Heinrich Wölfflin once said, “not everything is possible in every period.” Ernst Gombrich takes this statement as a starting point in his book Art and Illusion (first published in 1960 but based on a lecture series from 1956) and asks: if artists always want to represent what they see as accurately as possible, why do their results look so different from period to period? In other words: why does art have a history?
Gombrich’s main explanation is the “tenacity of conventions”. There is an “enormous pull in man to repeat what he has learned”, and only “exceptional beings” are occasionally able to “break this spell and make a significant advance” (pp. 24-25). That’s why, according to Gombrich, art has slowly become to look more and more lifelike.
So where do comics fit into this model? Gombrich actually mentions them as part of the “victory and vulgarization of representational skills” (p. 8), which had become ubiquitous in the (i.e. Gombrich’s) present day. Is that really the end of the story? Do comics constitute the apex of artistic illusion? Probably not. I think comics retain some features which Gombrich assigns to earlier periods, or “primitive” art.
One of these features is the treatment of local colour. In earlier times, artists didn’t try to represent the colour of an object as they perceived it (not an easy task, as Gombrich explains), but instead chose colours that fit into a specific overall tonality, as was demanded by the taste of the public of that day (chapter I).
Another feature of primitive art is to show each object in its “characteristic shape” (p. 302) – a horse from the side, a coin from above, etc. -, rather than to employ a “purely visual mode” (p. 19) with foreshortenings and intersections. (In a way, this “visual mode” is similar to Svetlana Alpers’s “Albertian mode”.)
When I was reading the first Justice League Dark trade paperback the other day, I found traces of both of these “primitive” characteristics in it. Justice League Dark was (at that time, 2011) written by Peter Milligan, drawn by Mikel Janin, and coloured by Ulises Arreola. I’ll write more about this interesting series in a later post, but for today, let’s look at p. 9 of the third issue.
In this scene, the superhero Deadman tries to stop June Moone, who is possessed by an evil witch, to jump from a rooftop. In panels 1, 3 and 4, the colour of that rooftop is black. In the last panel, its colour has suddenly changed to a textured grey, even though the lighting remains unchanged. Which is the “right” colour of the roof surface? That’s hardly possible to say. If it’s made of concrete or a similar material, it could appear greyish in daylight, but the scene takes place at night, so it might just look black. The reason for the colour change is, I believe, a pragmatic one: in the last panel, the roof on which Deadman and June are standing (or falling from, respectively) has to be distinguishable from the other rooftops of the surrounding buildings, so one is grey (thus adding the benefit of allowing to show shadows) and the others are black. (Another instance of a colour change is the reddish background of the second panel, but then again, Justice League Dark is full of such eerie glows.)
As for the second feature, June Moone’s face (her “characteristic shape”) is always shown more or less frontally on this page, regardless of where her body is orientated towards. Often, the figures in this comic are reminiscent of amateur stage actors who are taught to always face the audience so that they can be heard better. Deadman is shown twice frontally and twice from behind, but his red-and-white shape is easily recognised from any point of view – unlike June Moone, the unexceptional supporting character. We need to see her face to instantly recognise her.
I’m not saying the art in Justice League Dark is bad. It’s just in accord with the current style in mainstream superhero comics, which values clarity higher than accurate representation of the artists’ perception. The readers need to be enabled to easily read the comic and identify the important elements on the pictures, in order to understand what’s going on in the story. The story, not the look, conveys the atmosphere. In such a paradigm, there is no place for ambiguity.
That being said, Art and Illusion is a rich and diverse book. For a different connection to comics, see Nicolas Labarre’s article “Art and Illusion in Blutch’s Mitchum” at The Comics Grid. Also of interest to the comics researcher might be chapter X of Art and Illusion, “The Experiment of Caricature”.
In 1983, the same year in which her book The Art of Describing came out, Svetlana Alpers published an interesting essay in the journal Representations, titled “Interpretation without Representation, or, the Viewing of Las Meninas”. In this article, she describes “two modes of representation that are central in Western art” (i.e. pictorial art, primarily figurative painting). Basically, these modes are specific constellations of perspectival, deictical and above all diegetic dispositions.
A picture in the first mode – the “Albertian” mode, after Leon Battista Alberti – “is conceived to be like a window on the perceived world”. That is, the recipient is aware that the picture was made by an artist, and from a particular point of view. Thus, the recipient in front of the picture takes on the same position as the artist when he or she beheld the scene that would become the subject of the picture. (In contrast, Alpers writes that “the artist positions himself on the viewer’s side of the picture surface”, but I’m describing things from the reception side here.) Cues for this mode can be artificial compositions of objects framed to be painted, such as a sitter posing for a portrait, or objects which appear to have related to the artist in some way, such as a figure seemingly looking at the beholder/artist.
The second mode of representation – “the northern or descriptive mode” – is characterised by the perceived absence of the artist. The picture looks as if “the world produces its own image”, a “replicative image”. Such a picture “does not assume the existence of viewers prior to and external to it”, but it may contain “a figure situated as a looker within” who does not appear to look out of the picture, or at least not towards the position of the artist/viewer. In a way, such a device is similar to Michael Fried‘s concept of absorption.
