspect that complete removal will undermine what I have just done. Intuition kicks in again and, after allowing the surface to dry, I decide to apply a glaze of Van Dyck brown (known as ‘Cassel earth’ in the Ferrario range) to the entire painting. I thin the paint with Roberson satin glaze medium, and then rub it into the surface firmly with a soft cotton cloth. I finish this stage by taking a clean cloth and wiping this glaze out of the highlights so that it only remains in the depressions in the canvas in these areas.
I’m getting fairly excited now as the painting starts to take on the look of a 19th century Russian piece. This impression is reinforced by the slight halo effect around the head, reminiscent as it is of traditional icon painting. Somehow that seems to fit the atmosphere around Paul quite perfectly. I get a little more symmetry into the eyes and then brighten the highlights. The shadow on the nose is adjusted, and then I work more highlights into the beard and hair. At this point I start to wonder whether the piece might be finished, and decide to put it aside for a while to think about it. There may be more I can do, but until I am sure exactly what, I don’t want to touch this piece again.
The first thing I noticed about John was his excellent profile. I could easily imagine his image on a Roman coin, so I took a series of photographs, deliberately posing him in an attitude that echoed this thought. I drew a profile on the canvas in graphite, and then decided to get the background in straight away. I just wanted to be able to see that silhouette and how it might work against its background before making any other decisions. I am still working on two other pieces, and the relatively easy task of setting up John #1 gives me some valuable thinking time. I find that working on a single piece can result in a self-imposed pressure to make progress. With several pieces in progress I can allow pure intuition to guide me as to what to work on at any particular point in time.
After some further graphite drawing I get started on the head. The initial flesh tones are rather exaggerated. I am using a basic mixture of Venetian red and burnt sienna. Highlights and shadows are achieved with flake white and burnt sienna blended into this mixture in varying amounts to give the required tones. I also work a thin green glaze into the background. This brings the head forward and creates a feeling of depth, even at this early stage. This initial modelling of the features is taken fairly slowly and seems to go well, giving me a degree of confidence that I’m on-track. I can already visualise the final piece, and, importantly, the path to completion. I need to play these little psychological games with myself – there is nothing worse than standing in the studio confronted by a series of canvases and not knowing what to do next.
One of my favourite quotations is from the Italian artist Francesco Clemente: “I paint because I have to be astonished, and there aren’t things in the world that astonish me enough.” This comes close to describing the sense of surprised excitement I sometimes feel as a portrait begins to take shape and have a life of its own. I start to under-paint the hair and darken the back of the neck with raw and burnt umber, then lighten the flesh tones and the background with thin velaturas of flake white and Naples yellow. When I return to the studio the next day and look at John #3 I am surprised by the way the character of the piece has changed, and excited to find that I am seeing skin where I had previously seen only paint. This motivates me to carry straight on with the piece, driven by the sense of astonishment that Clemente described.
Working quickly and rather loosely, I darken the hair further with a mixture of lamp black and raw umber, trying to get a little more form into this area. I then switch to the t-shirt, carefully modelling the shadows and folds with thin glazes of Payne’s grey, and then, laying in flake white for the highlights. I use a small sable fan brush to blend these areas together and get the fall of the light looking as natural as possible. Taking stock of the overall effect, I decide to go back into the hair with a mixture of Payne’s grey and flake white. This unifies the painting somewhat by keeping a similar ‘temperature’ to the shadows in both the hair and the t-shirt. I think this cooling down is necessary to offset the subject’s rather florid complexion, an impression that I want to preserve.
First, I make some adjustments to the modelling of the ear. Then, for the next stage of the hair I make a thin mixture of flake white, a touch of Payne’s grey, Sansodor and a small amount of Liquin. I use sable rigger brushes to apply the paint with thin strokes in different directions, building up an impression of hair that hasn’t been too rigorously combed. The process requires a degree of discipline: I have to make sure the brush is not too heavily loaded in order to prevent the initial strokes being too thick; these strokes give a bright effect that is suitable for highlights, but because the mixture is thin, the final strokes from each loading are translucent and more suitable for the darker areas, particularly those at the back of the head. I concentrate on getting the overall tonal balance of the hair right rather than achieving a finished effect.
At this final stage I need to make a series of important adjustments. The background has to be rebalanced to fit with the direction of the principal light source to the left of the canvas. Then, using the rigger brushes again, I have to turn up the contrast of the hair: highlights in pure flake white, and shadows with pure Payne’s grey thinned with just enough Sansodor to run freely from the brush. Finally, I decide that the skin tone is too red – a judgment that only becomes clear when the other parts of the painting are complete. I make up a very thin glaze of Naples yellow and raw Sienna with stand oil and fine detail Liquin (quite runny compared with the gelatinous composition of the standard product), and float this over the skin areas with fan brushes. Clemente’s words come back to me again as I decide that the portrait is completed.
