When Art Worlds Collide: Meredith Monk’s Atlas and the Institutional Limits of Operatic
Experimentation
Ryan Ebright
In May 1990, Houston Grand Opera general director David Gockley penned a lengthy, emphatic
letter to Barbara Dufty, the executive director of Meredith Monk’s House Foundation for the Arts.
Monk, a downtown New York City interdisciplinary artist, by the 1980s had established a reputation
as an innovative and unorthodox performer, composer, director, and choreographer, known
especially for her visually arresting site-specific multimedia productions and wordless vocal
compositions. She had then parlayed the success of her 1979 Dolmen Music into an opera commission
with Houston Grand Opera (which I’ll refer to throughout as HGO). However, with less than nine
months to go before the premiere, Gockley and other HGO personnel were concerned. Costs were
spiraling, co-producers were getting anxious, and the opera still lacked a definite title, musical score,
and even plot. Whatever did exist was embodied in Monk and a few other performers. “We must
deal with reality,” Gockley wrote to Dufty, “and then go forward to the next phase: that of
Meredith’s unencumbered creativity within the parameters that we jointly have established.”1
These parameters, hammered out over three years, exemplify the relationships forged
between many avant-garde artists and established opera companies during the twentieth century’s
closing decades, a so-called “golden age” of American opera. While artists gained access to new
institutional resources and companies capitalized on these artists’ prestige, their different production
methods and aesthetics sometimes collided. The Monk-HGO relationship is a case in point. The co-
commissioning contract for Atlas, reached in November 1989, allowed for sixteen weeks of group
rehearsal—an allotment that constrained Monk’s usual collaborative creative process while marking
an unprecedented personnel expense for HGO.
In my paper this morning I trace the multi-year development of Atlas, focusing especially on
frictions that arose between Monk (represented by the House Foundation) and HGO. I see the Monk-HGO partnership and its frictions as a collision of art worlds, each of which held different
sets of conventions. These areas of resistance, occasioned by Monk’s creative processes, help to
reveal the ordinary and thus often hidden workings of opera production. As I’ve argued elsewhere,
the tensions between artist and institution—which first played out in the quotidian realities of
contract and post-contract negotiations and later in the creation of the opera’s score—reflect
different aesthetic positions on the role of the body in opera. For Monk, the human voice and body
are generative, central not only to operatic performance but also to its formation.
My concern here today, though, is less with the underlying aesthetics of the art worlds
represented by Monk and HGO, and more with the conventions that stem from these aesthetics.
Conventions, Howard Becker reminds us in his sociological deconstruction of art worlds, “provide
the basis on which art world participants can act together efficiently to produce works characteristic
of those worlds.” 2 By the standards of American opera production in the late twentieth century,
Atlas was spectacularly inefficient. But it was also unique among its peers, showcasing an astonishing
variety of vocal timbres and textures along with an idiosyncratic approach to narrative and
movement. It also demonstrated a different way of doing things, one that could produce innovative
results. By and large, though, the art world of American opera has not embraced the patterns of
cooperation modeled by Atlas. Too unorthodox for opera houses and too operatic for
multidisciplinary arts presenters, the opera would have to wait more than a quarter of a century for a
revival, in a venue outside the opera house. Ultimately, close examination of Atlas and its genesis
reveals how institutional identity, as reflected through conventions, might inhibit aesthetic
experimentation within the opera house.
For much of the twentieth century, one primary barrier to new opera was money. Although
a surge of funding fueled the exponential growth of opera in the United States during the postwar
period, much of it went toward the establishment of local and regional companies, as well as service
organizations like Opera America and the National Opera Institute. Relatively little went to new
opera. This began to change in the late 1970s with the formation of the NEA’s Opera-Musical Theater program, which actively promoted new American works, sometimes in conjunction with the
aforementioned organizations.
