Digital Project

The Natural Enemy of the Librarian

Uplifting monument or waste of space? Philip Johnson’s Bobst Library and a conflict between professions, a shift from book warehouses to social hubs. Photographs by Andrea Geyer.


I have worked in Elmer Holmes Bobst Library and Study Center, New York University’s main research library, on Washington Square Park, for twenty-four years. I have always found the design of the building beautiful—more so than just about any other library building I have been in. Every day I walk through the revolving doors and gaze immediately upward toward the series of cascading bronze stairs, which ascend twelve stories. I stand in the ten-thousand-square-foot chasm, which is encircled with clinical precision by shimmering catwalks. The pattern of black, white, and gray marble on the ground floor resembles an Escher drawing viewed through the lens of the Italian Renaissance. The stark simplicity of the railings and the harsh, clean lines remind me of a Mondrian painting. My chest feels a little lighter and my head swims a bit, as when stepping into a cathedral and being drawn heavenward.

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".

Nearly every day, however, I hear someone complain that the atrium is a “waste of space.” This complaint goes back to 1965, when a group of head librarians from around the country were invited to review the architect Philip Johnson’s design. Among the librarians was Ralph Ellsworth, the director of libraries at the University of Colorado, who voiced his objections to Martin Beck, NYU’s director of planning. The enormous atrium meant that the floors would be U-shaped, which would minimize the amount of storage and inconvenience readers, he asserted. He called the design “a throwback to the 19th century conditions” and “a fantastic architectural anachronism,” comparable to Boeing putting “buggy whip holders on the front of a B-727.”

Ellsworth’s vitriolic letter set the tone, and librarians continue to vehemently denounce the building to this day. They allege that Johnson, like so many architects, failed to appreciate the purpose of the building or draw on the knowledge of librarians. They resent that the needs of researchers, and imperatives of storage and preservation, were deemed to be less important than the desire for grandeur and monumentality. And, unknowingly, they express an abiding tension between practical design and aesthetics, between librarians and architects, which has a curious history.

To make sense of the persistence of this narrative, I dug into the history of American academic library design and the discourse between architects and librarians. I sought to understand the epistemological implications of Johnson’s design of the library—which also is a social hub. Was Johnson moving away from modernist notions of architecture and the arrangement of books and documents? Was he inaugurating an organization of space that was centered on dynamic relations between people in space, rather than on monolithic systems of communication? Does Johnson’s focus on the interaction of people to create knowledge, instead of to preserve and protect culture, signify a shift toward the postmodern? What is the genealogy of the librarian-architect dynamic?

The idea that a library like Bobst might reflect a shift in the expression of knowledge was presented to me by David Kaser, who taught a class on the history of library architecture when I was a student at the school of library and information science at Indiana University in the mid-1980s. Kaser was the kind of scholar and librarian that was quickly disappearing. In addition to teaching courses on the history of printing in the United States, he researched library architecture and the status of libraries in the academy in the nineteenth and twentieth centuries. To those more interested in “informatics,” Kaser was an anachronism, but I found him fascinating for his knowledge of libraries across the country. He never looked beyond the administrative implications of building designs, but he opened my eyes to the possibility that libraries reflect the history of how we express knowledge.

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".

In The Evolution of the American Academic Library Building (1997), Kaser argues that libraries should be designed to preserve materials, facilitate functionality, and attain beauty, but he admits that “very few have done all three.” As Kaser explains, there were no academic libraries in the United States until the mid-nineteenth century. Early American colleges tended to be small and religious, so at Yale and Brown, for example, the library shared space with the chapel. (Round libraries, such as the one designed by Thomas Jefferson for the University of Virginia in 1826, were also common.) In 1840, the University of South Carolina erected the first freestanding library, a classical edifice with four imposing columns at the facade. The next were built, in the Gothic style, by Harvard in 1841 and Yale in 1846. Harvard’s library was modeled after Kings College Chapel, built in Cambridge in 1446, and Yale’s on Trinity College, built in Dublin in 1732; both chapel-like structures exemplify the influence of ecclesiastical architecture on library design. While the earliest continental libraries were rectangular with perimeter shelving, the Classical and Gothic revival libraries in the United States featured an “alcoved hall with double-faced book presses extending inward between the windows in the two longer halls,” writes Kaser. Some libraries added clerestories with galleries that allowed for more shelving and also gave the impression of cathedrals.

