Lecture: Depicting Three Dimensions in Mediaeval Illuminations.

 

Depicting three dimensionality in mediaeval manuscripts (with a brief foray into the fourth dimension!)

This talk was given at Virtual Raglan 2020, which, due to COVID, was an online event.  It was presented a Zoom presentation with Powerpoint slides.

 

 

Slide 1 Introduction: The issue: we live in a three-dimensional world, and illuminated manuscripts, like most other painted media, are two-dimensional.  The aim of this talk is to see how (and indeed, whether) mediaeval manuscript artists depicted the three-dimensional world in their work.  Had this talk been presented at “non-virtual” event, my plan was to follow it up with another, more practical session in which I would teach the techniques and give participants a chance to practice them.  That can’t happen today, but maybe will at some future time.

 I am restricting this talk to the manuscript art of Western Europe, the cultural sphere of Catholic Christendom, from what the SCA generally reckons as its core period of 600 – 1600 AD.  That will give me more than enough to work on, but I don’t doubt the same study could be done with other cultures and periods, such as Japanese, Chinese, Persian, and Orthodox Christendom.  And although I am looking at manuscript illumination, the basics of what I will say will often apply to other forms of painted or drawn art and ornamentation in the period.

 

One implication of studying this culture in this period, is that one gets a pretty linear progression in terms of understanding of, and application of, the techniques that produce the illusion of three-dimensionality.  That would not be true if we extended our period backwards, to the Roman era, when such techniques were applied with considerable sophistication, or forwards to the twentieth century, where (probably as a reaction to the camera, which can do the job a lot better) many artists ceased to consider representation of three-dimensionality to be something they were bothered about doing.  But for our limited period, we will see a progression from unconcern with, or unawareness of, how to depict three dimensionality, to considerable sophistication in technique.

But before I start on the art-history bit, I want to do the science bit.  We need to think about what the problem is, before we can think about the solutions that mediaeval artists came up with.

 

Slide 2 How the eye works

Discerning three-dimensionality is a problem of human perception.  We rely principally on our eyesight to tell us what the world is like… seeing is believing.  Eyesight works by taking the reflected light from objects in the real world and projecting them onto a screen, the retina.  Although the retina is curved, it is a single surface and thus not unlike a two-dimensional page.  The brain then has to interpret that  projected light and from it deduce what the three-dimensional world is like.

 

Slide 3 Sources of information on depth.

There are five key types of information that the eye gives and the brain draws upon to interpret a three-dimensional universe.  We will look briefly at them all, though only four of them are relevant to a study of mediaeval painting. They are:

·         Stereoscopy

·         Focus

·         Shading

·         Aerial perspective

·         Perspective

 

Slide 4 Stereoscopy is derived from the fact that we have two eyes set slightly apart, so that they each view an object from a slightly different angle.  In modern times, this has been used to provide three-dimensionality to two-dimensional images such as those binoculars into which one used to insert two photographs taken from slightly different angles, or more recently the films you watch wearing red and blue glasses.  However, I am aware of no mediaeval antecedents for these.

 

Slide 5 Focus is partly derived from stereoscopy… the angle at which the two eyes have to align themselves varies as to how far away an object is… which is why you look cross-eyed when you look at something really close.  It is also to do with the science of lenses, which as we know from cameras can be used to make objects at a certain range appear sharp whilst objects at other ranges are blurred.  More modern painting, and especially photography, can deliberately blur the background to emphasise three-dimensionality, but generally speaking, mediaeval painting does not… or at least not very much.  Backgrounds may perhaps be shown a little more sketchily than foregrounds, but compared to later painters, mediaeval illuminators’ backgrounds tend to be remarkably detailed.

 

Slide 6 Shading is derived from the fact that the same light-source will reflect off a three-dimensional object to the viewer with varying intensity depending on how its surfaces are angled relative to that light-source.  At its crudest, it means that those parts of an object angled towards the light source are likely to appear lighter than those angled away from it.  Though varying in sophistication, some acknowledgement of shading is found in the majority of medieval illuminations, as we will see later in this talk. 