Monk’s Atlas owes its existence, in part, to Opera America’s pioneering “Opera for the
Eighties and Beyond” initiative. As Sasha Metcalf has detailed, a handful of opera administrators and
philanthropists in the early 1980s—concerned about the dearth of new American operas and a
stagnating repertory—hoped to imbue opera with new life by tapping into the world of avant-garde
music and theater.3 Pointing to Philip Glass’s successes with Einstein on the Beach and Satyagraha as
models of this successful synthesis, they launched a program that provided funding for new works at
various developmental stages. When Opera America launched “Opera for the 80s and Beyond” in
December 1983 in New York City at its annual convention, it also provided attendees with a taste of
the avant-garde. At a concert titled “New Directions in Opera” that was curated by Mary Macarthur
of The Kitchen and the Brooklyn Academy of Music’s Harvey Lichtenstein, attendees heard pieces
by Anthony Davis, Robert Ashley, Bob Telson, and Meredith Monk. There, Monk and her ensemble
performed Dolmen Music, a piece for vocal sextet, cello, and percussion.
David Gockley, then general director of HGO, recalls that he “responded positively” to
Dolmen Music when he heard it at the conference. He told me that he found it “amusing and
structurally interesting, listenable,” but also “radical.”4 This prompted him to initiate a working
relationship with Monk, with an eye toward an eventual commission. Under Gockley’s leadership,
HGO had by the early 80s begun to establish a reputation as a fledgling haven for new opera, with
world or American premieres of works by Carlisle Floyd, Leonard Bernstein, and Philip Glass.
Gockley was eager to expand his pool of composers. First, however, he wanted a trial run with
Monk via a two-week workshop with HGO’s young artist studio. In a May 1985 grant application to
Opera America, Gockley wrote:
The purpose of this project is to see whether a) conventionally-trained opera singers can
assimilate Meredith Monk’s musical style and techniques, b) whether Monk can work within
3 Metcalf, “Institutions and Patrons in American Opera: The Reception of Philip Glass, 1976-1992” (Ph.D. diss., University of California at Santa Barbara, 2015), 1.
the confines of a “conventional” opera company/music theater organization, and c) whether
Monk has an aptitude for music theater unrelated to her abilities as a performer. To put it
another way, is a collaboration between Meredith Monk and an opera company a possibility
for the future.5
Here, Gockley identifies two key conventions in play: operatic singing and what I’ll refer to broadly
as operatic production. Together, they required reciprocal transformations—opera singers into
Monk singers; Monk into a more conventional operatic creator. The latter was perhaps the more
important transformation. In a letter inviting the HGO presidential council to the September 1985
performance that resulted from the workshop, Gockley pointed to Philip Glass’s success at moving
beyond his own ensemble and entrusting his music to a producing opera company. (HGO had just
given the American premiere of Akhnaten.) Gockley intended to see if Monk could do likewise. “Up
to now,” he wrote, “her stage works have been performed by her and/or her hand-picked group of
collaborators.” If the experiment was successful, Gockley noted, “we might have a new composer
for the opera world. If not, we can be consoled by 6 La bohème on October 17.”
Gockley’s gambit paid off. The September concerts, which featured solo works performed
by Monk as well as Dolmen Music performed by HGO studio artists, were well received. More
importantly, the two-week workshop had been a success. In a later reference letter, Gockley wrote,
“When she came to work with our Houston Opera Studio, we fully expected to find our classically
trained young opera singers huddled in a corner with hands wrapped protectively around their
golden throats…The rapidity with which she won them over and got them to participate in a staging
of her work for singers and solo cello called Dolmen Music testifies not only to her powers of
persuasion, but also to her winning and unpretentious personality.”7 Yet as I’ll demonstrate, Monk
never fully assimilated into the traditional operatic production model in the way that Glass had. The
second criterion that Gockley identified earlier—the ability to “work within the confines of a
5 Gockley, OFTEAM special circumstances grant application, 8 May 1985. HGO archives. 6 Gockley, letter to HGO presidential council, 19 September 1985. HGO archives.
‘conventional’ opera company”—proved to be a transformation that Monk was unable or unwilling
to undertake.