For Alpers, the choice of mode is a meaningful and almost ideological one. The Albertian mode betrays a “commanding attitude toward the world”, whereas in the descriptive mode, “the world seen has priority”. In other words, “the artist of the first kind claims that ‘I see the world’ while that of the second shows rather that the world is ‘being seen.'”
Does it make sense to look for these representational modes in multi-image works such as comics? It might, if we assess the mode for individual images and then extrapolate to determine the predominant mode of the work as a whole. In Alpers’s main example, Velázquez’s Las Meninas, the two modes are intertwined, but for simplicity’s sake let’s assume here that as a rule, either one mode or the other is employed in any given picture. I’ll contrast the two modes of representation with each other by using two older manga as examples, Barefoot Gen by Keiji Nakazawa, and Haine by Kyōta Kita and Keiko Ogata.
Let’s stark with the more obscure one, ハイネ: 愛と革命の詩人 / Haine: ai to kakumei no shijin (“Heine: poet of love and revolution”, published in German as Heine in Japan in 1988). It is the biography of the German poet Heinrich Heine (1797 – 1856), who actually had never been to Japan. In these two panels from p. 35, we see the young Heine in conversation with his aunt and uncle. The figures in the top panel look orthogonally out of the picture so that we, the readers, feel as if they were looking at us, and as if we’d look through Heine’s eyes. In the bottom panel, the arrangement of the figures makes us imagine we stand right next to Heine’s aunt, as if we were a fourth family member taking part in the conversation. Both panels are clearly composed in the Albertian mode of representation, as they give the impression of a plausible human point of view. This mode is predominant throughout the comic.
In contrast, consider these two panels from the beginning of the classic Barefoot Gen (はだしのゲン / Hadashi no Gen), on p. 3 of the German one-volume edition from 1982. The top panel shows running feet, seen from slightly above ground level. On the bottom panel, we see the Nakaoka family fleeing into an air raid shelter, depicted from a slanted view from above. What these two points of view have in common is that they can hardly be said to show the action from a human perspective: it is unlikely that a figure within the story could perceive things from the same position as we, the readers, do. Therefore, we can say the descriptive mode is employed here. Particularly the oblique and elevated point of view of the bottom panel is a frequent device in Barefoot Gen to freshen up traditional “shot/reverse shot” dialogue scenes.
Our analysis of the modes of representation shows that in each of the two comics, a mode is chosen that complements it: Barefoot Gen is based on its author’s personal experiences, but in order to show that the horrible events depicted in this comic are neither purely fictitious nor singular, the descriptive mode of representation is used to make them less personal and more credible. In Haine, on the other hand, the Albertian mode is used as an invigorating device that puts the reader into the action and almost turns this second-hand account into an eyewitness account, thus bridging the gap of 150 years between our time and Heine’s.
In his essay “Perspective as Symbolic Form” (“Die Perspektive als ‘symbolische Form'”, 1927), Erwin Panofsky dispels the myth that artists didn’t know anything about perspectival construction before the Renaissance. He shows that the Ancient Greeks and Romans just employed a different system than the vanishing point system we are used to today, and neither can be said to be more “correct” than the other. For Panofsky, the transition from the old to the new system was a paradigm shift. As long as a perspectival paradigm is upheld, artists will construct their pictures in that system. Once a paradigm shift occurs, there is no turning back to the old system – artists don’t choose between different systems.
If we follow Panofsky so far, one question remains: is the vanishing point system still the uncontested paradigm today, or has another shift occurred in the last 85 years? Let’s look at a random comic book to see how perspective is handled there. It’s been a while since I last reviewed Astonishing X-Men, and I’m going to properly review the current issues in a later post, but for today I pick issue #57 from December 2012 (cover dated February 2013). With this issue, Gabriel Hernandez Walta took over as the regular artist from Mike Perkins. It is a harsh transition, as their art styles are so different: none of the lines in Hernandez Walta’s art are exactly straight; they are all slighty irregular and seem nervous, vibrant and sketchy. And yet, see how he constructs perspective in the first panel on the second page:
If we trace the lines indicated by rows of windows or sidewalk seams (traced in red by myself here), which would be parallel to each other in real three-dimensional space, they converge in a single vanishing point when Hernandez Walta projects them onto the two-dimensional space of his panel. Furthermore, this vanishing point is at the same location in the picture as the head of the character Warbird, the protagonist of this story.
This comic book is full of such obtrusively constructed vanishing point perspectives, often including floor tiles or other grid patterns that help to convey a feeling of depth. However, this is not the only perspectival system employed by Hernandez Walta. In several panels, he switches to isometric projection. For instance, in the third panel on p. 3:
Parallel lines formed by the furniture in the depicted room (assuming the furniture is meant to be rectangular and arranged in parallel) stay parallel in the projection – they never converge even if we extend them beyond the panel borders.
Overall, the predominant perspectival construction system in Astonishing X-Men #57 is still the vanishing point system. But maybe Hernandez Walta’s little isometric deviations are a sign that the vanishing point paradigm isn’t quite as uncontested nowadays as it used to be.