“Category error! It’s a category error!” That cry rang through the studio in James Stirling’s former art gallery space at Harvard University recently during an event called Opacity. Organized by the international architecture and engineering firm HDR, it brought together a group of four jurors (Harvard’s Preston Scott Cohen, Allison Williams, FAIA, founder of the San Francisco-based firm AGWms_studio, ARCHITECT editor-in-chief Paul Makovsky, Chicago-based writer, ARCHITECT contributing editor, and architect Ed Keegan, AIA, and myself) to review work the firm had done in the last year. It was Cohen who first introduced the idea, commenting on a building in which a part of the façade didn’t seem to know if it was a plane, a column, a block, or something in-between, but pretty soon we were seeing these glitches in the architecture matrix everywhere.
In fact, when I left the gallery to have dinner, I walked by Robert A.M. Stern Architects’ Harvard Law School building, and stopped in my tracks. Here was one of the best purveyors of a stripped-down (and sometimes rather well-dressed) Neoclassicism and neo-whatever making such mistakes. There are the almost symmetrical corners, one of which seems to be both an addition to the main body of the structure and the end point of a side wing. The confusion is elided by a curved window. At the other end of the building, the arches, lifted from somewhere else (perhaps a homage to the nearby H.H. Richardson Sever Hall), are neither directional nor apparently part of any axis. The corner of the main blocks sits on top of what should be a void. Then there is the main façade, whose length is broken up into panels, leaving the top floors above them to step back. No doubt this reduces the building’s visual weight, but it also confuses the status of the main façade. Are these reduced bays or sliced segments of a skin? What’s more, the space between them is sliced open by a lancet window that lines up not with a window above, but with what should be structure.
OK, so that was a super-formal description of what is otherwise a perfectly imposing and nicely detailed building, Moreover, I can almost hear the always witty Bob Stern tsk-tsking as I write this. Haven’t I read my Robert Venturi? Don’t I know that all of this was intentional, both because of the functioning of the building and to amuse and delight careful (but not careful enough) lookers such as myself?
Even if I ignore that voice, I have to ask myself, do category errors even matter? If it is only architects and critics who can spot them, and if they might even be deliberate ways to play with the structure and thus delight the knowing observer, are they to be not only tolerated but actually celebrated?
I posed my question to my usual sparring partner, Charles Dilworth, FAIA, who instantly went into a discussion of the “lies” told by the façade of the Seagram Building, whose grid of glass, steel, and bronze is as different from the Law School as you can get: “One of the great problems with a “real” steel column at the corner is the fact that it is not bilaterally symmetrical. The “fake” columns solve that problem,” he pointed out. “But the real column is going to be much larger than the fake ones, and this reduction of the visible member size produces subconscious anxiety because the building “sits” and “floats.” Also, “real” structural columns usually have ungainly size changes every few floors as they rise, and are often detailed indelicately–for very significant cost and performance issues. So, when people do exposed intumescent-painted steel or exposed concrete frames “neatly,” they also have to lie a lot, and waste a bunch of resources.”
All of that is very true, but I would argue that this is not a category error, which is to say, a problem of trying to build apples out of oranges, but rather a problem inherent in the chosen mode of design itself. If you want to design an expressed steel structure, you are going to run into exactly the kind of problems Dilworth points out. You could say that Ludwig Mies van der Rohe solved these problems rather elegantly, if indeed at great cost, both in terms of dollars and the environment. Unless:
The Seagram Building Construction Ecology (Actar Publishers, 2021),an excellent new book by Kiel Moe on the building’s ecology, argues exactly that point.
A category error is something different. In the case of the Seagram building, I pointed out to Dilworth, Mies van der Rohe could have crossed the grid in front of the windows, could have turned one of the columns into concrete, or could have extended the diagonal interior bracing to the outside, revealing what, in the logic of the system he chose, should be hidden. Such strategies of collage or partial expression are quite common in many Postmodern office structures.
What bothers me, as it did, I think, at least some of the other jurors at the Opacity event, is when the effect is either unintentional or poorly carried out. I would argue that, despite my respect for both HDR and the Stern firm, that is the case at the Harvard Law School building, and we found quite a few examples of such problems at the HDR firm. From my window at Virginia Tech, I can also see examples of such contradictions that have been caused by the University’s insistence on the use of “Hokie Stone,” quarried locally, on any new building. The results include such examples as a stone base that has coursing at the corner, showing that material to be just a cladding, supporting an unadorned, stucco-covered stair tower that appears to float, rather than sit on that base.
The problem is that you can find category errors almost everywhere, though. It makes it clear how difficult it is to make an internally consistent design given the pressures of costs, technology, and codes, not to mention what might be the less than adequate training of some architects. Like problems in punctuation in a text or inconsistencies in a politician’s speech, such mistakes or elisions might be expected and even acceptable, but they point to an underlying problem: If you can’t get the basic elements right, the whole logic and beauty of the building comes into question. Good architecture, no matter what its style, gets it right, down to commas and the consistency of what it says about the world.
Aaron Betsky is a regularly featured columnist whose views and conclusions are not necessarily those of ARCHITECT magazine nor of the American Institute of Architects.
