Notes on Roden Crater


Shinsuke Shimojo
Published in James Turrell:
Where does the light in our dreams come from ?
The official catalog book for the exhibition tour in Japan, 1997.

Date: August 23, 1997.
Time of observation: 18:00-19:00.
Weather: clear with clouds.


There had been the usual monsoon in the afternoon, but it was clear when we arrived. The air was transparent, and the sky was clear though there were clouds with various patterns and textures. It turned out to be the ideal condition for observation. Michael Bond, the chief assistant of James Turrell, or rather his "site foreman", kindly guided us.

The bottom of the crater was shallower than my original impression, formed from photos, but deep enough so that when we stood in the center, the rim appeared to be well above us. The shape of the rim had been modified by scraping and adding soil, such that its height was roughly uniform except for a slightly lower part on the eastern side, and its shape was elliptic but very close to a circle. The grit inside was dark red.

The following observations were made by myself while reclined on my back at the bottom center of the crater, with my head slightly lower than my feet, and my head slightly tilted backwards.

(1) The vault effect. It has been established that the sky is perceived by an observer, who is standing upright, as the inside surface of a vault which is closest to the observer at its top and farthest at the horizon. The presence of the rim and the observer’s lying posture enhanced these effects in the crater. The sky appeared almost as though it were a flat ceiling surface. The standing posture generally leads to a greater flattening effect than the lying posture on ordinary ground, but this relationship seemed to be reversed at the bottom of the crater owing to the rim (compare Figure a) and b)).

(2) The ellipsoid effect. In the lying posture, the perceived shape of the "sky vault" was not totally concentric, but rather cylindrical. The rim of the crater also appeared as an elongated ellipse, being consistent with the cylindrical shape of the vault. Both their longer axes were aligned with that of the observer’s body, i.e., the direction of the backbone. When the direction the observer’s body was lying in changed (for example, lying on his back in the direction of 9-3 o’clock as opposed to the original 0-6 o’clock direction), the perceived shapes of the sky vault and the rim changed, and again became cylindrical and elliptical, with their longer axes following the main axis of the body. It would be natural to assume that the portion of the rim which is closest to the observer distorts in shape first, and then this distortion spreads beyond the rim to include the sky, although there was no solid observation consistent with this (not much has been available about these phenomena in the literature either, to my knowledge). When the observer, still lying on his back, rotated his head and looked sideways or kept his head facing straight above but looked sideways with an eye rotation, the effects were weakened but still remained. Thus, factors limited to the eyes, such as vergence or accommodation, would not be solely sufficient to explain the effects. Factors related to the whole body should be taken into account. Could it be related to size constancy and the moon illusion? Maybe or maybe not, but there is little doubt that these effects are related to, or influenced by our visual experience while we are standing and walking in daily life.

3) The surface color effect. When the observer was standing, not only the clouds but also the blue parts of the sky had a volume in the dimension of depth, i.e. the direction of the line of sight. When the observer was lying on his back, however, the sky became flat and opaque, appearing almost as though everything had been painted on the inside surface of a dome. This impression became even more prominent right after sunset, when variation of the colors in the yellow-blue axis increased. There were enhanced contrasts: between the bright sky and the dark interior of the crater, and between the orange-yellow clouds and the blue sky. The whole sky appeared as a figure, delineated by the sharp edges of the rim. The texture created by flowing clouds made it appear even more like a brush painting. The initial volume color was framed by the rim and turned into aperture color or film color, and even into surface color, owing to the optical condition at the time of sunset. It may be noteworthy that the vault and the ellipsoid effects that I mentioned above are also limited to the component of surface color, not those of volume or aperture (film) color, although the sky potentially has all three of these components (see the footnote 2 in Perception as thinking without symbols). The component of surface color became overwhelmingly dominant when the observer was lying on his back. After sunset, only the upper layers of clouds reflected the sunlight in red, while the lower layers remained dark. Thus, they appeared as two surfaces with depth between them. By having the observer lie on his back, however, they were again compressed onto a single surface, which appeared as though it were a canvas with overlaid paints. When the observer was lying on his back and rotated his head and looked sideways or kept his head facing straight above but looked sideways with eye rotation, the effects were weakened but still remained, just as in the case of the ellipsoid effect.

(4) The homogeneous illumination effect. When the clouds in the east side of the sky turned orange, the luminance was well matched in all directions of the sky for a brief period of time. This, together with the rim of the crater, in effect, created nearly homogeneous illumination throughout the top half of the visual field. As a result, the body’s shadow completely disappeared from the ground, even when the observer was standing. This was a purely optical effect, but even before the observer realized that shadows were missing, he was vaguely aware of the unusual illumination environment and of having a sort of supernatural impression.

Through his past pieces of work, I knew already that it was Turrell’s favorite technique to turn the volume or aperture color of the sky into surface color by manipulating the observer’s frame of reference. But I had never imagined that he could realize such a powerful effect on such a big scale, only by utilizing naturally occurring structures and lights. The situation which is brought about here is literally unprecedented in the history of experimental psychology on visual perception since it: a) is outdoors, b) is huge in scale, c) requires only natural light, d) is perceived in the posture of lying on one’s back, and e) delineates half of the visual field with physically symmetrical and concentric edges. I see a large, never explored domain of research here.