But I’m getting ahead of myself. The HGO studio artists had successfully mounted Dolmen
Music, and the path was thus clear for a commission. Over the course of 1986 and 1987, the House
Foundation and HGO engaged in preliminary discussions. Dufty, writing for Monk, offered hints of
what the opera, tentatively titled Ghost Stories, might be about:
One [idea] is for a ghost story, either contemporary Japanese or from Hassidic literature. The
work also could be structured as three separate stories. Another idea is for a piece on the
desert fathers and their temptations, which could incorporate ghost tales as well. The last
idea is for an opera on La Malinche’s betrayal of her people to the Spanish conquerors.
Meredith would like to mount the work in a workshop fashion over a period of time.8
Many of the elements suggested in this description ultimately found their way into Atlas. For our
purposes, however, I want to draw a connection between that last sentence and those that precede
it. Absent the bodies and voices of the performing artists—which could only be present in
workshops—Monk could only offer possibilities of what the opera might be, but nothing more
concrete than that. HGO personnel were quick to express concerns over her working method. In
response to Monk’s request for four three-week rehearsal periods, HGO artistic administrator and
dramaturg Scott Heumann asked Dufty, “Does she really need all this time?”9
Monk finally received a formal commission offer in August 1987. In addition to laying out
several parameters, Gockley proposed an initial four-week workshop period with the singers and a
six-week rehearsal period leading up to the proposed premiere date of Fall 1990. He also noted that
the ensemble would blend House and HGO performing forces: five or six of Meredith’s usual
performers, supplemented by HGO singers. These parameters—personnel and rehearsals—would
prove to be key points of contention, and over the next three years, while Monk began early
conceptual work on the opera, Dufty and Gockley hammered out the details of the commission,
8 Barbara Dufty, letter to David Gockley, 30 April 1986. Box 34, folder 3, Meredith Monk archive, NYPL. 9 Heumann, letter to Dufty, 29 September 1986. Meredith Monk Archive, LPA Mss 2006-001, New York Public Library. which eventually came from HGO as well as Minneapolis’s Walker Arts Center and Philadelphia’s
American Music Theater Festival.
Frictions between Monk’s and HGO’s working methods first appear in drafts of the
commission. One draft, from August 1989, shows Dufty requesting double the number of singers
intended by HGO (38 instead of 19), because the size of the cast remained unknown to Monk. 10
Elsewhere in the same draft, Dufty pushed for longer workshop rehearsal periods, demanding an
initial eight-week period in Houston in May/June 1990 followed by another eight weeks in
September/October, rather than the total of 10 weeks proposed by HGO.
Monk largely won out in the final co-commissioning agreement, dated 30 November 1989.
In it, she was contracted as both composer and director. Dufty said that because it was Monk’s first
major contract, she and other House board members were adamant about getting everything that
Monk wanted. Terms that specified the size of the cast and orchestra were removed entirely in favor
of language that suggested Monk had free reign in these areas as long as she remained within budget.
The final version also removed language that gave the co-commissioners equal rights in determining
the casting. Most significantly for Monk, the contract guaranteed a total of sixteen weeks of
rehearsals in Houston, broken into three periods over the twelve months leading up to the premiere.
(For comparison, most new operas at that point might get six-to-eight weeks of rehearsals
immediately before the premiere.)
Monk began working extensively with artist Yoshio Yabara on the “conceptual design” of
the opera in September 1989, just before the commission was finalized. Rather than working with a
libretto or even a general story—typical starting points for opera composers—Monk would voice
ideas to Yabara, who would visualize them on paper. For Monk, composing meant improvising
vocally while accompanying herself on the piano, recording those ideas, and developing them
further, at times with her Fostex recorder’s multitrack capabilities. In the remaining months of the
year, Monk developed material for the duet Facing North with singer Robert Een, which would
eventually serve as material for the Arctic Bar scene in Atlas. She also held preliminary workshops with some of her existing vocal ensemble members in which she “concentrate[d] on the staging for
the production, visualizing her ideas on people rather than paper.”11
Yet because of Monk’s process, she couldn’t truly begin creating the piece until she had
assembled her cast. Between January and May 1990, Monk auditioned more than 400 people.
