The most talked-about Canadian work of architecture of the last decade is located not in Canada but in France: the new Opéra Bastille in Paris, designed by the Toronto architect Carlos Ott. From the beginning, the Opéra Bastille was the subject of lively controversy. For one thing, critics were skeptical about the whole idea of a new opera house because Paris already had two venues for lyric performances, the Opéra Comique and the famous Paris Opera, also known as the Palais Garnier. The Garnier, an 1875 Second Empire building, does have some technical drawbacks, but it is widely admired and loved. There were also those who maintained that the new opera had been built in the wrong place, that it should have been located where Pierre Boulez wanted, as part of the new music complex at La Villette, rather than being shoehorned into the cramped and awkward site of an old railway station in a working-class district, beside the Place de la Bastille.

However, other than town-planning considerations had led to locating the new opera on this historic spot that every year is the site of a popular street festival. When the idea of building a new opera house had been proposed to the government in 1981, it was argued that the Palais Garnier was an old-fashioned, elitist institution and that there was a need for a more progressive opéra populaire, hence the symbolic (cynics would say public-relations) import of the Bastille site.

The idea of a people’s opera probably appealed intellectually to the socialist president Mitterrand, even though he is not known to be an opera lover, but the concept is a mushy one. It’s true that French opera could do with a boost – it does not currently rank high with the French public – and since 1945 the Paris Opera has slipped from the first rank to mediocrity. But would a new hall really make a difference? Wouldn’t that be like trying to save a corporation from bankruptcy by building a new headquarters? And just because the Palais Garnier has chandeliers and gilt, does that really make it elitist? After all, in Italy, where opera has a mass following, it’s presented in neoclassical buildings like Milan’s La Scala, which was inaugurated in 1778, or Venice’s La Fenice, which opened in 1792. In any case, judging from the international celebrity of star opera singers like Beverly Sills and Luciano Pavarotti and the prominence of opera on public television, opera – that is to say, classical opera – is arguably the most popular of the fine arts. This raises another contradiction: the proponents of a “people’s opera” have argued that it would present a more modern repertoire and would not rely on the international star system, yet it’s precisely the nineteenth-century operas and the superstars that the general public desires.

If Parisians were lukewarm to the Opéra Bastille, it might also have been because of a growing sense of exasperation. It was not merely a question of the building’s cost, which the government now admits was not $540 million but at least $775 million. The Opéra Bastille was ill-starred from the start. In 1984, for two months, Jacques Chirac, the right-wing mayor of the city of Paris, refused to grant a building permit for the left-wing president’s new opera house. In 1985, the newly appointed artistic director, Jean-Pierre Brossmann, resigned, apparently unwilling to bend to one of the exigencies of a people’s opera – fewer rehearsals and more performances. In July 1986, the building site was shut down completely for two weeks; political wrangling had broken out again between Chirac, newly elected as prime minister, and Mitterrand, and it threatened to scuttle the opera completely. In 1988, Mitterrand won a second term as president, the socialists were returned to power, and a plan to build a reduced version of the Opéra was revived – it remained to complete the building for its opening on Bastille Day, July 14, 1989, the bicentennial of the French Revolution. Then, in January 1989, the Israeli conductor Daniel Barenboim, who had been named artistic director only two years before, was abruptly fired; his programming ideas had been judged too “elitist” (Barenboim had proposed Mozart!). His dismissal caused an international stir: prominent conductors such as Herbert von Karajan, Zubin Mehta, and Sir Georg Solti said that they would have to reconsider their association with the Paris Opera; Pierre Boulez, the director Patrice Chéreau, and the singer Jessye Norman (who was to sing at the inaugural) all resigned in protest. “What’s the difference between the Titanic and the Opéra Bastille?” went a Parisian joke. “The Titanic had an orchestra.”

Well, the Opéra didn’t sink, and it did acquire a new conductor, albeit not a famous one: Myung-Whun Chung, a young Korean-American previously best known as the younger brother of the violinist Kyung-Wha Chung. But Parisians were still not satisfied. I had the feeling that what most disturbed the people I talked with about the new opera house was the architecture itself. On this there was general agreement: the Opéra Bastille was too big for its site, it was an awkward composition, it lacked style and grace (Le Monde had called it “a rhinoceros in a bathtub”), it was, in a word, moche – ugly.

