Richard Serra

“Masterpieces come late.” I was stunned to hear this pronouncement, last spring, in the course of a phenomenal lecture by T. J. Clark on de Kooning’s 1958 Suburb in Havana. Let’s bypass my puzzlement at the word “masterpiece” from an art historian whose past work would seem to preclude the use of such an ill-defined notion. The tempo of the lecture was fast, and at the time I could make only one hasty association: that was with Matisse, and with my unlikely preference, in his whole oeuvre, for the 1947–48 interiors, among his last canvases.

This association helped me grasp what I take to be the historical reasoning behind Clark’s remark. “Masterpieces come late,” that is, when there is nothing more to prove, when the theater of rivalry is finished or the burden of tradition lifted, when novelty is no longer an issue. Such rare conditions give an air of apparent ease to the work, a looseness all the more striking in that it is not so simple for the artist to accept liberation from the historical superego, including his or her own standards. It took Matisse seven years, after the 1941 surgery that nearly cost him his life, to reach that point with The Egyptian Curtain.

I think Richard Serra has also reached that point in the three Torqued Ellipses, massive yet elusive volumes of unidentifiable shape. Torqued Ellipses I is the most agitated of the group, the spatial to-and-fro of its rusty “wall” the most extreme, though the effect is subdued by the oil treatment, which gives the surface a velvety quality (the rotation of the interior space, sandwiched between two ovals at top and bottom, is greater than in the other two works); Torqued Ellipses II is at the other end of the spectrum, with the overhang of the wall the least pronounced (the rusty metal is left untreated); Double Torqued Ellipse, which combines two ellipses, one inside the other, is the harshest. Its inside chamber appears higher than the others because it is less elongated, and the shiny metal seems almost fresh from the mill.

Are these Serra’s masterpieces? If so, we should examine their belatedness, what’s not new in them, what they summon that has been present in Serra’s work for a quarter century but that they paradoxically, and seemingly without effort, enhance and emphasize. For it is in part such an unapologetic oldfangled-ness that opens them to a different order of experience—it is exactly what makes them possible masterpieces.

There are two major principles in Serra’s work that are carried to an extreme in the Torqued Ellipses: first, the incommensurability between knowledge, particularly the type of eidetic knowledge that is the stuff of geometry, and perception; and second, the presupposition of a moving viewer. To trace the application of these principles in Serra’s production is to tell its entire history after the early “Prop” pieces. In an essay I wrote years ago, I mused about my shock at discovering after the fact that the placement of the plates in Spin Out (for Robert Smithson), 1972–73, made a perfectly regular geometric figure (it was after having spent a full afternoon in situ wondering whether this was so, and concluding negatively). Exchange, a 23.5-meter-high vertical sculpture created last year in Luxembourg, poses the same riddle. When I recently visited it with two architect friends, neither could meet my challenge to properly survey the piece and produce a drawing that would explain the odd relationship between plan and elevation (mortified, they had to declare forfeit). As for the second principle, the walking beholder was assumed as early as Shift, 1970–72, and it was such a viewer who would discover, passing through Clara-Clara, 1983, that the two sides of his or her body were out of sync. Among Serra’s best essays is a short text he wrote about circling the long-gone Rotary Arc (either by car or on foot): it is the verbal equivalent of a storyboard. It has been clear for some time that one cannot look at a post-1970 Serra while stationary.

What, then, makes these principles even more operative in the Torqued Ellipses? Perhaps the fact that they were finally taken for granted by the artist as he worked on these baffling sculptures. Being relaxed about the very principles on which his art is grounded—a confidence that comes with the aforementioned belatedness—allowed Serra to think about the process of engendering sculptural forms in ways he would likely have condemned before. Indeed, the purely geometrical issue he began with—“how to relate the top and bottom ellipse to each other and how to rotate the void,” as he comments in the Dia catalogue—might explain why, for once, the rhetoric of weight that has always (unnecessarily) encumbered the appreciation of his work is irrelevant. “Each sculpture [weighs] forty tons, yet they don’t seem heavy.... There’s a certain weightlessness to them,” he says, and the plates “could be thinner.” How unusual of Serra, and how remarkably unself-conscious, to be able to speak that way. It almost sounds like Greenberg’s old notion of sculpture as optical mirage, something that Serra, among others, has always rejected, and for good reason.