Although more conventional opera auditions began in Houston, Monk soon realized this format did
not meet her needs. In Houston, she reflected, “people seemed to have a kind of disembodied
voice.” 12 Instead, Monk auditioned performers in groups of twenty at her Tribeca loft in New York.
These two-hour audition-cum-workshop sessions involved extended vocal and physical warm-ups,
learning music by ear, and other collaborative, improvisatory exercises. According to singer Katie
Geissinger, who joined Monk’s vocal ensemble for Atlas and has remained a key member since then,
this workshop format allowed Monk to “see how you performed over time, how you collaborated,
how you improvised, how quick you were to pick things up, vocally and dramatically and
movement-wise.”13 Geissinger adapted quickly, but other singers were less comfortable with this
unorthodox format. Pablo Vela, a longtime Monk collaborator who served as assistant director for
Atlas, noted a somewhat common reaction among the more conventionally-trained auditionees
when told that neither the opera’s roles or its music had been determined yet: “You’d sort of see this
wild eye starting back and forth, like, ‘Let me out of here.’”14 After a five-month audition and
callback period, Monk eventually settled on a company of eighteen performers, only three of whom
she had worked with before.
Even with the contract negotiated, many details remained to be clarified. Records from a
meeting with Dufty and the various co-commissioners in February 1990 hint at these difficulties,
citing “many different opinions and misunderstandings regarding this project.”15 The budget worried
everyone. Union role classification and pay issues also posed a difficulty. The House Foundation had
11 House Foundation, NEA Phase III grant application, Summer 1989. Series V Box 34, Folder 4. LPA Mss 2006-001, New York Public Library.
12 K. Robert Schwarz, “Remaking American Opera,” Newsletter: Institute for Studies in American Music XXIV:2 (Spring 1995): 14.
13 Katie Geissinger, email correspondence with the author, 15 February 2017. 14 Pablo Vela, interview with Nancy Smithner, 3 July 2002. MGZMT 3-2378, New York Public Library. always used a favored nations approach, which meant that every performer was paid the same
amount, no matter how long they had been with the company.16 AGMA union rules, however,
stipulated different amounts for different role classifications. Jim Ireland, HGO’s managing director,
suggested in exasperation, “perhaps we should just give $500k to [Monk] and say, see you in
Houston in February. Please have a show ready!”
In the end, that’s pretty much what HGO did, handing over control of most of the budget
to the House Foundation, allowing it, rather than HGO, to determine how much money should go
toward sets, costumes, rehearsals, and so on. To stretch the budget, Monk held rehearsals in her
New York City loft, which allowed her to not pay per diems to the New York-based cast.17 The
amount of money spent on the rehearsal process necessitated tightening in other areas, especially in
terms of the sets and costumes.
Those sixteen weeks of group rehearsal, though, were arguably well-spent, and marked
Monk’s most significant departure from the conventions of operatic production. According to
countertenor Randall Wong, who sang as the Lonely Spirit as well as one of the Ice Demons,
different scenes from the opera developed out of different working methods. Many of the ideas in
the Ice Demons scene, he said, came out of rehearsal improvisations on melodic ideas that Monk
introduced, whereas the Lonely Spirit scene was more fully developed by Monk before she
introduced it to Wong. Based on the cassette tapes in Monk’s archives, it appears she would then
solidify or develop ideas further using her own voice. In tapes dating from August and September
1990, for example—after the initial summer workshop period—we can hear Monk singing all three
of the Ice Demon parts.18 While the form largely matches that of the recorded version, there are also
small differences in phrasing, timbre, and vowel vocalizations. [play excerpt] This is just a small
sample, and the nearly 15 hours of Atlas cassette tapes—some featuring Monk alone, others with the
16 Dufty, phone interview with the author.
17 Gockley, phone interview with the author.
18 Meredith Monk, “Opera—Additional Material,” Cassette Tape ID 237276, September 1990. LPA Mss 2006-001, New larger ensemble—show that Monk consistently maintained an attitude of experimentation
throughout the creative process, even as she began shaping the material more firmly in later stages.