 

Slide 7 Aerial perspective is derived from the fact that air is not totally transparent.  It contains elements, notably water vapour, that do in fact reflect light, though to a very small degree.  Specifically, they tend to reflect blue light.  The further away an object is, the greater the thickness of air you are looking at it through, so the bluer it appears.  So distant hills always seem to be blue, whilst objects in the foreground are visible in full colour.  Aerial perspective is only really relevant, therefore, in landscapes where there are long vistas into the distance.  Such subject-matter is only found in illuminations towards the end of the mediaeval period, but during the fifteenth century one does start to see landscapes, occasionally as subjects in their own right and more often as back-drops to the main action of the scene, and the use of blued landscapes is often very pronounced, and very beautiful.

 

It is worth noting, though, that even though the effect of aerial perspective is only discernible at long ranges, the fact that the brain registers it as “a thing” means that you can cheat with a bit!  If you want to make something in the foreground stand out from the background, toning down the background colours with blue or grey and heightening the foreground colours will help create that sense of three-dimensionality.

 

Slide 8 Perspective derives from the fact that objects further away appear smaller.  Provided the brain knows roughly how big a thing is meant to be, it can estimate its range from the size of the image that is projected onto the retina.

 Excuse the choice of subject-matter, but military cemeteries are great for demonstrating this!

 

Slide 9.  The principle of perspective is that any parallel set of lines will, if traced and extended,  converge and meet together at the same point, which we call the “vanishing point”  Different sets of parallels will each have their own different vanishing point.  The vanishing point will always be at a height equivalent to the viewpoint of the onlooker… thus in the picture, where the viewer is rather taller than the crosses, it is positioned a little above the crosses.  The net effect of this is that the further away something is, the smaller it will appear.  You know this anyway, but here is the proof:

 

Slide 10.  With the wonders of Adobe photoshop, I have highlighted one of these crosses.

 

Slide 11. I have now copied and pasted it to only one row back, but you can see that where it now stands, it appears much bigger.  Yet it is exactly the same cross as the one in front. The brain is fooled because it is so used to the idea that “further away” should look smaller. Some brains may be better at this than others:

 

Slide 12

Father Ted and Father Dougal… “These cows are small.  Those cows are far away”.

https://www.youtube.com/watch?v=MMiKyfd6hA0

 

Slide 13 But why do far-away cows look smaller than nearby cows?  Here is a demonstration.  The screen nearest the cow has a small hole in it.  It is, perhaps, the pupil of a human eye, or the pinhole in a camera obscura.  The further screen is perhaps the retina of the human eye, or the projection screen of the camera obscura. The reflected light from the cow that passes through the pinhole is projected onto the screen.  The thicker line shows us how big the image of the cow is on the screen.

 

Slide 14 Move the cow back a bit, and we see how the image it projects is significantly smaller than the original image.

 The geometric principles behind this were probably known to the Romans, but in the mediaeval period they only came to be studied again as part of that general revival of interest in all things Roman that we call the Renaissance.  So we really only start to see scientifically accurate perspective appearing in illuminations from the late 15^th century, and even then not universally.  But this is the time when the principle of having a consistent “vanishing point” onto which all parallel lines will converge as they recede away from the viewer, comes to be appreciated.

 

All this was known to the Romans,  We now take it for granted.  I think I first learned it in primary school.  So did you… I doubt you learned anything new in what I have told you, I just reminded you of stuff you already knew.

 But actually, it isn’t obvious.  We only have to look at the work of mediaeval artists to see that the knowledge of these things, like so much scientific knowledge, was lost or ignored in the Dark ages.  It only began to be re-discovered in the mediaeval period, and reached a climax at the Renaissance.

 You can see the artists wrestling with the questions: I can see that things further away look smaller, but I know they are actually the same size.  I can see that parallel lines which recede from me look as though they get closer together, but I know that actually they do not.  And so you end up with a rather bodged half-way house, which acknowledges the existence of perspective but is not quite sure what to do with it.