I went to see for myself. There is no question that the site chosen for the new opera is too small. The Place de la Bastille, a historic spot but not a very attractive urban space, lies between the Marais, a seventeenth-century quartier that has recently been restored, and the twelfth arrondissement, a gritty, working-class neighborhood. The massive Opéra in this residential landscape resembles a beached supertanker. The chief feature of the main facade facing the Place de la Bastille is a colossal curved wall, partly of glass and partly of stainless steel panels. The main entrance is located in the middle of this wall and is approached by a large exterior staircase. The staircase, as well as a forbidding square arch sheathed in black granite, is slightly askew to take into account the commemorative Colonne de Juillet in the center of the circular Place de la Bastille. (The column commemorates the Parisians who fell in the popular uprising of July 1830, which led to the downfall of Charles X, not the destruction of the notorious prison, which occurred in July 1789.) From the Place, the building stretches back along the rue de Lyon for more than two hundred meters, an undistinguished collage of columns, office-building-type glazing, and blank walls, interrupted by a curved volume that marks an experimental performance space that is as yet unfinished.

So tight is the site that there is no space from which the new building can be seen to advantage, except perhaps from the base of the column, were one courageous enough to brave the hazardous traffic. To make matters worse, the main facade of the Opéra is partially obscured by a small, undistinguished building housing a brasserie. At the time of construction, historians believed that a nineteenth-century building on this site had originally been a seventeenth-century neighbor of the Bastille prison. This turned out not to be the case, but by then the building had been torn down, so a replica, based on an old engraving, was built in its place.

What about the architecture of the Opéra? Carlos Ott has described it as “a functional project which is not essentially aesthetic.” Indeed, as much as such a thing is possible, Ott has reduced the aesthetic experience to a minimum. This is a building in which everything that is not granite is stainless steel, everything that is not white is black, and everything, absolutely everything, is obsessively arranged according to a square grid – the window mullions, the seams of the granite slabs and the stainless steel panels, the joints of the paving, even the supports of the railings. The same graph-paper motif and the same palette, if one can call it that, are continued in the interior.

The lobbies are located immediately behind the curved glass wall and take advantage of the view in a manner common to many modern concert halls like Place des Arts in Montreal and Roy Thomson Hall in Toronto. But neither of these buildings enjoys much of a view. At night, the homely Place de la Bastille achieves a magical quality with its spotlit column topped by a gilt Hermes, and Ott’s chief architectural conceit becomes apparent: to establish a dialogue between the building and the square by emphasizing the transparency of this huge building. I hadn’t much liked the Opéra during the day, but nighttime improved it; if not magical like the Place, it at least managed to appear dramatic.

The heart of an opera house, at least for the audience, is the hall itself. The greatest constraint on the design of any performance space is its size: the greater the number of seats, the more difficult it is to achieve visual and acoustic intimacy. Some postwar opera houses, like Berlin’s Deutscher Oper, which was built in 1961, have limited their capacity to fewer than two thousand seats, which happens to be about the size of La Scala (2,015) and the Palais Garnier (1,991). At the other end of the scale are enormous modern halls like New York’s Metropolitan Opera, which can accommodate 3,800 persons. At 2,700 seats, the Opéra Bastille steers a middle course. Although there are several tiers of loges, the layout, unlike the horseshoe-shaped La Scala or the Palais Garnier, is predominantly frontal, with two steep balconies.

I attended a performance of Arthur Honegger’s dramatic oratorio, Joan of Arc at the Stake, a moving if necessarily lugubrious work, whose gloomy atmosphere was heightened by the sight of drably attired actors and singers slogging across a stage that was covered ankle-deep in what appeared to be mud. The music, however, was glorious, and with an expanded orchestra and an eighty-five-voice choir it easily filled the cavernous space. From the first balcony, where I was sitting, the stage was far away, but the sound was good, at least to my amateurish ears. (I have been told that there are some acoustical blind spots among the front rows in the orchestra.) I asked Arthur Kaptainis, the music critic of the Montreal Gazette, what he thought of the acoustics. “The Opéra Bastille has what you could call a modern sound: clear but not especially resonant,” he said. “I thought that the sound lacked warmth,” he added, “but perhaps that’s a psychological reaction.”