The four months leading up to the premiere on February 22, 1991 were a flurry of activity.
The opera underwent a title change during this time, from Ghost Stories to Atlas, in order to better
reflect the narrative that had taken shape over the course of the rehearsals, particularly during the
final period. A November 1990 HGO memo warned ominously, “Per Emily, only three pieces have
been worked on. Nothing is written down. She cautions us about whether or not the piece will be
ready come February.”19 Although we can and should read this memo with a grain of salt given
other evidence, the score likely was far from ready. A memo from a few weeks earlier confirms how
much remained to be done: “Meredith doesn’t see how she can get us the orchestral parts by
February 1.” And, in fact, the music score, normally an early step in the process of an opera’s
creation, was ultimately the last. Wayne Hankin, the music director for Atlas, revealed that the
creation of the score—at least for the orchestra—came down to opening night: He said, “Somebody
asked me, ‘When was the score actually finished?’ as what they heard, and the answer was, pretty
much, somewhere between first and second act [on opening night].”20
Monk’s unconventional creative process produced an unconventional opera. Like so many
composers of the late twentieth century, Monk framed her opera as an attempt to “redefine the
parameters” of the genre. The story of the opera, inspired loosely by the life of the early twentieth-
century Belgian-French explorer Alexandra David-Néel, follows a young woman over three acts as
she leaves home, undertakes a journey with companions to a number of geographical and spiritual
waypoints, and ultimately ascends to a higher plane of existence before returning home. In light of
Monk’s previous large-scale music-theater works, the linearity and conventionality of this hero’s-
journey narrative was arguably her largest concession to opera art world. The means by which the
narrative is conveyed, however, are more radical, as you heard a minute ago. The plot is conveyed
19 Jim Ireland, HGO production memos, 14 November and 26 October, 1990. HGO archives. primarily through movement and vocalise, which is in turn supported by a ten-player ensemble of
strings, synthesizer keyboards, horn, clarinets, and an assortment of winds.
Another substantive concession Monk made to operatic convention was the performance
venue. Debby Lee Cohen, the co-set designer with Yoshio Yabara, said that when she and Yabara
first visited Houston Grand Opera, they immediately thought that the opera should be performed in
the lobby rather than the Cullen Theater. “That opportunity didn’t exist, but that’s really where we
should have done Atlas,” Cohen said. “For years after we did Atlas in Paris at the Odeon, I dreamed
that we did it in reverse, with the audience sitting on stage and we would be performing it with some
sort of crazy set on the balcony.”21 In fact, Monk (via Dufty) had asked HGO about the possibility
of an alternative performance space early on, but nothing seems to have come of it. A design like
that would have been more in keeping with Monk’s and Cohen’s previous work, and Cohen noted
that the design meetings, like the rehearsals, were incredibly labor-intensive, often lasting hours, as
the team tried to figure out how to dialogue with the performance space. Performer Robert Een told
me that in some ways Atlas “was squeezed into a proscenium,” perhaps to its detriment.
Despites all of these difficulties, in the end both Gockley and the House Foundation viewed
Atlas as a success. For all the tensions that exist in the documentary record, twenty-five years after
the fact Gockley recalled Atlas being, “a remarkably problem-free project,” and multiple cast
members I spoke with described him as wonderfully supportive, believing in their work throughout
the process. Indeed, Gockley’s vision and support seem to have overridden—or perhaps
alleviated—the concerns of other personnel at HGO. But his willingness to give Monk artistic
freedom, within limits, did come at some cost, beyond the obvious loss of budgetary management
oversight. He estimated that 20% of their regular opera patrons walked out during the first
intermission each night. (In fact, local music critic Charles Ward enjoined his readers, “Don’t Leave
Early.”) If true, that meant that Atlas played to an auditorium less than half-full every night; most
performances barely filled one-third of the hall. These consequences, however, were acceptable for
Gockley, who through his commissions was principally invested in, in his words, “bequeath[ing] an artistic legacy to the future,” which Gockley did by cultivating what he called “risk-intensive arts
philanthropy.”
