Villa of Poppaea, Bedroom

	Thoughts on Vanishing Point Perspective in the Villa of Poppaea	Home Up
	William Storage and Laura Maish Email us about this page

(Click the images for larger versions)

oecus	cubiculum
diata, west wall	dieta, north wall

Many of us learned that vanishing-point perspective was a product of the Renaissance. This is terribly wrong. Plato and Aristotle discussed it in regard to theater; and Polybius cites Greek historians who discuss its use in architectural rendering. Vitruvius discussed it in a manual of architectural training in the first century BC. Vitruvius explains in de Architectura:

Thus adhering to the principles which I found in those of their works adapted to my purpose, I have endeavored to advance further. Agatharcus, at the time when �schylus taught at Athens the rules of tragic poetry, was the first who contrived scenery, upon which subject he left a treatise. This led Democritus and Anaxagoras, who wrote thereon, to explain how the points of sight and distance ought to guide the lines, as in nature, to a centre; so that by means of pictorial deception, the real appearances of buildings appear on the scene, which, painted on a flat vertical surface, seem, nevertheless, to advance and recede.
- de Architectura, Book VII - Introduction, Par. 11

In chapter two, he briefly explains vanishing point:

Fitness is the adjustment of size of the several parts to their several uses, and required due regard to the general proportions of the fabric: it arises out of dimension (quantitas), which the Greeks call ποσότης. Dimension regulated the general scale of the work, so that the parts may all tell and be effective. Arrangement is the disposition in their just and proper places of all the parts of the building, and the pleasing effect of the same; keeping in view its appropriate character. It is divisible into three heads, which, considered together, constitute design: these, by the Greeks, are named iδέαi: they are called ichnography, orthography, and scenography. The first is the representation on a plane of the ground-plan of the work, drawn by rule and compasses. The second is the elevation of the front, slightly shadowed, and showing the forms of the intended building. The last exhibits the front and a receding side properly shadowed, the lines being drawn to their proper vanishing points. These three are the result of thought and invention. Thought is an effort of the mind, ever incited by the pleasure attendant on success in compassing an object. Invention is the effect of this effort; which throws a new light on things the most recondite, and produces them to answer the intended purpose. These are the ends of arrangement.
- de Architectura, Book I � Chapter 2, Paragraph 2

That the ancients were capable of executing single point perspective is acknowledged by many art historians and archaeologists, and is obvious from a number of extant paintings, including the oecus in the Villa of Poppaea. It seems more often than not, however, that Roman painters of 2nd Pompeian style, either lacked understanding of the geometric principles of vanishing-point or chose to depart from them for an unknown reason. It appears certain that some of the Roman painters were only "eye-balling" it, and produced some bizarre results that could have inspired Escher. But the notion that the chief Roman painters had only an intuitive command of perspective seems highly unlikely; the construction lines used in the above oecus could only result from a conscious effort; its central perspective approach is almost identical to that used by Masolino in the 15th century. Why they might chose to depart from the simple rules for part of a painting and adhere to them in another is the subject of speculation. Most scholars who have addressed this point agree that it has something to do with the paintings' being designed to be viewed from a specific sweet spot. H.G. Beyen wrote a detailed analysis of this phenomenon in 1939, concluding that the ancients possessed sufficient knowledge of optics and geometry. Alan Little took it a bit further in 1971. It seems to us that both of them, despite their considerable analysis, failed to address several possible reasons for the departure from standard central-point perspective. John R. Clarke gives a good overview, citing many earlier writers, in The Houses of Roman Italy, 100 BC - AD 250.

We added construction lines to four of the paintings from Oplontis. On one of them, the oecus, it appears that all lines converge. In the other three, lines representing those perpendicular to the surface of the painting converge at two or three points along a vertical line through the center of the image. In most cases, lines extending from corners (of fictive three-dimensional structures) closest to the head of a standing viewers relate to the highest point of convergence; and points farther away - particularly those near the floor - converge at lower points. Beyen concluded from a statement from Vitruvius regarding receding lines intersecting at the center of a circle that Vitruvius used a complex scheme involving circular guide lines in the painted image. But he did not seem to us to succeed in showing how this manifests itself in the extant wall paintings. We'll also note here that some art historians appear overly fond of drawing a large number of arcs and circles on ancient structures and paintings, seeing meaning in intersections for which their is no evidence of intent by the artist. For example, see Richard Brilliant's bizarre proposal for a geometric design scheme for the Arch of Septimius Severus, which involves dozens of complex intersecting arcs, that, of course, coincidentally intersect at key points of the arch.

The scheme used in the bedroom (cubiculum, room 11) seems straightforward, considering the small size of the room . Plato noted that, when viewed at close range instead of from a seat in the theater, the use of perspective resulted in a very distorted appearance of the painting. Beyen may not give this point enough weight. The oecus is a large room, where a seated viewer could be sufficiently distant from the wall - say, a distance equal to the wall height. The bedroom is tiny, however. Even without furniture, a viewer would strain to be more than a body length from the painted wall. At such distance, a viewer observing the bottom half of the image would be looking down at a steep angle. This angle results in vertical perspective effects; that is, he would see image-vertical lines (the pillars - painted lines representing virtual vertical lines) converging toward a vanishing point somewhere below eye level on the wall. This effect would be equivalent to that produced by standing close to actual 3-dimensional pillars.

If proper central-point perspective were used in the bedroom, all image-horizontal lines perpendicular to the plane of the wall would intersect at a central point. In the below image the green lines indicate some of the lines in the painting that intersect the central point. Red lines indicate painting lines that intersect below the central point. Blue lines indicate where the red lines would be if rendered in correct single-point perspective.

The below image shows how the above image appears if viewed downward from a point close to it. It shows perspective applied to the 2-dimensional image above. Note that because of this perspective, the blue lines in figure 2 are at a substantially greater angle above horizontal than the equivalent lines in figure 1. They should not be, however; because they represent horizontal lines perpendicular to the surface of the wall. If a viewer were to look down 3-dimensional pillars at a steep angle, the blue lines would not be affected by the second vanishing point in the vertical plane, as they are orthogonal to it. So for these lines to have a natural appearance - when projected onto a 2-dimensional wall and then viewed downward at a steep angle - they would need to be drawn at a somewhat lower angle above horizontal. This is exactly what the painter did in the small bedroom.

Another way of looking at this is that "proper" vanishing point technique is only valid when an image is viewed from a sufficient distance that the surface of the image is roughly perpendicular to the viewer's line of sight. For a standing viewer in the bedroom, this is not the case. The artist thus adjusted for this by lowering the intersection point of the blue lines. The same rationale could explain the scheme used on the north wall of the diaeta, but not the west wall of the same room.