 And that all raises a fundamental question which is answered in different ways throughout the mediaeval period.  What do we even want to depict?  What is, or what appears to be? The answer is not a foregone conclusion.  If all cows are the same size, why shouldn’t we depict them all the same size? If we know an object is red, why shouldn’t we paint it red, even if the side towards the sun appears pink and the side away from it appears maroon?  In a sense, what “appears-to-be” is actually a lie, and why should we necessarily go along with that lie?

 But then again, when we say “what is”, what do we actually mean?  From a mediaeval perspective, even more than a modern one, that isn’t straightforward.  I know the King is 5ft 4 tall and I know a peasant is 5ft 4 tall.  But I also know that the king is a far more important person than a peasant.  Which truth do I want to portray? In ancient Egyptian art, the answer is clear: I paint the pharaoh bigger than any human, but (probably) smaller than the gods.  You don’t quite see that in mediaeval art, but as we will see, there can be a link between bigger and more important (or at least, more interesting) which trumps concerns for perspective. And there are other questions about truth to be asked. I know that the evil things in this world are caused by demons.  I have never actually seen a demon because demons are invisible, but if I do not depict them in my painting, I am missing an aspect of the truth.  Do demons have size? Or, to put the question another way, how many angels can dance on the head of a pin?

 And then again, isn’t this overthinking it a bit?  Why am I really doing this painting?  Perhaps, because I know this book, maybe a Bible, is something important, and I want to make it look pretty and precious because I know it is important.  I don’t need to add truth to the Bible… it already has it.  But I do want to show that I know that it is an important book. Or maybe, this book is a gift for somebody and I have been commissioned to make it look as pretty as possible because then it will make a more impressive gift.  Or maybe the person who has commissioned me just wants to show off that he can afford books of this quality.  In that case, I’ll just do whatever is pretty, and whatever is the current style so that my patron doesn’t look like some backwards hick-from-the-sticks.

 So the notion that illuminations even should make an attempt at depicting a three-dimensional world is not one we can necessarily take for granted. There may be other priorities for what a painted image in a book is meant to do. In very broad terms, we can say that during that 600-1600 period there is a move from the assumption that depicting three-dimensionality doesn’t matter, to the assumption that it does.  There is a general direction in this movement, but not a constant speed. That move is paralleled in other art forms, but there are some specific conventions of manuscript illumination that play on this.  I’m going to look at some examples of famous manuscripts during this period… it is far from an exhaustive study, but is a selection of highlights demonstrating the general trend.

 

Slide 15 Lindisfarne Gospels c715 In the earlier period, in very broad terms, the emphasis is on decorative borders. Figurative elements may appear in those borders or even as full-page pictures, but of course a border is not something that is attempting to replicate real, three-dimensional space.  If anything, it is a celebration of the two-dimensionality of the page and of decorative pattern.   Here, in the Lindisfarne Gospels,  in the image on the left, much is pure pattern.  We can see depictions of recognisable things… the intertwined birds and the cat’s head.  But they are part of the pattern… they don’t even pretend to exist in real space, so even to ask questions about how they depict three-dimensionality is to ask a non-question.  The picture on the right is also from the Lindisfarne Gospels, but in a very different style.  That is because it quite consciously copies an earlier, Roman-inspired illumination.  Some attempt at setting this figure in space has been attempted, but it is clearly not a priority.  The bench on which he sits is rendered in a sort of false perspective… enough to tell us that one end of it is further away from the other, by the use of diagonal lines.  But the artist has little understanding of how this should work. 

Slide 16 In fact, if you trace the lines that ought to converge on a vanishing point, they actually diverge. Note also that the figure has been painted in very flat colours.  No attempt has been made to use lighting and shading to imply three-dimensionality.  Either the artist doesn’t know how to do it, or it just isn’t a priority.

 

Slide 17. Benedictional of St Aethelwold c 975 In this later Saxon manuscript, decorative borders remain the priority, but within them we sometimes see pictures, of which this one, showing the assumption of Mary, is among most complex. Shading is starting to appear, most obviously in the drapery of Mary’s dress.  Although this is more decorative than anything, it does start to suggest that there is depth to the drapery. Note, though there is no very consistent sense of from which direction the light is shining. Rather, lighter areas are nearer to us, darker ones further away. It is thus not “accurate” shading… but it does nevertheless suggest three-dimensionality. The faces, however, are painted flat.  The figures are not really set in any actual space… the green behind the apostles might perhaps be a very basic suggestion of a background of trees, but really the terms background and foreground don’t apply here.  The apostles are all shown the same height, whether they are in the front row or behind.  Mary, it is true, is shown larger than the grieving women behind her, but this is almost certainly because she is more important, not because she is in front.