HGO and the House Foundation struggled to find additional venues in the U.S. beyond
those of the co-commissioners and the Brooklyn Academy of Music. Arts presenters like UC
Berkeley’s Cal Performances, the Kennedy Center, and others who otherwise supported Monk’s
work shied away from Atlas, and efforts at booking a 1994–95 season tour failed to materialize. Ella
Baff at Cal Performances explained that it was “not the right piece to support for [Monk],” citing
the huge costs involved; instead, they were interested more in Monk’s solo or site-specific work.22
Atlas, then, suffered from a lack of distribution channels or venues; it was, as I suggested earlier, too
operatic for multidisciplinary arts presenters and not operatic enough for opera companies. Instead,
Monk’s vocal ensemble toured a concert version of Atlas—a strategy also adopted by other creators
of iconoclastic American opera in the late twentieth century, such as Anthony Davis and Steve
Reich.
Given that Atlas and its process of creation upended conventions for both HGO and Monk,
what did each party gain from the partnership? Monk arguably saw the most benefit, allowing her to
retain her working methods even though it came with small concessions in terms of narrative and
venue. Most importantly, the partnership allowed her to work on a larger scale, with more resources
than she was accustomed to. Atlas gave her a taste for a more varied instrumental palette, and also
established the new vocal ensemble with whom she would work for the next two decades. But, by
adhering to her normal process, Monk arguably limited the ability of future artists to perform Atlas.
Monk’s method, in which the score was a byproduct of the performance, had ramifications for the
opera’s reproducibility. The score that did eventually exist was partial, lacking or approximating most
of the vocal parts, and of course it failed to inscribe the relationships between music and movement
that are fundamental to Monk’s work, as all scores must.23 It would take more than twenty-five years
22 Memo. Box 33, Folder 2. Meredith Monk Archive, LPA Mss 2006-001, New York Public Library. 23 Meredith Monk, ATLAS, “Ice Demons,” working score. Series VI, Box 60, Folder 2. LPA Mss 2006-001, New York Public Library. for a usable score to be created that could serve as the basis for a new production, under Yuval
Sharon, which Megan will talk a bit about shortly.
For HGO, the benefits were less obvious, and the relationship entailed more drastic changes
to its conventions, arguably even to its identity. Given the control that Monk maintained over all
aspects of the production, in many ways HGO functioned more as a presenting organization that a
producer. In terms of the artist-institution relationship, Monk’s Atlas at HGO was more like Glass’s
Einstein on the Beach at the Metropolitan Opera than, say John Adams’s Nixon in China at HGO.
HGO had relatively little control over casting, personnel, scenic design, and other parameters;
instead it mainly provided funding and a venue. I see this as being at odds with their identity as a
leading producer of new American opera, one fashioned through dozens of premieres throughout
the 1980s and 90s. Monk’s creative conventions meant that other opera companies, had they wanted
to program Atlas, would have functioned similarly—and here it’s worth remembering that all of
Atlas’s co-commissioners besides HGO were arts presenters or festivals, not traditional opera
companies.
The history of Atlas helps to highlight just how much the few success stories of the
American opera renaissance of the 80s and 90s are predicated upon a traditional production model.
In the last decade, though, some U.S. companies, like L.A. Opera and Opera Philadelphia, have
adopted a more bifurcated identity, acting sometimes as a traditional producer, at other times as a
presenter through partnerships with independent producers like Beth Morrison Projects. If opera
creators and institutions today wish to cultivate new forms and means of musico-theatrical
expression, they would do well to begin by questioning their conventions, perhaps leaning into their
inefficiencies, as composer Neo Muyanga suggested in a keynote this past summer. Such a process
might be transformative.