The only purpose of these early American libraries was to store books. Many used a classification system that was simply enumerative. Rather than books being grouped by subject, the handwritten catalogue often listed books seriatim, which was acceptable because the number of volumes was so minuscule. Eventually, alcoves were created to increase the amount of space available for shelving, which led to the adoption of general subject classifications.


Librarians and architects were already at odds in the late nineteenth century, when librarianship and architectural practice were being professionalized. (The American Library Association was founded in 1876, the American Institute for Architects in 1856.) Many librarians felt that architects ignored their needs and created buildings that emphasized grandeur over functionality. William Frederick Poole, the librarian at the Chicago Historical Society and the founder of Poole’s Index to Periodical Literature, was one of the most outspoken opponents of such designs, which he saw as wasting space and pointlessly imitating churches. At a meeting of the ALA in 1881, Poole delivered a fiery speech against the “vacuity” of the new Peabody Institute Library in Baltimore. “The nave is empty and serves no purpose that contributes to the architectural effect,” he argued. “Is not this an expensive luxury?”1

Poole went on to propose that additional floors be created and that books no longer be relegated to the aisles, which would allow for the storage of 717,000 instead of 150,000 volumes. He suggested that books be classified according to “four grand divisions or departments of knowledge,” with each getting a separate floor and reading room, accessible by elevator. Then he returned to the long-standing relationship between libraries and religious edifices:

Why library architecture should have been yoked to ecclesiastical architecture, and the two have been made to walk down the ages pari passu, is not obvious, unless it be that librarians in the past needed this stimulus to their religious emotions. The present state of piety in the profession renders the union no longer necessary, and it is time that a bill was filed for divorce. The same secular common-sense and the same adaptation of means to ends, which have built the modern grain-elevator and reaper are needed for the reform of library construction.

Poole’s writing expresses a particularly American practicality, which also characterizes the work of his colleague Charles Ammi Cutter, who for twenty-five years helmed the Boston Athenæum’s library. Cutter fostered interlibrary loan programs, advocated for the provision of proper lighting and control of humidity, and invented his own classification system, which influenced the Library of Congress and is used throughout the world. Despite the remarkable achievements of librarians like Poole and Cutter, their focus on functionality and pragmatism has hindered serious theoretical writing about librarianship and archival practice in the United States. How different might the situation be if they had been attentive to the ways in which social relations and cultural values might be communicated, even engineered, through the design of buildings that are not merely machines for the storage and circulation of books?

Of course, the ire of Poole and his colleagues was not just a response to the impracticality of architects. At the same conference, the architect J. L. Smithmeyer, who was designing the Library of Congress, fanned the flames by proposing exactly what Poole detested: a grand building in the ecclesiastical style that didn’t incorporate any of the ideas of librarians.2 Poole was livid; even the plans for the national library were being developed without input from librarians. But Poole wasn’t about to give up. After Smithmeyer’s presentation, he proposed that the ALA pass a resolution in support of “a radical modification to the prevailing typical style of library building, and the adoption of a style of construction better suited to economy and practical utility.” The resolution passed.

Poole was not the only librarian who bitterly fought with architects about building design. Cutter, as the editor of Library Journal, which Melville Dewey had founded as the trade journal of the profession in 1876, threw down the gauntlet. He declared: “The architect is the natural enemy of the librarian.”

Some architects responded to the criticisms and tried to ameliorate the situation. In 1888, the librarian William Isaac Fletcher published a letter in American Architect that acknowledged his colleagues were “perhaps a trifle long-haired,” appealed for dialogue, and calmly explained his profession’s central principles. Invited to speak before the ALA that year, the architect Normand Patton admitted: “It would be a singular mistake to plan a church and forget the convenience of the minister; and yet many a library has been planned, and apparently the librarian has been left out of the calculation.” A détente ensued and librarians began working with architects, if somewhat grudgingly.

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".


Between 1875 and 1890, as Kaser notes, university instruction began to shift. American universities came to favor seminars, which required students—especially graduate students—to refer to many primary and secondary sources. This resulted in a staggering increase in the collections of libraries, which required a new design approach. Architects began to use core stacks made of iron and, often, glass, which allowed for the flow of light. This model, taken from German universities, won the approval of Poole and his compatriots.

“Enter the Behemoths,” as Kaser calls the new libraries. The first iron book stacks were built in 1877 as an addition to Harvard’s Gore Hall. Partitions established separate spaces for books, seminar rooms, technical services, periodicals, and reference. Because of the size of the collections and the ways in which they were being used, new systems were required for organizing and locating books. The William Rainey Harper Memorial Library at the University of Chicago, completed in 1912, was the first to distribute the collections by subject and provide adjacent seminar rooms for students. Harper Library pointed to a new understanding of the shape of knowledge, as reflected by systems like Dewey Decimal Classification and Library of Congress Classification. In determining the subject of each volume, librarians also expressed the epistemes of their time.