 

 

Slide 18 Winchester Bible c 1150-1175 Many, though not all, of the pictures in the Winchester Bible appear within the borders.  The picture on the right is from a border.  This sets them up from the start as being primarily decorative in function, though their content is relevant to the Biblical text.  Though it is stylised, shading is used here, visible now not only in the drapery but also in the faces.  The figures are not set in space; that is, there is no attempt to depict a background, instead a solid screen of flat gold-leaf avoids the question.  The page to the left, by a different artist but from the same book, includes such elements of background as are essential to the story, but no more.  We are told that David took his sling-stones from a brook, so a brook and stones are included, but there is no real sense of them being back-ground, foreground, or indeed relating to three-dimensional space in any way.  Generally, a flat coloured background takes away any question about how these figures fit into space. All figures are the same size, except of course for Goliath the giant, who really needs to be shown as big, and David as a youth, who really is small… perspective is not the reason for any size differences.

 

Slide 19 Macclesfield Psalter c 1325-1350.  Moving on a couple of centuries, characteristic of the fourteenth century is the “bar-and-ivy” style, of which the Macclesfield Psalter is perhaps the finest example.  Although there are “pictures” in the Macclesfield Psalter, it mostly consists of decorated borders. Though these still give no scope for a consideration of perspective, the figures within them start to be rendered more three-dimensional by the use of shading.  This can be quite sophisticated.  For example, we tend to think of Michelangelo as one of the greatest exponents of using shading to render the complex three-dimensionality of his contorted, muscular bodies, such as the famous Adam in the Sistine chapel ceiling painted in about 1510.  Yet an un-named English illuminator had already produced this figure in the Macclesfield Psalter nearly 200 years before!

 

Slide 20 Lisle Psalter c 1308-1340 Also from the fourteenth century, but this time showing what happens when pictures appear in their own right, rather than as elements in borders. In fact, two centuries on from the Winchester Bible, things are much the same… progress was not consistent in this period.  Again, figures are mostly shown the same size… where there is a difference, as in the crucified Christ and Adam at his feet, it is about importance, not how far away they are.. Adam, after all, is in front of the cross.  Complex diapering backgrounds take the place of a landscape background, taking away any need for considerations of how figures fit into space.  Shading is beautifully done, but still works mostly on the principle that light means forward, dark means backward, rather than on the basis of light coming from any particular direction.  Although the beautiful delicate drawing of this manuscript shows how far drawing has evolved, the representation of three dimensionality has actually progressed very minimally thus far in our period.  But, as we enter the fifteenth century, that will change dramatically. 

 

Slide 21 Belle Heures du Duc de Berry 1405-1409

Enter the Limbourg brothers, perhaps the finest and well known of manuscript illuminators.  They made their name with this Book of Hours.  Exquisite as it is, it is still very fourteenth century in style.  There is a little background scenery, but it is nominal, and essentially the background is still flat diapering.  But we get a hint of perspective… the figures are definitely smaller as they get further away, and the artist is prepared to do this even though it makes the Virgin Mary and St John larger in scale than the most important person in the scene, Jesus.  There is also some constancy to the light source in the shading… it comes from top left.

 

 

Slide 22 Tres Riches Heures du Duc de Berry c 1410-1415

So pleased was the Duc de Berry with the work, he commissioned another Book of Hours from the Limbourgs, the famous “Tres Riches Heures”.  Suddenly we are in a different world, with full landscapes and therefore recession of space.  It is not scientific perspective, it is intuitive. In this scene for February, the sheep-shed is shown tapering away towards the distance… but there is no consistent vanishing point.  The man chopping down the tree in the background is, as one would expect, smaller than the figures in the foreground… but not really by enough. In fact, he is about the same size as the furthest of the three figures in the house.  The trees themselves are really rather tiny… because, to be honest, they are less interesting than the man himself.  Yet, without a doubt, we are now within a three-dimensional world, with landscape as a background.  And again, note the shading, consistently from top left.