A few days after starting the portrait of Gordon Chaston I began this one so that I could toggle between the two while waiting for glazes to dry. As a personality, David represented a rather different challenge. At first I was intrigued by his resemblance to the philosopher, Ludwig Wittgenstein: the same bone structure and shape of face, the same bushy hair creating an atmosphere of unruly charm. Then I started to notice a deeper connection. Something in the eyes, a fey quality suggesting that they each inhabit a private world that overlaps with our own only partially, and rather superficially at that. There is a boyish quality about them both, one that stands in contrast to the deeply lined faces and creates an ambiguity about their age, adding to that fey quality and leaving judgment suspended as to character traits and prevailing mood.
With the same burnt Sienna ground that I most frequently use, I start the initial drawing. I don’t refer directly to photographs of Wittgenstein, but his appearance is well-known to me, and stays at the back of my mind as a guiding impression. If I feel that my portrait of David loses its resemblance to that impression then I know it will be time to call a halt and review progress. The initial drawing with its exaggerated highlights has a ghostly quality that makes me reluctant to move on. David #1 stays on an easel untouched in a corner of the studio for over a week. There is something about it that I don’t want to change or disturb. My memories of studying Wittgenstein’s work have their effect, encouraging me to adopt a more analytical approach than I might usually do at this stage – no bad thing…
Eventually I come to a decision, but not before starting initial work on a third portrait, one with a very different feel and approach. My first step with David #2 is to push in that very pale background. This helps me to get my highlights into perspective and start to lose some of them. I start this process off by glazing some raw sienna into several areas, mainly on the left side of the face. I start building up the flesh tones with a series of thin glazes, some tending towards pink and others toward yellow. As I go along I constantly re-draw the shadow/dark areas of the face with that 50/50 mix of burnt and raw umber that I favour. These small adjustments are made to take me closer to the likeness I want to achieve, but also to build the right facial expression.
I need to move away from the dead grey tones of David #2, so the process of glazing in the flesh tones goes on as I try to breathe more life into the skin. After some detailed work in local areas I adjust the mouth slightly and then apply an overall glaze of burnt sienna, cadmium red and flake white. This is thinned liberally with stand oil and turpentine in order to offset the opacity of the flake white. Conventional advice is always to avoid the use of white in glazing, but this a rule that I regularly ignore when I need to achieve lighter tones. Finally, I start to work on the eyes because I have reached a stage where I need to have them better defined in order to judge the facial expression. The irises are a little too small at the first pass, but the steely stare feels right so I take some time out now to assess things.
After a few days reflection I start work again, emphasising the highlights. I then turn my attention to some more subtle adjustments in the mouth area. John Singer Sargent reportedly described a portrait as “a likeness in which there is something wrong about the mouth.” My theory is that the older the subject the more the mouth tells us about their character and personality, and that this is because, over time, the muscles around the mouth create lines and shadows that reveal prevailing moods. This makes the mouth and the area around it the most important feature in describing the sitter’s psychology. At David #4 I am starting to try to restore the slight suggestion of an enigmatic, reflective smile that I believe is important in conveying the fey quality that I spoke of earlier. This was lost somewhere between David #2 and David #3 and I know that I have to get it back.
At this stage I am fortunate to encounter David again in a social setting and I have the chance to observe him interacting with other people and then sitting quietly on his own. This is far better for me than any formal sitting, and it increases my confidence in the progress I have made thus far. I decide that there is far too much of the bright burnt sienna ground in view so my next step is to work up the jacket and shirt. Further glazing to lighten the flesh tones then feels easier to judge, and I continue the very slight adjustments to the mouth. I am proceeding cautiously now, using a variety of photographs as references, because I am comfortable with the likeness and rather afraid of compromising it by trying anything too radical. To be truthful, I feel as if I am walking on eggshells…
I darken the blue of the jacket and re-model the shirt before starting on the hair. This is accomplished in several discrete stages, gradually building up a structured complexity as I go along. Between these stages I re-work the flesh tones, highlights and shadows, finding them much easier to judge with the bushy mass of steely grey hair starting to form. For me this is an entirely intuitive process, going on in painfully small steps over a three to four week period. Every little success engenders a nervousness that the next step might be the false one that ruins everything. I reach the point where I believe the piece is finished, and yet I am dissatisfied. I am working on other pieces, but constantly returning to this one – just looking at it and trying to analyse the source of my own discontent. Eventually I decide that the problem isn’t a local one pertaining to a particular feature, but rather an overall one: the complexion is slightly too sallow, too yellow. I decide, with some trepidation, to attempt a final thin pink glaze, with a dash of crimson in it. In just a few minutes the sallowness dissolves, the flesh comes to life, and, unexpectedly, all is unified.
I am really happy with this piece now, but I cannot articulate precisely why. Notoriously, Wittgenstein said “Whereof one cannot speak, thereof one must be silent.” Perhaps he should have tried painting.