After the opening of Harper Library, iron stack towers became the norm, as did the classification of books by subject. Nearly every major university library built between 1900 and 1945 had a stack tower, and the capacity for up to two million volumes. (The stacks were typically closed, meaning that pages retrieved books for patrons, who consulted them in grand reading rooms.)

This was the model for Butler Library at Columbia University, which opened in 1934. Following the Chicago model, the massive new structure was designed around a core stack with radiating reading rooms, organized by subject. The behemoths were incarnations of Poole’s grain elevator; they were storehouses for especially valuable commodities. At Butler, a prominent mural by the American artist Eugene Savage depicts Athena as defending knowledge—the cultural record of the upper classes—and the Columbia campus from the ignorant, who are portrayed as buff, dark, green-skinned monsters. Columbia, after all, retreated uptown to escape the squalor of the city and to build the “Acropolis of New York.” Occasionally, an undergraduate recognizes the meaning of the painting and complains, but most fail to see the implications of the allegory.

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".


In “The Design of the Bobst Library and the State of the Art Today” (1973), Johnson describes Butler as embodying “the revolution against the great room in favor of the warehouse.” Butler’s predecessor, Low Library, across the plaza at the center of campus, was one of the last great buildings in the old style, a neoclassical temple with a colonnaded facade. Designed by Charles Follen McKim and erected in 1895, the building established a sense of the university as a place, and of the library’s monumental staircase as the heart of campus life. In his essay, sent along with a letter to James Hester, the president of NYU, Johnson embraces “a new functionalism”; but he also recommends revisiting the great room in the form of an atrium. “For so many books, so many students, simple orientation requires a great orientation device,” he writes. “Just as a gridiron plan city requires Washington Squares and Central Parks, just as labyrinths like great urban stations such as Pennsylvania Station and Grand Central need huge center spaces, so libraries need space to find the way, space to relax the eye, space to recognize others in that lonely maze—a university library.”

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".

In his design of Bobst Library, Johnson was looking back to the magnificent edifices of the nineteenth century and the grandeur of ecclesiastical architecture. He believed that architects had gone too far in the direction of the utilitarian warehouse and neglected those who would inhabit the building. He wanted a building that could serve as the core of the university, a gathering place where people could “see each other working,” as Mervin Dilts, a professor emeritus of classics at NYU, told me.

In another letter to Hester, dated eight years earlier, Johnson explains his desire to “build a type of interior campus, a space which, in Charles Wrightman’s words, would give the undergraduate a feeling of, ‘This is a university, this is where I want to come back.’” The key to achieving this was the atrium, which would reach from the street to the ceiling, giving the building “a unity and a grandeur second to none in our city.” Johnson compares his design to Low Library, “which still today gives architectural focus and spatial grandeur to the Columbia campus.” And, in conclusion, he notes that he has “met all reasonable requirements of the library profession.”

The librarians didn’t agree. They elicited the help of David Tishman, a university trustee and vice chairman of the board of Tishman Realty and Construction Company. Tishman became an outspoken opponent of Johnson’s plan. He expresses his consternation in a 1965 letter to Hester: “You said the librarians’ views on the library planning are subordinate to those of the architect. I cannot accept this premise.” Tishman chides the architects for failing to consider the librarians’ opinions regarding the “function, maintenance, and operation of the library.” He assures Hester that a library with a conventional stack at the core will be one of “the world’s great educational buildings,” and “a structure incorporating the last word in the science of library design and functions.”

Tishman convinced Hester that other distinguished librarians from across the country should be brought in as consultants. Each of them rejected Johnson’s plan. William Dix, the head librarian at Princeton, declared: “I feel obliged to state flatly that the open area in the center of the building can be achieved only with the sacrifice of functional aspects of the building.” John Berthel, the chief librarian at Johns Hopkins, insisted on “the omission of a shaft similar to the Peabody.” Stephen McCarthy, Cornell’s librarian, strongly preferred “the omission of the shaft and mezzanines” in favor of the “loft-type space since it is preferable for any subdivision in the future.”

Tishman bundled all these comments into a report, again attacking Johnson: “The use of open space as designed results in being too public and bustling a place, which should be a place for quiet study. Here the business of the library is done.”