Slide 23 Bible Historiale: Proverbs to the Apocalypse, 1411

Another painting, from the same milieu as the Limbourg is this one of King Solomon. The panels of the throne are shown with a receding angles that almost come to a consistent vanishing point… the artist seems aware of the theory of perspective.  But he doesn’t fully understand it.  Note the tiles.  The artist knows that tiles look smaller the further away they are.  But he also knows that tiles are right-angled and arranged in rows and columns which are parallel.  So how does he solve that conundrum?  It’s not true perspective, but it is instinctive perspective.  The shading on the tiles has also been used to suggest recession.

 

Slide 24 Tres Riches Heures du Duc de Berry c 1485-1489

The Limbourgs died before they could finish the Tres Riches Heures, and by the time it was completed in the 1480s, the Renaissance was well established, especially in Italy where the book was now located.  An artist named Jean Colombe was one of two commissioned to complete the book.  In most respects, Jean Colombe is a far inferior artist to the Limbourgs.  But his perspective is a little more convincing, and look at his aerial perspective!  His use of blue shades to set the background of his paintings into the far distance is masterly… a device hardly found in the work of the Limbourgs.  The lighting is really quite sophisticated… and always from top left.

 

Slide 25 Gentile Bellini: Procession in Piazza San Marco 1496

In fact, the convention that light is from top left is by now very strong.  Though picture in fact shows a panel painting and not a manuscript, it does illustrate well how strong it was. I once did a scroll based on this painting, and in doing so noticed something odd.  From where Bellini must have been standing to get this view, north is to the left.  The sun never shines from the north in Venice, or anywhere north of the tropics.  But so strong was the convention that light source is top left, that is how he painted it. 

 

Slide 26 Sforza Hours c1490

I have mentioned that decorative borders to some extent deny three-dimensionality.  But by the High Renaissance, expressing virtuosity in depicting three dimensions is so important, that the border gets transformed.  As with so much in the Renaissance, this is due to a rediscovery of Roman decorative motifs, which are shaded to emphasise their three-dimensionality.  Birago, working in Milan in the late 15^th century, is one of the great exponents of this.  He continues the tradition of combining abstract shapes, often based on flowers but now also with classical motifs such as urns, with grotesque figures.  The figures too are given a classical twist.  But the abstract shapes and the figures are now, more than ever before, rendered with virtuoso shading and perspective, as though Birago is boasting, “I’ve read the text-books, I know how this stuff is done!”

 But note not only the shapes in the borders, but also the background.  It is a simple, two-dimensional block of colour… but the decorative objects in the foreground are casting their shadows onto it.  We are moving into the realm of trompe l’oeil.  We are acknowledging the existence of the flat surface of the page, but pretending that the painted images are in front of it.  We’ll come on to that in a minute, but just before we move on from this image, take note of the use of aerial perspective in the landscape with the rabbits.

 

Slide 27 Book of Hours of the Blessed Virgin Mary “Flora” c 1483 - 1498

As we move into towards the 16^th century, a popular style of border is to paint incredibly detailed, lifelike images of flowers and insects against a flat background of gold leaf. This is an early, and particularly fine, example of the style, by another famous illuminator, Simon Marmion.  Note how a dark shadow is painted over the gold to give that trompe l’oeil effect… it really looks as though these flowers and insects have been laid on top of the gold sheet.  Meanwhile, on the facing page, we have to remind ourselves that the frame around the painting is not really made of carved wood.

By now, the images contained within those trompe-l'oeil borders are very sophisticated in their use of shading, perspective and aerial perspective.