In mid-2014 a number of things came together. I had built a new studio at the end of my garden and I had a new project to work on as artist-in-residence at the St. Mungo’s Broadway Trust: creating portraits of the trust’s clients, past and present, for an exhibition and charity auction. The new studio was important because I find it difficult to work systematically in oils without a dedicated space, and the project called for systematic working methods, being able to have several pieces in progress at any one time. I had been creating portraits of homeless people for several years, working in graphite, acrylic ink and watercolour. It was time for something a little more ambitious.The choice of the first portrait subject was a no-brainer: Gordon Chaston, a former client, was one of the first people I met at St. Mungo’s. A real character by any reckoning, he helped me recruit subjects and offered to write commentaries on them or do interviews for the exhibition. Thanks to Gordon I was off to a flying start.
I started off with a simple drawing on a burnt sienna ground, using raw and burnt umber with a thin flake white wash for the highlights. The thin wash of burnt Sienna seemed a little fierce so I pushed the pale yellow/green background in early on. I took my time getting this initial drawing right, concentrating on the relative positioning of the key features. I did this mainly in order to give myself a level of confidence around being on the right track. A friend stopped by and remarked on the fact that all the life drawing I had done in recent years seemed to be paying off. I hadn’t thought about it in those terms, but I suppose he was right. At this point it isn’t a painting, but rather a drawing done with oil paint. Having said that, I find the distinction between painting and drawing rather arbitrary, and one that I try to ignore as far as possible.
The next stage (Gordon #2) is the gradual laying in of a series of glazes and washes to build up a complexity of flesh tones. The first stage of my drawing had been done with a soft graphite pencil and I had forgotten to fix this before applying any paint to the canvas. As a result my washes picked up some of the graphite and gave a darker effect than I had intended. Not all of the consequences are negative, but I make a mental note not to repeat this error. Quite a bit of the burnt sienna ground is left in place as a mid-tone, allowing the warmth of its glow to show through subsequent layers of paint. The main concern at this stage is an overall tonal relationship, together with a gradual refinement of the features. As I go along I refine the drawing with the placing and adjustment of shadows using raw umber.
Now I’ve worked into the background to bring it a little closer to the shade of green that I want and made an important adjustment to the flesh tones by glazing with caput mortuum. This isn’t an easy pigment to get hold of and I get mine in powdered form from Cornelissen, mixing it with stand oil and then a little Liquin to thin the glaze. The effect is an unusual one at this point with an almost metallic quality, but the rationale for it becomes clearer in subsequent stages as I use a further series of thin glazes that are much closer to natural skin tones. I am still refining the drawing and work on some of the shadow areas with burnt umber while the glaze is still wet. This is a delicate process and I tend to use much smaller soft sable brushes than most artists might usually choose.
Lots to think about now as it starts to become clear to me where the inspiration for this piece came from. I had taken dozens of photographs of Gordon, but I now know why I settled on this particular pose and expression. It had reminded me, subconsciously, of the work of Semyon Chuikov (1902-80), a Russian Socialist Realist painter whose work I first saw in the State Tretyakov Gallery, Moscow in 1993. Chuikov’s most famous work, The daughter of Soviet Kirgizia, was the subject of a well-known Soviet postage stamp. In a move typical of the Socialist Realist school, the slight upward angle and the lighting create a heroic image out of an otherwise mundane subject. Gordon is a politically sophisticated individual with strongly egalitarian views. Somehow, I imagined him as a political commissar, painted by Chuikov in the 1950s…
With this now at the back of my mind I know what I need to do with the light to make the portrait work. I work briefly on the shirt using a mixture of cobalt blue and Prussian blue, and then concentrate on a series of thin glazes to build up the flesh tones into a more naturalistic look. I’m using burnt umber, burnt sienna, Naples yellow, cadmium red and flake white in various mixtures to achieve this. Between these glazes I redefine the darkest shadow areas with raw umber to retain the tonal balance. The application of the previous glaze of caput mortuum has acted as a warm grey underpainting for this stage. I note that as the relative strength of the highlights changes, so does the facial expression. It becomes important to stand back from the canvas periodically in order to retain control of this effect.
I now have a clear direction in mind, and start to feel as though I’m on the home run. After further work on the background to make the green less chemical and more organic, I work on the hair with a simple palette of lamp black, raw umber and flake white. Then I lighten the blue of the shirt, improving the modelling of the fabric, while changing the emphasis of the highlights so that they accord with what I have in mind for the lighting of the face and neck. The first step here is a thin glaze of raw sienna with a touch of flake white. Before this glaze dries I work into it with touches of flake white for the highlights (including a slight, but important, widening of the upper part of the nose) and a 50/50 mix of raw and burnt umber for the shadows.
All that remains is a series of final adjustments, particularly to highlights which need to be emphasised. Gordon comes to the studio for a final sitting, but there is less to do than I imagined. I am keen to ensure that I get some flavour of penetrating intelligence into his gaze – this is, after all, a man who teaches courses in astronomy and cosmology at the St. Mungo’s Recovery College (a unique and remarkable institution of which more in future posts). Finally, I decide that the piece is finished. Of course, there is more that I could do, but I now have the feeling that I’m just as likely to detract from the work as I am to improve it – a sure sign that it is time to stop.