The geometrical point of departure only underlines how unrelated to any fact of geometry is one’s experience of the pieces. Serra mentions that someone asked him “what a cross-section would be halfway up Torqued Ellipse I.” I did. It was a dumb question: the same oval the plates trace on the floor or above our heads would appear, with the exact same proportions, regardless of where the sculpture is cross-sectioned (although differently positioned on the intersecting horizontal plane). As a matter of fact, no ellipse is torqued whatsoever. (If anything is, it is the “walls” that connect the identical top and bottom elliptical circumferences.) At any rate, receiving the answer to my dumb question was no help at all since I could not see the asserted identity and thus relate it to my own experience. Serra’s work has challenged the identity principle right from the start: the Torqued Ellipses just do it ever more so.

The question is, Why? It took a bit of a detour for me to understand. It took, more precisely, an excursion to Washington, DC, where, next to the East Building of the National Gallery of Art, the elements of Tony Smith’s Wandering Rocks, 1967, are scattered on the lawn. Examining this sculpture is an exercise in induction. At first one simply cannot understand what exactly are the polyhedra that one beholds—the solids are cut so that their edges look continuous even though the planes they define change direction. It takes several turns around the solids to get that they are all fragments of a single “beam” of lozengic section, each oriented differently and thus catching the light in contradictory ways. But the point is that one finally “gets it,” gets the geometric identity (the lozenge) behind it all and the way the specific configurations of the crystalline shapes relate to this common denominator. Then—Eureka!—one properly reads the different facets of the solids, reads their varying orientation, and marvels at the sculptor’s manipulative skills.

There’s no such trick in the Torqued Ellipses, nothing to “get” in this manner. There are many dissimilarities between Smith’s sculpture and the Serra pieces, but none more important than the fact that Smith plays with the arris (per OED, “the sharp edge formed by the angular contact of two plane or curved surfaces”). He transforms it into a double contour (hence the conundrum), while Serra eliminates it altogether. The limits between abutting planes are always clean-cut in Smith’s sculpture, which is why one can in the end solve the conundrum and know what one sees. One can draw or photograph every possible vista of any solid in Wandering Rocks and pinpoint the moment at which a spatial articulation becomes ambiguous and the moment at which everything falls into place. The logic is binary, and the turning point is always marked by an arris. Not so with the Torqued Ellipses: it is impossible to delimit such articulations because there aren’t any. The “walls” are constantly shifting, and because the surface is continuous yet irregular (or of a kind of geometric regularity that can be grasped only by a computer), one is never sure of the direction in which they swerve. In short, one is left without any spatial coordinates, incapable of measuring the distance between oneself and any point on the “walls.”

This dissolution of coordinates is even more striking when circling the exteriors of the Torqued Ellipses than when entering their rings—but then, it is true that the spatial torsion puts the inside/outside distinction to a hard test. In the catalogue, Serra convincingly underlines their critical relationship to architectural space (as in all his work—here again an effect of their “belatedness”) and why he realized them in steel rather than concrete (so that they would not be perceived, purely and simply, as architecture). But the only moment I felt the skewing of the “walls” as strongly inside one of the sculptures was when walking in the space between the two membranes constituting Double Torqued Ellipse. Advancing in this curved corridor that expands and contracts at the same time, as if in the digestive organ of some monstrous creature, I felt literally seasick.

“I wanted to keep them within the definition of sculpture,” says Serra of his Torqued Ellipses. And this he does, even though he overturns quite a few rules along the way. I am especially thankful to him for having led me to understand, retroactively, what had always puzzled me about one of Matisse’s greatest sculptures, his 1909–10 Seated Nude (Olga). Of course, there’s no question of geometry in such a figurative work, nor of a “sharp edge formed by the angular contact between two plane or curved surfaces,” but there is a suppression of the articulation between front and back, such an articulation being a kind of anatomic equivalent of the geometric arris. When turning around Olga, the beholder is always caught realizing that he or she is suddenly looking at her back while a split second before it was her belly that was in view, and no matter how many times the journey is rewound, the viewer never finds the moment at which the shift occurs.

There is perhaps no better definition of a masterpiece than its power to make previous works of art intelligible. If I am right, the Torqued Ellipses are in pretty good standing.

Yve-Alain Bois is Joseph Pulitzer, Jr., Professor of Modern Art at Harvard University.