What Kaptainis was referring to is the cool decor: the walls covered in gray granite and black wood, an undulating ceiling of white glass, and seats upholstered in black fabric. It’s true that decor matters little when the lights are out, but an opera house should not merely function as a background to the spectacle; it should create an atmosphere of anticipation. To say that “the place looks like a gymnasium,” as the soprano June Anderson remarked after singing at the opening, is perhaps ungenerous, but the interior of the Opéra is distinctly impersonal – imperturbable and sleek in a corporate-boardroom sort of way, which perhaps reflects the architect’s previous experience managing projects for a real estate developer.

The Opéra Bastille is obviously intended to be a modern rethinking of the traditional opera house, but in turning away from la grande cuisine bourgeoise of the Palais Garnier, Carlos Ott has eschewed nouvelle cuisine and instead has provided the Parisian public with the architectural equivalent of bread and water. Moreover, because many of the details are crude and the workmanship is sloppy, the bread is not even a crusty baguette; this is American-style sliced bread.

If truth be told, American style, or at least American expertise, is what the jury that picked Ott’s project as one of 3 finalists from among 756 entries in an international architectural competition thought it was getting. According to Michèle Audon, director general of the state body that oversaw the Opéra Bastille project, several of the jurors voted for the Ott project assuming that its anonymous author was the renowned American architect Richard Meier, to whose retro-modern style Ott’s entry did bear a superficial resemblance. (Meier has since built the Parisian headquarters of a cable television company; the result suggests that a Meier opera house would probably have been just as monochromatic but carried out with a lighter touch than Ott’s unwieldy design.) In fact, Meier had entered the opera competition but was eliminated in the first cut, together with other architectural stars such as Charles Moore, Kisho Kurokawa, and the Miami firm Arquitectonica. As designers often do, these architects had taken liberties in interpreting the competition program. The French bureaucrats who had originally promoted the idea of a modern people’s opera and who were advising the jury were having none of that. The bureaucrats had written a 423-page competition program minutely describing the new opera (including a schematic plan of the building), and they expected it to be slavishly followed. That is what Ott – and he alone – had done.

In the end, the French got what they wanted: not the most beautiful opera house in the world, but the biggest (despite its smaller seating capacity, the Opéra Bastille complex is three times larger than the Met) and technologically the most advanced. The French continue to have an abiding faith in new technology – which they often invent with considerable skill – and what is most innovative about the Opéra Bastille is not the architecture but the engineering. More than half of the Bastille site is taken up by enormous backstage facilities, which include not only a rehearsal hall, a mobile orchestra pit, a turntable, and a mobile stage that is also an elevator but also eleven ancillary scenery stages on two levels, joined together by an automated system of motorized trolleys. The purpose of all this space and machinery is to permit the rapid rotation of different operas: while one is being performed, another can be in rehearsal, and scenery for a third can be made ready on the lower level. It is a marvel of engineering, and despite some opening-night mishaps it all does appear to function as intended.

Whether such complexity is really required in an opera house is another story. Moving scenery around at dizzying speeds was supposed to provide a larger repertoire and a more varied program – a different opera every night, as many as 450 performances annually! But, as Maryvonne de Saint-Pulgent, a French journalist, points out in her fascinating account, Le syndrome de l’opéra, in 1990 Parisian concert halls were trying to sell twelve thousand tickets nightly to an opera-going public that barely exceeded thirty thousand persons, each of whom would have had to go to a concert three times a week to keep the halls full. Hugues Gall, a Frenchman who was the director of Geneva’s Grand Théâtre, called the Opéra Bastille “the wrong answer to a problem that doesn’t exist.” (Now the wrong answer is Gall’s problem, too; earlier this year, he replaced Chung, who was fired as music director after months of well-publicized wrangling with the Opéra’s chairman.) There are already signs that in practice the people’s opera house will not function in a manner much different from opera houses in New York, Berlin, or Milan, except that so far it has presented fewer operas and fewer performances. After a 40 percent price hike in 1990, the price of a ticket is as expensive as it had been at the Palais Garnier; there’s an increasing reliance on stars (the leading role in Joan of Arc at the Stake was taken by Isabelle Huppert, a popular film actress); and the second season included The Magic Flute – pace Barenboim.