Johnson responded to the criticisms in a letter to Hester, writing, “We do not claim that a cheaper library could not be built. Most assuredly, a warehouse is cheaper than a palace.” Johnson argues that at NYU, which lacks a campus like Columbia’s, there is an imperative to “put our great space under glass, indoors,” in order to give the university “a sense of monumentality.” He continues:

I’m not at all ashamed to use the word, monumentality, glory, pride, but nonetheless you have those feelings whether one admits them or not. It’s only permitted now in churches and military establishments, but there is no reason why universities should not and could not have a central pride. They certainly did in the nineteenth century.

Hester had opponents not only at NYU but in Greenwich Village, where residents saw the building as a town/gown issue. Four people—including the author and activist Jane Jacobs and the congressman (and, later, mayor) Edward Koch—tried to hinder the construction by suing the city to prevent the necessary variances from being issued. In the end, the case was thrown out. Hester, who had stated that the library would be his most important legacy, and Johnson rallied their allies and persuaded Mayor John Lindsay to grant approval.

However, the troubles did not end. Construction was plagued by setbacks, incompetent construction agencies, the bankruptcy of the quarry that was supplying the red sandstone that Johnson chose for the exterior, and labor strikes. Bobst finally opened on September 12, 1973, nearly ten years after being designed.

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".


When Kaser remarked that very few libraries had managed to shelter their contents and facilitate use and enhance aesthetics, he was summarizing the debate around the purpose and functionality of libraries—which continues to inform the criticism of Bobst. The vitriol expressed by Poole in his speech against the Library of Congress is not so different from what I’ve heard from librarians as I’ve mentioned my research on the subject. They’ve all taken the opportunity to tell me how much they hate the building, and how the atrium is a “waste of space.” On the other hand, tourists and local visitors often stop by the library to see the atrium and take photographs and selfies. Like NYU students and staff, they experience the “monumentality” and “pride” that Johnson was after. And they get the “sense of awe” that was articulated by students in 1973 in the Washington Square Journal. “Although the atrium has often been criticized as wasting space, the building simply would not be half as stunning as it is without it.”

Andrea Geyer, Passing Lines, 2017, series of digital C-prints, each 15" × 10".

Even as Bobst has become the center of the campus life, NYU has outgrown the library, which more than ten thousand people pass through each day. No library remains sufficiently spacious after thirty years, and Johnson’s design predates the personal computer; the digital age has turned the library into an entirely different beast.

That said, Johnson enabled patrons to see one another and be integrated into the shelving; to engage in the creation of knowledge and not just access a warehouse. Rather than fix subjects in place, the library encourages them—and not just books—to circulate and relate to one another. As critics are quick to point out, Bobst, like so many contemporary libraries, is a space for social interaction, not just quiet contemplation. That, however, is now the nature of research, teaching, and knowledge. Yet, to some, the building feels like a panopticon, with each visitor aware of the possibility of being watched at all times, if not by centralized guards then by other library users. This new criticism of the atrium, though cast in postmodern—and, specifically, Foucauldian—rhetoric, perfectly fits within the long-standing librarian-architect dynamic.

Without quite knowing it, Johnson seems to have moved from the modern to the postmodern and marked an epistemological shift—which is apparent in our treatment of knowledge as well as our design of libraries. In recent renovations, the university has added classrooms, collaborative study spaces, and research commons, which have made the building more and more “user-friendly.” Johnson foreshadowed these changes in his desire to make the library a place of activity and not a warehouse for knowledge.

Of course, other changes to Bobst have coincided with an effort to combat suicides after students flung themselves from the upper floors to the atrium in 2003 and 2009. These deaths pulled the building back into the modernist camp. Johnson’s genius design of the atrium allowed for the ease of the suicides. His vision overshadowed the needs of humans living with the building. Fortunately, the architectural intervention to prevent suicides maintains the grandeur and even amplifies the symbolism of Johnson’s original design. Architects hired by the university enhanced renderings that Johnson made for panels to be installed at the New York State Theater, creating thirty different designs. These were turned into massive bronze cutouts and installed in the atrium to prevent anyone from jumping. Today, looking upward from the lobby at the atrium, you not only have the feeling of awe but the impression of glittering waterfalls of digital data flowing from the top floor.

1 Poole later published his speech as an article, “The Construction of Library Buildings,” in the April 1881 issue of Library Journal.

2 Smithmeyer’s presentation was published in the same issue of Library Journal as Poole’s.