 Slide 28 Opera Porphyrios” 1483 The most wonderful example of trompe l’oeil I have seen is this, an edition of Aristotle’s “Opera Porphyrios” from 1483 by Girolama of Cremona or his school.  It is not strictly an illuminated manuscript.  The text is in fact printed, though the illumination is hand-painted just as it would have been in a more traditional manuscript.  But look how the artist has played with the interface of the flat surface of the page and the illusion of three-dimensionality in the images.  We get the impression that the paper of the printed text has been torn away to reveal the painting underneath.  The revenge of the book illuminator on what will be his nemesis, the printed book!

 

So by the end of the 15^th century we have seen the question of how to depict three dimensions on a two-dimensional page well-and-truly tackled. I could go on to show examples from the sixteenth, but they would reveal nothing new.

 

“But”, I hear the nerds among you complaining, “there aren’t three dimensions, there are four!”.  Actually, the real nerds are probably complaining that there are even more than that, but that’s real esoteric stuff.  How, if at all, did mediaeval artists depict the fourth dimension, time, in illuminated manuscripts?

 

I was inspired to think about this when I was asked to produce some titles for the Insulae Draconis dance video.  If you think about it, animation is precisely an answer to the question of how you depict the fourth dimension on a two-dimensional surface, just as perspective and shading are answers to the question of how you depict the third dimension.  The problem is, mediaeval people didn’t have access to the technology we have that enabled me to animate an illuminated page.  What did they do?  Did they do anything?  Did they have their own answer to the question of how to depict the fourth dimension on a two-dimensional surface?  This, of course, is particularly important when telling a story, where movement through time is the essence.

 Slide 30, “Romuleon” 1480 Yes, they did.  The simplest and most common was to depict the same character at several different points in the story all in the same picture. In this example, telling the story of Romulus and Remus, we see a shepherd discovering the two brothers being suckled by a wolf in the foreground.  In the background, we see the same shepherd holding the two babies (they are there, even if they are tiny!) and presenting them to his wife.  Time has passed between the two incidents.  As with animation, we fill in the gaps in between.  We can tell that he has rescued the babies and taken them home with him.

 Before continuing to look at the fourth dimension, this picture is a very good re-cap of what we have learned about how manuscript artists tackled the other three.  We have shading, with the light coming from top left, most obvious in the rocks.  I haven’t really mentioned shadows yet, which make their appearance about the mid 15^th century, but they can clearly be seen at the feet of the foreground shepherd.  Reflections appear in the water… reflections can’t really be called a way of depicting three-dimensionality, but they show that by now artists have committed themselves to the notion that they will paint things as they appear, not as they are.  There are not really a house or trees in the river, but that is how they appear, just as people far away are not really smaller than people nearby, but that is how they appear.  In this painting, the shepherd in the distance is, as he should be, smaller than the one in the foreground… but not really by enough for this to be true perspective.  The house is shown rather too small for the couple to actually live in… a hangover of the idea that we paint things that are important or interesting larger than things that are not.  Despite this rather poor perspective drawing in the foreground, look at the background, where the towers of the city recede into the distance.  This is very convincing perspective!  And see beyond them, the blue landscape… aerial perspective.

 

Slide 31  Roman de Giron le Courtois (14^th Century)

And of course, when you have a manuscript, which has lots of pages bound together, you can also depict successive events in pictures on successive pages.  That’s clearly going on here in the 14^th Century manuscript of a romance… a story-book, where the illustrations assist the telling of a story like a modern picture-book.

 

Slide 32 Sigenot 1470 An extreme example of this is the medieval codex Sigenot (circa 1470) which has has sequential illuminations with relatively short intervals between different phases of action. Each page has a picture inside a frame above the text, with great consistency in size and position throughout the book (with a consistent difference in size for the recto and verso sides of each page). Does this look familiar?  Do you remember those flick-books you had as a child, where you flicked the pages so as to rapidly see one image after another?  Low-tech animation!

 

I have no idea whether it would be possible to do that with this book.  Flick books tend to have their images in the corners, it would be hard to flick a whole page fast enough to get the desired effect.  I have no idea how springy parchment is, or whether the book is bound in a way that would facilitate flicking.  I suspect that any request to the curator to flick through their extremely rare and valuable manuscript would be met with a very stony look.  But here is what happens if you animate a few of the illustrations:

Not conclusive, but interesting!