The Met, La Scala, and Covent Garden are merely opera houses – the Opéra Bastille is a grand projet. The Big Projects – there are nine of them – refer to a series of monumental architectural works in Paris undertaken by Mitterrand since his election in 1981. Mitterrand, the impact of whose presidency on the city has been compared to the grand siècle of Louis XIV, is an enthusiastic builder of somewhat erratic taste whose ambition vastly exceeds that of his immediate predecessors. Charles de Gaulle rebuilt Paris after the war but added little that was new except the donut-shaped Maison de la Radio, a broadcasting center; Georges Pompidou built highways along the Seine and replaced the market of Les Halles with the Centre Pompidou, which today, paint peeling and steel rusting, resembles an oil refinery more than ever; and Valéry Giscard d’Estaing converted the vast Gare d’Orsay into a polyglot museum of the nineteenth century. So far, in addition to building the new opera house, Mitterrand has moved the Ministry of Finance out of the Louvre and into a new building, renovated the Louvre itself, and endowed Paris with something called the Arab World Institute. At La Villette, on the northeast edge of the city, he has had built a music center and a park of architectural follies, and at La Défense, in the northwestern suburbs, he has erected an office building in the shape of an arch, a modern counterpart to the Arc de Triomphe. Construction has recently begun on an enormous new national library, a controversial building that will add over $1 billion dollars to the $3 billion that has already been spent on the grands projets.

If ever there was an argument against the hoary notion that each generation must feel an obligation to make its own distinct architectural contribution “symbolic of its time,” the Big Projects is it. With the exception of I.M. Pei’s elegant glass pyramid in the courtyard of the Louvre, and some of the historic restorations at La Villette (which were begun by Giscard), Mitterrand’s grands projets are not great architecture. The grandiose library will resemble four half-open books, a banal and simpleminded concept; the Parc de la Villette is a collection of silly-looking pavilions set amid arid landscaping; the new Ministry of Finance is an exercise in the kind of monumental modernism that has long been discredited elsewhere; and the bombastic government office building at La Défense is less like a triumphal arch than a huge, marble-clad coffee table. Unfortunately, Mitterrand is not Louis XIV, or rather, his architects haven’t lived up to the standards set by Claude Perrault’s east front of the Louvre, Jules Hardouin-Mansart’s Dôme des Invalides, and André Le Nôtre’s Tuileries Garden.

Or even to the standards of Charles Garnier, the designer of the old opera house. Garnier, like Ott, came out of nowhere to win an architectural competition for a new opera house and likewise did so at a tender age – both were thirty-six – and with little previous experience. Garnier also had to navigate the treacherous shoals of French politics in order to see his ideas realized, although it took him somewhat longer – thirteen years compared with Ott’s six-year odyssey. But Garnier’s was a different time. His opera house included innovations such as a cast-iron roof structure and an unusual foundation, but these were hidden behind a marble architecture of eclectic richness. In the nineteenth century, going to the opera was chiefly a social occasion, and Garnier devoted considerably more space to sumptuous, mirror-lined lobbies and a grand staircase than to the hall itself. Technical efficiency was given distinctly second place: the backstage areas are spartan, and a quarter of the seats have an inadequate view of the stage. Nevertheless, it is a building that, while it was criticized at first, eventually captured people’s affection. “I remember being disarmed by the warm, comforting acoustics of the Palais Garnier,” recalls Kaptainis. “The sound, at least in the good seats, was magnificent.” Perhaps one day, the Opéra Bastille, too, will evoke such sentimental reminiscences – time can be the architect’s best friend – but I wouldn’t count on it.

The much-loved Palais Garnier was merged with the Opéra Bastille, and following a complete restoration it is used for ballet and occasional operas, especially seasonal favorites such as La Cenerentola. As The Wall Street Journal observed, audiences “tend to dress up more for performances at the Garnier than at the 20th-century granite and glass opera house at Bastille.”